Sustainable Sanitation Systems

SUSTAINABLE SANITATION SYSTEMS:

UNDERSTANDING PRIORITIES, PROCESSES, AND PATHWAYS TO SUCCESS

ALLIE DAVIS

M.S. Civil Engineering, University of Colorado Boulder, 2017

B.S. Civil Engineering, California Polytechnic State University, 2015

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirement of the degree of

Doctor of Philosophy

Mortenson Center in Global Engineering

Department of Civil, Environmental, and Architectural Engineering

2019

This thesis entitled: Sustainable Sanitation Systems: Understanding Priorities, Processes, and Pathways to Success written by Allie Davis has been approved for the Department of Civil, Environmental, and Architectural Engineering

______Dr. Sherri Cook University of Colorado Boulder (co-committee chair)

______Dr. Amy Javernick-Will University of Colorado Boulder (co-committee chair)

______Dr. Elizabeth Jordan United States Agency for International Development

______Dr. Leidy Klotz University of Virginia

______Dr. Anu Ramaswami University of Minnesota

______Dr. JoAnn Silverstein University of Colorado Boulder

Date: ______

The final copy of this thesis has been examined by the signatories, and we find that both the content and the form meet acceptable presentation standards of scholarly work in the above- mentioned discipline.

IRB Protocol #16-0026

ii Abstract

Davis, Allie (PhD in Civil Engineering; Department of Civil, Environmental, and Architectural

Engineering)

Sustainable Sanitation Systems: Understanding Priorities, Processes, and Pathways to Success

Dissertation directed by Assistant Professor Sherri M. Cook and Associate Professor Amy

Javernick-Will

Despite the unprecedented time, attention, and finances dedicated to increasing global access to sanitation, sanitation systems continue to fail at unacceptably high rates. Sanitation failure is a global concern, especially due to diminished public and environmental health and heightened economic costs. A commonly cited reason for sanitation failure is the selection of inappropriate technologies that do not address local priorities. To further evaluate this possible reason for failure, this dissertation employed a multi-method approach in 20 resource-limited communities in southern India with small-scale sanitation systems. Since sanitation priorities are mostly unknown, this research first focused on how to best identify a community’s priorities.

Findings demonstrated that interviews with community members most effectively identified sanitation priorities, contrary to common practices of using focus groups or interviewing community leaders. In addition, most priorities, and their ranked order, were community-specific, highlighting the need to conduct priority assessments in each community. If priorities are important for sanitation systems to be used and maintained, there need to be better ways to evaluate the potential of different sanitation technologies to address these priorities. As a result, a new social sustainability assessment was created to evaluate the ability of existing sanitation systems to address priorities and the potential of sanitation technologies to improve addressment. Application of this protocol showed that addressment could be improved by ensuring systems function

iii according to their intended designs, increasing knowledge of priorities in planning, or adding resource recovery when communities value those benefits. To better understand how priority addressment, and other factors, contribute to sanitation failure and success, a qualitative comparative analysis was conducted. Results revealed that sanitation success required adequate maintenance funds, clear maintenance plans, technical support, stakeholder engagement, and community buy-in; failed systems lacked these important factors. Finally, to evaluate how sanitation sustainability is and should be measured, six frameworks were evaluated and applied to implemented systems. While these frameworks include multiple pillars of sustainability, numerous and varied indicators are used, and results do not clearly distinguish between sustainable and unsustainable systems. Therefore, framework effectiveness could be improved by ensuring indicators have clear metrics, sustainable sanitation is defined, and frameworks account for context-specific differences. Overall, the results from this dissertation provide tools and recommendations to help implementing organizations, communities, and municipalities improve sustainable sanitation in resource-limited communities.

Dedication

To my grandma, Margy Giles.

Educator. Traveler. Pray-er. Independent Woman.

v Acknowledgements

My work and my time in graduate school has been shaped by so many individuals, too many to fully name and adequately thank. I cried every time I began to write this, a reflection of how deeply impactful the many people were who have shaped me these four years. I would like to highlight the immeasurable support and guidance from a few of these wonderful individuals.

First, to my advisors, Sherri Cook and Amy Javernick-Will. The chance to work with these incredible women was the reason I came to CU and the reason I pursued a PhD. You set an incredibly high bar for me and my work, and I am so grateful that you always held me to it. You believed in me, advocated for me, pushed me, guided me, encouraged me, invested in me, celebrated me, and taught me. Your guidance has strengthened my research skills, critical thinking, and confidence as a researcher and as a person. I can’t believe how much time you both dedicated to me weekly and that you never tired of reading countless versions of all my papers; your support and mentorship are unparalleled. I am so lucky to have had the chance to pursue research that I am passionate about with two advisors who work well together and complement each other perfectly.

You have left an indelible mark on me. It has been a true pleasure to work with the both of you.

I am also deeply grateful to my committee, Dr. Liz Jordan, Dr. JoAnn Silverstein, Dr. Leidy

Klotz, and Dr. Anu Ramaswami. Your insight and our discussions have proven immensely fruitful.

I am thankful for your time and contributions to my work and my professional development. I am so thankful to have a committee who has consistently pushed me to think deeper and has always been so invested and excited about my work.

One of the greatest highlights of graduate school has been to know and learn from Rita

Klees. If I have half the career and impact you have, then I will consider myself immensely lucky.

For the past four years, you have been my guiding star for my personal and professional dreams.

vi You have had far-reaching impact on the global stage and remain so humble and excited to invest in the next generation of leaders. Any success that I have had has been touched by you, and I am deeply grateful for your encouragement, networking on my behalf, letters of recommendation, guidance, and inspiration.

I have also been extremely lucky to have been a part of two wonderful research groups.

These individuals collectively have provided hundreds of hours of feedback on my work and have made me so much better for it. They read multiple drafts of every paper, watched every practice conference presentation or seminar, grilled me for interviews, commiserated amid disappointment, and celebrated in my successes. For this, I am thankful to: Andrew Tracy, Shaye Palagi, Casie

Venable, Aaron Opdyke, Erin Arneson, Cristina Poleacovschi, Eric Antillon, Alex Belenguer-

Gonzalez, Alex Zerio, James Harper, Caleb Cord, Pranav Chintalapati, Kimmy Pugel, Briar

Goldwyn, Kyle Kwiatkowski, Topher Jones, Pranoti Kikale, Kyle Thompson, Michelle Solomon,

Eric Pederson, Katie Chambers, Nick Valcourt, and Wesam Beitelmal.

This research would not have been possible without the support and guidance of several individuals and organizations in the WASH sector. Notably, Lukas Ulrich and his 4S project team from Eawag-Sandec, BORDA, and CDD Society fostered many of my initial connections to sanitation implementing organizations in India. Lukas also provided great insight into my research design and analyses, always pushing me to strengthen my defense of my methods. I am, however, most grateful to Lukas for the countless conversations and masala dosas we shared that helped me survive even the most difficult moments of my fieldwork. Marius Klinger, Geeta Singhal, and

Rohit Chandragiri also provided much-needed optimism, encouragement, and friendship during my fieldwork in India and the many months since then.

vii This research would also have certainly been impossible without the dedication and hard work of my two research assistants, Sridhar Selvaraj and Vijay Kumar. They both provided incredible insight into our work, willingly woke up at 4 am to commute for hours by bus or train alongside me to a different location nearly every day, braved the 43˚C heat of Trichy in April, persisted to secure interviews and meetings when even I was close to giving up, hunted down increasingly spicier food, and taught me so much about the beauties and complexities of southern

India—all memories I will cherish forever. I also had the pleasure of working with three undergraduate research assistants, Garrett Geer, Vanessa Thompson, and Tesia Golec. Their hard work and dedication made this work possible, I am indebted to their contributions.

I am also deeply thankful for the hundreds of individuals in India who hold together the foundation of this research. The communities, municipalities, implementing organizations, and universities taught me far more than I can ever repay. I hope my work reflects the brilliance and perseverance of many who are hard at work providing sanitation in India and that these findings can help increase safe sanitation access in the future.

In addition to these wonderful mentors, partners, collaborators, and teachers, I am surrounded by a community of people who keep me moving forward daily. First, and most importantly, to my family. I love you all and could not have begun to imagine this journey without you, your support, and the hundreds of ways you have removed obstacles and given me immense privileges. To my dad, Rob Davis. I never imagined to follow in your footsteps to pursue a PhD, let alone at CU Boulder. I realize now this may have implicitly always been the path. Your dedication, hard work, honesty, and leadership have long guided what I do and the decisions I make. I am grateful to attend an institution that has grown and benefited from your service, commitment to diversity, high standards, and ambitious vision. I am also grateful you listened

viii when I said there could only be one Davis on campus (at least, one each campus). It has been such a gift to share weekly lunches and summer backpacking—for this, a detour to graduate school has been so worth it. To my mom, Shirley Davis. You raised us to be confident, independent, and resilient. I did not believe you in high school when you told me I was a natural leader, but perhaps

I had yet to emerge from my Dark Ages. Your leadership was what allowed me to recognize those abilities in myself. I am so thankful for the millions of opportunities you have given me and for your unconditional love and support. I have loved having you witness my graduate school journey and know my friends. Thank you for feeding me, doing my laundry, loaning me the car, and listening to my highs and lows, all while you expanded your career and remained the most hospitable person I know. You read every word of my comprehensive proposal and will be one of the very few to read every word of this dissertation. I look forward to one day attending your PhD defense. To my sister, Grub Davis. You are the true academic in the family and technically my first research adviser (although, sanitation research was quite the departure from catching capuchins and chasing after spider monkey pee in Panama). I wish it had been possible to also attend the same graduate school, but I am so grateful to have had many chances to visit each other and hike, travel, and dive together. Any time someone commends my productivity or capacity to balance a million things, I think that little do they know how much more you accomplish. I have fed off of your laser-focused passion and enthusiasm and have now found my own. You have always been there, and you will always remain my greatest inspiration, best friend, and twin.

A mentor of mine in undergrad, Chip Appel, accurately predicted that graduate school would bring some of the most fruitful and long-lasting relationships of my life. To Andrew Tracy.

I lack the words to do justice to the depth and meaning of your friendship. You have been by my side during some of the hardest and best moments of my entire life. You have already given me a

ix lifetime worth of friendship, and I am ever so grateful to see that grow forever. I am possibly even more grateful that you share my (un)healthy obsession with politics and are always ready to listen to my litany of foreign policy facts and government dreams. You are family to me, and your belief in me means more than I can say. To Shaye Palagi. It was such a gift when Amy decided to take us on at the same time. I cannot imagine this journey without passing through nearly every milestone in unison, getting to celebrate each other’s accomplishments and returns in a notoriously disruptive way. You are a constant source of inspiration for what it means to be a strong woman, a thoughtful citizen, and a compassionate human. You may be the only one who actually understands theoretical contributions, and I have learned so much from your deep thinking and zest for impactful research. You have a beautiful and deep spirit and have been one of my greatest encouragers and cheerleaders, for which I am so grateful. To Casie Venable. I have learned more about pop culture and college basketball than I ever expected when starting graduate school. You have a brilliant mind and such a high capacity to understand so many disparate and complex topics.

I have loved sharing the high-intensity academia, fun and competitive bartending, and soft and vulnerable personal pieces of my world with you. You are an inspiration of confidence, will-power, and goal-driven pursuit. I am so grateful Boulder did not entirely scare you away because I truly treasure your friendship. See you in D.C. To Taygra Longstaff. Even a month in it felt like I had known you for a lifetime. You are my kindred spirit and level-headed counsel. You share so much of my perspective, thought process, and personality type and this has grown my confidence in myself so much. You have pushed me to embrace my strengths and pursue my dreams. I am so grateful for your loyalty and your friendship. To Erin Arneson. You brought wisdom and humor to our pod. I am so thankful for how you stand up for your beliefs and protect the people you care about. You are an incredible teacher and friend. To Sid Bhandari. It has been so wonderful to move

x beyond simply being officemates to becoming close friends. Your depressing thoughts of the day have uplifted my spirits, your advice has guided my finish, your enthusiasm and investment has made me feel celebrated.

And finally, to the ones outside of school who have held me together for these four years and beyond. To my Grace Modisett, you have known me longer than most in this list. Your humor, love, support, cheerleading, counsel, and steadiness pushed me through to the finish line and constantly reminded me of the beautiful world that exists outside of academia. I have been so spoiled to have had another set of years living near you and sharing life with you. To Clam

Shannon and Jessica Jansen. You will laugh, but you have been my wise women council and counsel for the past eight years. It has meant more than the world to me to have your constant support and friendship and to be a part of a strong woman community (and no end to absurd humor and entertainment) that makes it so hard to imagine finding such a community elsewhere. To Emily

Crider. You have been such a close friend and mentor to me, and it has been so special to watch the ways that both of our lives have transformed over the past four years. I am thankful for your leadership, compassion, generosity, confidence, stubbornness, wisdom, faith, and son Owen. To

Mike Tyska. You joined in near the peak of the whirlwind and have been such a steady source of optimism and encouragement. Thank you for believing in me, being patient with me, and making the final stretch so fun. To my many other friends and colleagues whose time, support, and encouragement have and will continue to shape me forever, I am so grateful: Evan Dicks, Kaity

Mattos, Matt Bentley, Ruth Bentley, Lilly Bussell, Julia Core, Thad Webb, Laura MacDonald,

Sharaya Jones, Vidya Venkataramanan, Nina Ball, Annie Rumbles, Ellie Moeller, Erica Kaminski,

Jessy Darnell, Kiersten Hill, Neely Rettig, Salome Philip, and Sarah D’Souza.

xi Funding for this research was provided by: the Mortenson Center in Global Engineering, the Beverly Sears Graduate Student Research Grant, the Engineering Excellence Fund, Rotary

District 5450, Eawag-Sandec, the College of Engineering & Applied Science at CU Boulder, the

Department of Civil, Environmental, and Architectural Engineering at CU Boulder, the CU

Boulder Graduate School, and the United Government of Graduate Students at CU Boulder. Any opinions, findings, and conclusions or recommendations expressed in this material are those of

Allie Davis and do not necessarily reflect the views of the funding organizations.

xii Table of Contents

Chapter 1: Introduction ...... 1 Chapter 2: A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems ...... 9 Chapter 3: Priority Addressment Protocol: Understanding the ability and potential of sanitation systems to address priorities ...... 44 Chapter 4: The use of qualitative comparative analysis to identify pathways to successful and failed sanitation systems ...... 76 Chapter 5: Analyzing sanitation sustainability assessment frameworks to identify ways to improve how sanitation sustainability is measured in resource-limited communities ...... 109 Chapter 6: Research Contributions ...... 140 Complete Dissertation References ...... 151 Appendix A: Glossary of Terms ...... 172 Appendix B: Case Summaries ...... 174 Appendix C: Interview and Survey Questionnaire Scripts ...... 184 Appendix D: Observation Checklist ...... 204 Appendix E: Published Supporting Information for Chapter 2 ...... 223 Appendix F: Published Supporting Information for Chapter 3 ...... 239 Appendix G: Published and Expanded Supporting Information for Chapter 4 ...... 259 Appendix H: Supporting Information for Chapter 5...... 311 Appendix I: Reflections, Lessons Learned, and Suggestions ...... 341

xiii List of Tables

Table 1.1 Summary of literature gaps, research questions, and dissertation chapters...... 6

Table 2.1. Percent participation by respondent sex and community roles for each priority identification method in all 20 communities...... 18

Table 2.2. Comparison of the average time and the resources needed for interviews, photovoice, and focus groups in all 20 communities...... 30

Table 2.3. Results of a comparison of the most effective role from all 20 communities based on identification of the greatest number of unique priorities ...... 34

Table 3.1. Description of Current, Intended Design, and Added Resource Recovery Scenarios. 51

Table 3.2. The 52 sanitation priorities identified across all 20 cases, and the number of cases that shared each priority ...... 61

Table 4.1. Causal conditions hypothesized to influence sanitation system success and failure. . 85

Table 4.2. The truth table summarizes the fuzzy scores for each causal condition and the outcome for all 20 cases...... 89

Table 5.1. Summary of the six sanitation sustainability frameworks selected...... 113

Table 5.2. Summary of the main types of indicators included in the selected sanitation sustainability assessment frameworks ...... 118

Table 5.3. Summary of the results from TAF...... 128

Table 6.1. Summary of the results and research contributions from each chapter...... 141

Table A1. Glossary of terms used throughout the dissertation...... 17272

Table C1. Summary of interview, photovoice, and focus group scripts, durations, and topics...... 18484

Table E1. Summary of community and sanitation system information...... 22424

Table E2. Participation demographics for all communities ...... 225

xiv Table E3. Participation demographics for AHP groups for all 20 communities...... 2266

Table E4. Summary of the total number of unique community and unique sanitation priorities identified in each community...... 2277

Table E5. Comparison of method performance in all 20 communities for identifying the total number of unique community priorities and sanitation priorities ...... 2277

Table E6. Comparison of method performance in all 20 communities for identifying the total number of unique physical priorities and unique abstract priorities ...... 2288

Table E7. Comparison of method performance in all 20 communities for identifying the total number of unique community priorities and sanitation priorities ...... 2288

Table E8. Results of a comparison of the most effective role from all 20 communities based on identification of the greatest number of the most important priorities...... 232

Table E9. Results of the effectiveness of respondent sexes for identifying the most unique priorities...... 2377

Table E10. Results of the most effective methods to use with each community role...... 2377

Table E11. Results of the most effective methods to use with each sex...... 2388

Table F1. Summary of each case’s sanitation system.a ...... 24040

Table F2. Coding dictionary for sanitation priorities...... 24141

Table F3. Coding dictionary for community priorities...... 24444

Table F4. Participation demographics for interviews evaluating the level of priority addressment for all 20 cases...... 2466

Table F5. Changes to sanitation priorities’ addressed values from the Current Scenario to the Intended Design Scenario...... 2477

Table F6. Changes to community priorities’ addressed values from the Current Scenario to the Intended Design Scenario...... 2488

Table F7. Changes to sanitation priorities’ addressed values from the Intended Design Scenario to the Added Resource Recovery Scenario...... 2499

xv Table F8. Changes to community priorities’ addressed values from the Intended Design Scenario to the Added Resource Recovery Scenario...... 25151

Table F9. Comparison of all cases’ total case sanitation scores between scenarios...... 25353

Table F10. Comparison of total case sanitation scores: Percent difference for the change in score from the Current and Intended Design Scenarios for each case ...... 25353

Table F11. Summary of community priorities from all 20 cases...... 25454

Table F12. Comparison of total case community scores: Percent difference for the change in score from the Current and Intended Design Scenarios for each case ...... 25555

Table F13. Results of comparison of total case community scores between scenarios for all cases...... 2566

Table F14. Comparison of the average and standard deviation of the percent differences between Added Resource Recovery Scenario total case scores when considering uncertainty...... 2588

Table G1. For all 20 cases: sanitation technology, implementing organization type, operation and maintenance (O&M) manager, outcome, and case summary...... 26060

Table G2. Summary of respondent demographics for interviews conducted to characterize causal conditions and outcomes for all 20 cases...... 27070

Table G3. Summary of system performance based on water quality analyses...... 27171

Table G4. List of hypothesized causal conditions from literature and case knowledge that may influence sanitation system success or failure...... 27272

Table G5. Indirect calibration for Behavior Change Education...... 27474

Table G6. Indirect calibration for Clear O&M Plan...... 274

Table G7. Indirect calibration for Community Participation in Planning...... 2766

Table G8. Indirect calibration for Construction Quality...... 2777

Table G9. Indirect calibration for Government Barriers...... 2777

Table G10. Indirect calibration for Municipality Involved in Planning...... 2788

xvi Table G11. Indirect calibration for Skilled Operator...... 2799

Table G12. Indirect calibration for External Technical Assistance...... 2799

Table G13. Summary of causal conditions’ indirect calibrations...... 28080

Table G14. Indirect Calibration for Performance...... 28585

Table G15. Necessity scores for hypothesized causal conditions for success that were excluded from the analysis because necessity scores were below 0.30...... 2899

Table G16. Summary of simplifying assumptions for success and failure of sanitation systems...... 29393

Table G17. Necessity and sufficiency of conditions for success...... 29494

Table G18. Preliminary pathways to success. The solution presented is the intermediate solution obtained using fs/QCA software (Ragin et al., 2017)...... 29494

Table G19. Final pathways to success. The solution presented is the intermediate solution obtained using fs/QCA software (Ragin et al., 2017)...... 2999

Table G20. Necessity and sufficiency of conditions for failure...... 300

Table G21. Preliminary pathways to failure...... 301

Table G22. Final pathways to failure...... 3066

Table H1. Summary of sanitation systems used to apply the sanitation sustainability frameworks...... 31212

Table H2. Complete list of indicators and adaptations from all six sustainability frameworks...... 31313

Table H3. Summary of numbers of indicators by the selected frameworks...... 33131

Table H4. Summary of the level of adaptation required by the selected frameworks...... 33131

Table H5. Summary of the total amount of adapted indicators based on the sustainability pillars...... 33232

xvii Table H6. Summary of the results from the UNICEF Sustainability Checks Framework .... 33333

Table H7. Summary of the results from the CDC/ARC WASH framework ...... 33434

Table H8. Summary of the results from the SEI Sustainability Checks ...... 33535

Table H9. Comparison of rankings from 1 (best) to 12 (worst) across selected frameworks. 3386

Table H10. Definitions of sustainability adopted by the six selected frameworks...... 3399

xviii List of Figures

Figure 2.1. Overview of the data collection and analysis methods used to compare the effectiveness of three priority identification methods and evaluate the impact of respondent sex and community role on effectiveness...... 15

Figure 2.2. Results of the percent difference between methods for identifying the greatest number of (a) unique sanitation priorities and (b) unique community priorities ...... 26

Figure 2.3. Results of the number of unique community priorities identified in interviews by each community role (community leader, community members, system manager)...... 33

Figure 3.1. Overview of the priority addressment protocol used to identify priorities and evaluate the current ability and potential of different sanitation systems to address priorities...... 49

Figure 4.1. Overview of data collection and analysis methods used to identify pathways to success and failure of sanitation systems...... 80

Figure 4.2. Results of the combinations of conditions that led to (a) success and (b) failure ..... 92

Figure 5.1. Comparison of weighted (a) and unweighted (b) results from the single-score aggregation framework, the Composite Indicator Approach...... 130

Figure 5.2. Ranking from 1 (best) to 12 (worst) for sanitation systems from the selected frameworks...... 132

Figure E1. Results of average method performance across all 20 communities for identifying the top priorities ...... 230

Figure E2. Comparison of the number of unique priorities identified and the time required for data collection for interviews, photovoice, and focus groups ...... 231

Figure E3. Results of community roles for identifying unique sanitation priorities through interviews for (a) community priorities and (b) sanitation priorities...... 233

xix Figure E4. Results of the number of unique sanitation priorities identified by all community members in interviews (n = 7 to 12) and the average community member in an interview for (a) community priorities and (b) sanitation priorities ...... 234

Figure E5. Results of the number of unique sanitation priorities identified in interviews by each community role (community leader, community members, system manager)...... 235

Figure E6. Results for identifying sanitation priorities when speaking with community leaders for (a) community priorities and (b) sanitation priorities ...... 236

Figure F1. Percent of each case’s population that used the sanitation system...... 252

Figure F2. Results of the AHP relative importance values for the most common sanitation priorities ...... 252

Figure F3. Results of the AHP relative importance values for community priorities that were expressed by 10 or more cases ...... 255

Figure F4. Comparison of the ability of the Current, Intended Design, and Added Resource Recovery Scenario to address community priorities...... 255

Figure F5. Comparison of the three scenarios’ average total case sanitation (a) and community (b) scores from all 20 cases ...... 257

Figure G1. (a) Anchor and crossover points for direct calibration for Addressed Sanitation Priorities. (b) Direct Calibration for Addressed Sanitation Priorities...... 283

Figure G2. (a) Anchor and crossover points for direct calibration for Sufficient O&M Funds. (b) Direct Calibration for Sufficient O&M Funds...... 284

Figure G3. (a) Anchor and crossover points for direct calibration for Use. (b) Direct Calibration for Use...... 286

Figure G4. (a) Anchor and crossover points for direct calibration for Maintenance. (b) Direct Calibration for Maintenance...... 286

Figure H1. Comparison of weighted and unweighted results from the TechSelect 1.0 Framework (Kalbar et al. 2012a)...... 337

xx Chapter 1: Introduction

Despite the unprecedented time, attention, and financial investment dedicated to increasing global access to sanitation (WSP 2017), sanitation systems continue to fail at unacceptably high rates. In resource-limited communities, up to 70 percent of sanitation systems fail within two years of construction (WHO and UNICEF 2017). This sanitation failure inhibits access to safely managed sanitation (United Nations 2015a) and leads to diminished public and environmental health (Mara et al. 2010) and heightened economic costs (WSP 2017). In response, the United

Nations’ Sustainable Development Goals (SDGs) have named universal access to safe sanitation as a primary goal (United Nations 2015a). To achieve universal access, it is important to enhance our understanding of the causes of sanitation failure and how to implement successful sanitation systems.

A lack of alignment between implemented sanitation systems and local priorities may be one reason for sanitation failure, as communities are unwilling to accept systems that do not meet their needs. Despite a recognition that identifying sanitation priorities is an important step in sanitation planning (Freudenberger 2011; Seymour 2014), an understanding of the most effective methods to identify priorities is still needed. To improve the effectiveness of priority assessments, there is a need to better understand which methods capture community priorities, including their most important priorities. Chapter 2 of this dissertation compared two priority identification methods used commonly by sanitation implementing organizations interviews and focus groups, and one method gaining traction in the sanitation sector, photovoice, to evaluate their effectiveness for identifying the most comprehensive list of sanitation priorities and for identifying the most important sanitation and community priorities. In priority identification processes, a respondent’s personal attributes such as sex or community role (leader, member, sanitation system manager) are

1 likely to influence a method’s effectiveness at capturing priorities (Kamyotra and Bhardwaj 2011;

Leder 2007; Wankhade 2015). Therefore, Chapter 2 also evaluated differences in respondent attributes (community roles and respondent sex) for priority identification. An understanding of the most effective methods to identify sanitation priorities can help implementing organizations efficiently use their limited resources to better address priorities, increase sanitation acceptance and success, and improve human and environmental health.

In addition to improving priority assessment processes, the sanitation sector must also understand the ability of existing sanitation technologies to address priorities. Systems that align with local priorities are thought to be strong determinants of continued use and maintenance

(Murphy et al. 2009; Palaniappan et al. 2008). In addition, the potential of different technologies to address priorities is needed. For instance, sanitation systems that recovery energy, nutrients, or water from wastewater (i.e., resource recovery systems) are thought to provide greater benefits than conventional systems (i.e., traditional sanitation systems without resource recovery) (Cornejo et al. 2013; Walekhwa et al. 2014). Yet, the potential benefits of resource recovery systems have not been evaluated within the context of a community’s goals and priorities. To address these needs, Chapter 3 created a “priority addressment protocol” that (1) identifies sanitation and community priorities; (2) evaluates how well existing sanitation systems address those priorities; and (3) analyzes the potential for conventional and resource recovery systems to improve priority addressment. This social sustainability assessment uses a quantitative score to measure priority addressment based on how well or poorly each sanitation system did—or could—address a community’s ranked lists of sanitation and community priorities. The results and new protocol can help identify the most appropriate sanitation systems and design improvements that align with community priorities to encourage greater sanitation acceptance, use, and success.

2 In addition to appropriate sanitation technologies that address context-specific priorities, many other factors may influence sanitation success or failure. For instance, sanitation success may be attributed to community participation in planning (Roma and Jeffrey 2010), affordability

(Mwirigi et al. 2014), maintenance support (Sansom 2011), and low maintenance complexity

(Brikké and Bredero 2003). Sanitation failure has been attributed to supply-driven approaches

(Starkl et al. 2013b), lack of maintenance (Katukiza et al. 2010), faulty designs (Sujaritpong and

Nitivattananon 2009), high costs (Cronin et al. 2014), lack of ongoing support (Eales et al. 2013), or inappropriate technologies (Murphy et al. 2009). However, persistent sanitation failure suggests that these factors alone do not fully explain failure causes. Analyzing the influence of individual factors on sanitation success or failure has not fully explained the causes of these outcomes.

Instead, comprehensive evaluations that evaluate combinations of multiple factors are needed to better understand how they influence sanitation success or failure. Therefore, Chapter 4 used qualitative comparative analysis (QCA) to determine combinations of factors (i.e., pathways) that led to sanitation system success or failure. These pathways can guide implementing organizations, communities, and governments on how to focus their limited resources on the factors that avoid failure-prone scenarios and promote success.

Beyond sanitation success, there is also a need to improve sanitation system sustainability, where systems minimize negative social, economic, and environmental impacts. There are numerous sanitation sustainability frameworks—tools to measure or compare sustainability— developed both in research and practice (Balkema et al. 2002; Lennartsson et al. 2009; Molinos-

Senante et al. 2014; UNICEF 2012). However, there is not consensus on a sustainability framework to use or build upon. Also, despite this plethora of frameworks, new sustainability frameworks continue to be developed as sanitation systems continue to fail. The limited use and

3 impact of existing frameworks may be because frameworks are not user friendly, are too complex, or require difficult data collection; are not easily adaptable to different contexts; lack alignment on indicators; and/or do not adequately or comprehensively measure sustainability. Therefore, to better understand how sanitation sustainability is currently measured by sustainability frameworks and how they might be improved, in Chapter 5, I evaluated six common sanitation sustainability frameworks to identify the ability of the frameworks to measure sanitation sustainability. I also applied these frameworks to successful and failed systems to generate further insight on the essential and most effective elements of sustainability frameworks. The results provide suggestions for framework improvements, which can increase framework use and ultimately sanitation sustainability overall.

Research Context Presently, India is home to the world’s fastest growing population but still faces some of the most significant challenges for successful sanitation. In India, 60 percent of the population lacks access to safely managed sanitation (WHO and UNICEF 2017). Less than 20 percent of wastewater generated is treated (Kamyotra and Bhardwaj 2011), 67 percent of the country still uses biomass for cooking fuel, and 23 percent of the population lacks access to electricity (WHO and UNICEF 2015). The central government has responded with ambitious national initiatives to increase sanitation access and end open defecation by October 2019 (Government of India 2016).

Improved approaches, such as robust methods to identify and address contextualized priorities, and identification of combined factors that enable success, are needed to improve sanitation systems in resource-limited contexts, such as India. To evaluate priorities, pathways to success and failure, and sanitation sustainability, 20 resource-limited communities in Karnataka and Tamil

Nadu, India were selected. Communities had the following similarities: all were peri-urban slum resettlements; had 800 to 1000 residents who were day laborers, had low incomes, and were from

4 India’s lowest castes; and had community-based sanitation systems, where community members were expected to be responsible for all or most of the maintenance needs. Overall, there were five successful and five failed conventional sanitation systems and five successful and five failed resource recovery systems. Resource recovery technologies included DEWATS (a type of decentralized wastewater treatment system with a biogas digester, baffled reactor, and gravel filter), Ecological Sanitation (EcoSan) urine diverting dry toilets, and settling tanks with gravel filters. Non-resource recovery technologies include DEWATS (with settling tanks instead of digesters) and settling tanks with gravel filters. More detailed community summaries are included in Appendix B.

Dissertation Organization Table 1.1 summarizes the research gaps, research questions, and the organization of this dissertation, which is written such that each chapter is a standalone journal article. Chapters 2, 3, and 4 are already published; Chapter 5 is in preparation for publication submission. Please cite the journal articles instead of this dissertation.

5 Table 1.1. Summary of literature gaps, research questions, and dissertation chapters. Supporting Chapter/Citation Gaps/Needs Research Questions Information Chapter 2: Davis, A., Javernick-Will, A., To quickly assess priorities, implementing (1) Which methods are the most effective Cook, S., 2018. A comparison of interviews, organizations often rely on a single method for for identifying the most unique sanitation focus groups, and photovoice to identify priority assessment without understanding its priorities and community priorities? Appendix E: Published sanitation priorities and increase success of effectiveness. (2) What differences are there between supporting community-based sanitation systems. Since community perspectives may differ across the methods, community roles, and information for Environmental Science: Water Research & role and sex, there is a need to understand how a respondent sex for identifying a Chapter 2 Technology 4, 1451–1463. respondent’s role or sex influences priority comprehensive list of priorities and those https://doi.org/10.1039/C8EW00391B identification. most important to the community? (1) What are communities’ priorities for Sanitation systems that address local priorities are sanitation systems? Chapter 3: Davis, A., Javernick-Will, A., thought to increase use and acceptance, but the Cook, S., 2019. Priority Addressment ability of current sanitation technologies to address (2) To what extent do sanitation systems Appendix F: Protocol: Understanding the Ability and priorities is not known. currently address community and Published Potential of Sanitation Systems to Address sanitation priorities? supporting Different conventional and resource recovery Priorities. Environmental Science & information for technologies are implemented without (3) What is the potential for different Technology 53, 401–411. conventional and resource recovery Chapter 3 https://doi.org/10.1021/acs.est.8b04761 understanding how these technologies align with or diverge from a community’s goals and priorities. technologies to improve priority addressment? Previous research has investigated the causes of Chapter 4: Davis, A., Javernick-Will, A., sanitation success or failure by often focusing on Cook, S., 2019. The use of qualitative (1) What combinations of social, Appendix G: factors in isolation, leading to mixed conclusions on comparative analysis to identify pathways to institutional, economic, and technical Published and a factor’s influence. Comprehensive evaluations successful and failed sanitation systems. factors are associated with successful expanded that identify which combinations of social, Science of The Total Environment 663, 507– sanitation systems and failed sanitation supporting institutional, technical, and economic factors are 517. systems? information for needed to better understand sanitation success and https://doi.org/10.1016/j.scitotenv.2019.01.291 Chapter 4 failure. Despite the abundance of sanitation sustainability (1) How is sanitation sustainability Chapter 5: Analyzing sanitation sustainability frameworks, many frameworks are not used, and currently measured? assessment frameworks to identify ways to new frameworks are still being developed. There is Appendix H: improve how sanitation sustainability is a lack of understanding of how sanitation (2) How can sanitation sustainability Supporting measured in resource-limited communities sustainability is currently measured, what are the frameworks be improved to accurately information for essential and effective elements of a framework, and and comprehensively measure Chapter 5 (publication forthcoming) how frameworks could be improved to better and sustainability? more completely measure sustainability.

6 Chapter 1 References Balkema, A.J., Preisig, H.A., Otterpohl, R., Lambert, F., 2002. Indicators for the sustainability assessment of wastewater treatment systems. Urban Water 4, 153–161. https://doi.org/10.1016/S1462-0758(02)00014-6 Brikké, F., Bredero, M., 2003. Linking technology choice with operation and maintenance in the context of community water supply and sanitation. WHO and IRC WASH, Geneva, Switzerland. Cornejo, P.K., Zhang, Q., Mihelcic, J.R., 2013. Quantifying benefits of resource recovery from sanitation provision in a developing world setting. J. Environ. Manage. 131, 7–15. https://doi.org/10.1016/j.jenvman.2013.09.043 Cronin, A.A., Ohikata, M., Kumar, M., 2014. Social and economic cost-benefit analysis of sanitation in Odisha State, India. J. Water Sanit. Hyg. Dev. 4, 521. https://doi.org/10.2166/washdev.2014.150 Eales, K., Blackett, I., Siregar, R., Febriani, E., 2013. Review of Community-Managed Decentralized Wastewater Treatment Systems in Indonesia. WSP. Freudenberger, K.S., 2011. Rapid Rural Appraisal and Participatory Rural Appraisal: A Manual for CRS field Workers and Partners (Manual). Catholic Relief Services, Baltimore, Maryland. Government of India, 2016. Swachh Bharat [WWW Document]. Swachh Bharat Mission My Clean India. URL https://swachhbharat.mygov.in/ (accessed 9.9.16). Kamyotra, J.S., Bhardwaj, R.M., 2011. Municipal wastewater management in India. India Infrastruct. Rep. 299. Katukiza, A.Y., Ronteltap, M., Oleja, A., Niwagaba, C.B., Kansiime, F., Lens, P.N.L., 2010. Selection of sustainable sanitation technologies for urban slums — A case of Bwaise III in Kampala, Uganda. Sci. Total Environ. 409, 52–62. https://doi.org/10.1016/j.scitotenv.2010.09.032 Leder, S., 2007. Gender issues in water and sanitation programmes: Lessons from India. South Asian Water Stud. 5, 59–62. Lennartsson, M., Kvarnström, E., Lundberg, T., Buenfil, J., Sawyer, R., 2009. Comparing sanitation systems using sustainability criteria. Stockholm Environment Institute, SEI, Stockholm. Mara, D., Lane, J., Scott, B., Trouba, D., 2010. Sanitation and Health. PLoS Med. 7, 1–7. https://doi.org/10.1371/journal.pmed.1000363 Molinos-Senante, M., Gómez, T., Garrido-Baserba, M., Caballero, R., Sala-Garrido, R., 2014. Assessing the sustainability of small wastewater treatment systems: A composite indicator approach. Sci. Total Environ. 497–498, 607–617. https://doi.org/10.1016/j.scitotenv.2014.08.026 Murphy, H.M., McBean, E.A., Farahbakhsh, K., 2009. Appropriate technology – A comprehensive approach for water and sanitation in the developing world. Technol. Soc. 31, 158–167. https://doi.org/10.1016/j.techsoc.2009.03.010 Mwirigi, J., Balana, B.B., Mugisha, J., Walekhwa, P., Melamu, R., Nakami, S., Makenzi, P., 2014. Socio-economic hurdles to widespread adoption of small-scale biogas digesters in Sub- Saharan Africa: A review. Biomass Bioenergy 70, 17–25. https://doi.org/10.1016/j.biombioe.2014.02.018 Palaniappan, M., Lang, M., Gleick, P.H., 2008. A Review of Decsion-Making Support Tools in the Water, Sanitation, and Hygiene Sector. The Pacific Institute, Oakland, California.

7 Roma, E., Jeffrey, P., 2010. Evaluation of community participation in the implementation of community-based sanitation systems: a case study from Indonesia. Water Sci. Technol. 62, 1028–1036. https://doi.org/10.2166/wst.2010.344 Sansom, K., 2011. Complementary roles? NGO-Government relations for community-based sanitation in South Asia. Public Adm. Dev. 31, 282–293. https://doi.org/10.1002/pad.609 Seymour, Z., 2014. Sanitation in developing countries: a systematic review of user preferences and motivations. J. Water Sanit. Hyg. Dev. 4, 681–691. Starkl, M., Brunner, N., Stenström, T.-A., 2013. Why Do Water and Sanitation Systems for the Poor Still Fail? Policy Analysis in Economically Advanced Developing Countries. Environ. Sci. Technol. 47, 6102–6110. https://doi.org/10.1021/es3048416 Sujaritpong, S., Nitivattananon, V., 2009. Factors influencing wastewater management performance: Case study of housing estates in suburban Bangkok, Thailand. J. Environ. Manage. 90, 455–465. https://doi.org/10.1016/j.jenvman.2007.11.006 UNICEF, 2012. Sustainability Checks: the UNICEF Experience in Eastern and Southern Africa. UNICEF. United Nations, 2015. Transforming our world: the 2030 Agenda for Sustainable Development [WWW Document]. Sustain. Dev. Knowl. Platf. URL https://sustainabledevelopment.un.org/post2015/transformingourworld (accessed 11.19.15). Walekhwa, P.N., Lars, D., Mugisha, J., 2014. Economic viability of biogas energy production from family-sized digesters in Uganda. Biomass Bioenergy 70, 26–39. https://doi.org/10.1016/j.biombioe.2014.03.008 Wankhade, K., 2015. Urban sanitation in India: key shifts in the national policy frame. Environ. Urban. 27, 555–572. WHO, UNICEF, 2017. Progress on Drinking Water, Sanitation and Hygiene: Update and SDG Baselines. Joint Monitoring Program, Geneva. WHO, UNICEF, 2015. Progress on sanitation and drinking water: 2015 Update and MDG Report, Joint Monitoring Programme. World Health Organization, Geneva. WSP, 2017. Economics of Sanitation Initiative [WWW Document]. World Bank. URL https://www.wsp.org/content/economic-impacts-sanitation (accessed 11.9.17).

8 Chapter 2: A comparison of interviews, focus groups, and photovoice to

identify sanitation priorities and increase success of community-based

sanitation systems

This chapter is the published journal article: Davis, A.; A. Javernick-Will; and S. Cook. A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems. Environmental Science: Water Research & Technology, 2018, 4, 1451–1463. https://doi.org/10.1039/C8EW00391B.

Abstract Art

Semi-structured interviews with females and community members identified the greatest number of unique and most priorities in resource-limited communities.

Water Impact Statement

In order to quickly assess a resource-limited community’s priorities and sanitation needs, implementers often rely on a single method for priority assessment without understanding its effectiveness. This work identifies the most effective qualitative priority identification methods and uncovers the methods’ tradeoffs to ensure priorities are addressed, increase acceptance, and ultimately improve human health and water quality through successful sanitation projects.

9 Abstract Addressing priorities is essential for the success of water and sanitation projects, especially in resource-limited communities. In order to quickly assess a resource-limited community’s priorities and sanitation needs, sanitation implementers often rely on a single method for priority assessment without understanding its effectiveness, such as its ability to identify the greatest number of unique and most important priorities. To increase the sector’s exposure to and understanding of more comprehensive and qualitative priority identification methods, this paper compares three qualitative priority identification methods: interviews, focus groups, and photovoice. Each method was used to identify community priorities (e.g., jobs, electricity) and sanitation priorities (e.g., cost, cleanliness) in twenty communities in India. A cross-case comparison of the three methods for both community and sanitation priorities found that interviews identified the greatest number of unique priorities and most important priorities. Since many respondent attributes are expected to impact a method’s effectiveness, the impact of each respondent’s sex and community role (leader, member, sanitation system manager) was also evaluated. Regardless of method, community members voiced more priorities, including more of the most important priorities, than community leaders or sanitation system managers. Leaders and system managers, as well as males and females, identified similar numbers of unique and most important priorities. These results are in contrast with current practice and support the use of interviews and the inclusion of community members during priority identification efforts.

Environmental engineers, researchers, government agencies, and sanitation implementers can use these results to better identify priorities, which, if addressed, can increase community acceptance and ultimately minimize project failure.

10 Introduction Lack of alignment between projects and priorities has contributed to increased project failure because communities are unwilling to accept systems that do not meet their needs. In resource-limited communities, where failure and project misalignment have drastic consequences, identifying priorities effectively and efficiently is particularly important to manage scarce resources. Identifying and addressing priorities is essential for sanitation projects, where community acceptance of a system is vital for use, maintenance, and success (Murphy et al. 2009).

However, due to time and resource limitations, sanitation implementers (e.g., implementing organizations, environmental engineers, practitioners) are often unable to perform extended, ethnographic data collection and some lack the knowledge on how to most effectively identify priorities. This poses a significant barrier to increasing access to sanitation, and consequently to meeting the United Nation’s Sustainable Development Goals (United Nations 2015b).

Sanitation implementers typically use supply-driven strategies (Breslin 2003; Starkl et al.

2013b), such as short, structured questionnaires that focus on collecting information rapidly and favor quantitative information. However, these strategies often fail to comprehensively identify priorities, in part due to the methods’ rapidity (McNall and Foster-Fishman 2007) and in part because they do not incorporate diverse community perspectives (Alkire 2005; Wood et al. 2016).

The use of qualitative methods for priority identification, such as semi-structured interviews or focus groups, help to overcome this barrier by allowing for open-ended questions and context- specific responses (Ryan, M et al. 2001; Scheyvens and Storey 2003) and are thus becoming a recommended approach. Since qualitative approaches are time-intensive (Alsaawi 2014), it is especially important to evaluate the relative effectiveness and limitations of each to decide which qualitative methods are appropriate to identify sanitation priorities. Knowledge of effective methods can help sanitation implementers determine how to use their limited resources effectively

11 to better address priorities, to increase project acceptance and success, and ultimately to improve positive environmental and health impacts. Despite the need, a comparison between qualitative priority identification methods is lacking.

Of the qualitative methods used by sanitation implementers, the two most common are semi-structured interviews (referred herein as interviews) and focus groups (Scheyvens and Storey

2003; Watkins et al. 2012). Interviews pose open-ended questions to a representative group of respondents to obtain qualitative responses (Harrell and Bradley 2009). While interviews can uncover high-quality and rich data (Alsaawi 2014), they can be time-intensive in both data collection and analysis (Stokes and Bergin 2006). In comparison, the focus group is a method where the organizer assembles a representative group of respondents and facilitates a discussion in response to preset prompts (Harrell and Bradley 2009). Focus groups are often used instead of interviews to reduce the total amount of data collection time required and to encourage consensus- building among respondents (Stokes and Bergin 2006). However, due to the complexity and intricacy of group dynamics, focus groups have also been found to suppress individual perspectives when social norms, culture, or power differentials encourage conformity and result in censorship (Carey 1994; Stewart and Shamdasani 2014; Stokes and Bergin 2006).

A relatively new method to the sanitation sector is photovoice, where individuals take pictures based upon a prompt and then describe their printed photographs in a follow-up interview

(Wang and Burris 1997). This process, compared to other qualitative methods, provides an individual more time to reflect and a different approach to sharing personal or community information, which may allow for greater insight into cultural aspects or sensitive issues. For example, photovoice has been used previously in other sectors, most commonly in healthcare

(Catalani and Minkler 2010); studies have found that photovoice resulted in comprehensive

12 information, even when the topic was sensitive or abstract, such as health and disease (Allotey et al. 2003; López et al. 2005). However, it is unclear if photovoice can also be a useful tool to identify priorities that are not easily visualized, such as education or safety (Kaminsky 2016; Wang and Burris 1997). In the sanitation sector, photovoice has been recently used to link household hygiene to pathogen exposure (Badowski et al. 2011) and to understand what influences community attitudes towards water, sanitation and health and how these attitudes influence behaviors (Bisung et al. 2015a; b). Overall, photovoice has been demonstrated to effectively explore social, cultural, and institutional aspects of individual and community life; despite this, its application for sanitation is still very limited, so its potential for priority identification is unknown.

Beyond identifying a comprehensive list of priorities, sanitation implementers often seek to uncover the most important priorities so that they can efficiently tailor interventions (Kamble

2014; Raymond et al. 2009). Understanding a method’s effectiveness at identifying the greatest number of unique and most important priorities should help sanitation implementers determine how to use the organization’s and the community’s limited resources effectively to better address priorities, to increase project acceptance and success, and ultimately to improve positive environmental and health impacts. Thus, the ability of a method to identify the most important priorities, versus just an unranked list of all priorities, should also be evaluated. A common way to determine the most important priorities, or the priorities of greatest need or value within a community or stakeholder group, from an unranked list of all priorities, is to use the analytical hierarchy process (AHP). The AHP is an established method used in multi-criteria decision analysis to elicit and assign numerical weights to a list of unranked priorities (Saaty 1987). While there are methods to identify priorities and methods to determine which ones are most important,

13 there is a dearth of literature that compares how well qualitative priority identification methods

(e.g., interviews) perform relative to each other.

In addition, there are multiple factors that are expected to impact the effectiveness of a priority identification method. For example, literature suggests a difference in the perspectives and experiences between community roles (Uddin et al. 2014) and males and females (Schwartz, S. and Rubel, T. 2005), and thus these groups can have different priorities. Community leaders have been found to often be inadequate spokespersons for their community, usually due to issues of power and politics (Platteau and Abraham 2002; Rao and Ibanez 2003; Sara and Katz 1998); however, evaluations of development projects show that sanitation implementers usually rely on leader opinions to a greater extent than other community roles (Platteau and Abraham 2002; Rao and Ibanez 2003; Sara and Katz 1998). Also, differences in perspectives can be magnified in certain cultural contexts; for example, in India, a significant body of research asserts that extreme gender and caste differences exist and that these differences have important implications for each group’s experience with sanitation (Brown 2010; Leder 2007; O’Reilly 2004; Schwartz, S. and

Rubel, T. 2005). Despite the recognition that respondent attributes, including their sex and role within their community, likely impact a priority identification method’s effectiveness, there has not been a systematic evaluation or quantification of this impact.

Overall, qualitative priority identification methods commonly used by sanitation implementers have not been directly or systematically compared within a sanitation context for their effectiveness at identifying the most comprehensive list of priorities and for identifying the most important priorities. Further, the impacts of respondent attributes on a method’s effectiveness are not well understood and have not been quantified. As a result, research is needed to understand how to best use sanitation implementers’ limited resources to identify priorities effectively. The

14 goals of this research were to determine which methods are the most effective for identifying the most unique sanitation priorities and community priorities and to determine if there are any differences between the methods, community roles, and respondent sex for identifying a comprehensive list of priorities and those most important to the community. These methods were used in 20 resource-limited communities, all of which had community-based sanitation systems, which were centralized treatment systems managed by communities.

Methods To accomplish these goals, this research compared the effectiveness of two commonly used methods, interviews and focus groups, and one method gaining traction in the sanitation sector, photovoice (Figure 2.1).

Community and Participant Selection Twenty resource-limited communities in Karnataka and Tamil Nadu, India were selected to compare the methods. Presently, India is home to the world’s fastest growing population and faces significant infrastructure gaps; 60 percent of the country lacks access to toilets and less than

20 percent of wastewater is treated adequately (Kamyotra and Bhardwaj 2011; Wankhade 2015).

Communities were selected based on comparable populations, demographics, and geographic location. Each community was a peri-urban slum resettlement, where communities lacked basic

15 services such as water and housing (UN-HABITAT 2010). Residents were from India’s lowest caste, had low incomes, and found employment primarily through day labor. Communities were also selected based on similarly sized sanitation systems but for a range of sanitation technologies and system performances. Each community had one centralized community-based sanitation system that served between 800 and 1000 users and was intended to provide 100 percent toilet coverage in the community. External organizations implemented these systems, and communities were responsible for operation and maintenance. Sanitation technologies included decentralized wastewater treatment systems (DEWATS), comprised of settling tanks or biogas digesters, anaerobic baffled reactors, and gravel filters; single-pass intermittent sand filters with settling tanks; community ecological sanitation systems (a type of urine diverting dry toilet); and baffled septic tanks with gravel filters (Tilley et al. 2014b). Success or failure of the sanitation system was evaluated in each community based on (1) compliance with local regulations, which was determined by measuring effluent pH, COD, and BOD; (2) how many people used the system continuously and exclusively, which was determined by triangulating researcher observations, documentation, and interviews with community members and sanitation implementers; and (3) adequate maintenance, which was determined by triangulating researcher observations, documentation, and interviews with system managers. Half of the communities had successful sanitation systems, and the other ten communities had failed sanitation systems (Table E1†). Since literature suggests a difference in the perspectives between males and females (Schwartz, S. and

Rubel, T. 2005) and between community roles (Uddin et al. 2014), male and female participation was solicited equally (Scheyvens and Storey 2003) and three different community roles

(community leader, sanitation system manager, community member) were included, when applicable, in each community (Table 2.1) to quantify and compare how these differences might

16 influence the identification of sanitation priorities and method effectiveness. In each community with a leader, the community leader was a male who was elected formally and served as the liaison between the community and outsiders, which is common in southern India. All but three communities had community leaders. The sanitation system manager was either (i) a male operator, who was a community member or a government employee that was paid to operate and maintain the sanitation system; or (ii) a women’s self-help group (WSHG), a formally organized group of 10 to 20 female community members that maintained the sanitation system and managed the income generated from that system. No community had both an operator and WSHG, and six communities had neither. Community leaders, operators, and WSHG members were identified using contact information provided by the local government or the implementing organization; they confirmed their participation in these roles and were further validated by community members. Community members were residents without a formal title or role, who also confirmed their identification with this role.

Community members were solicited by going door-to-door, which is how most sanitation implementers perform data collection (Kamble 2014). Door-to-door sampling can bias results based on who is home during data collection and may omit the perspectives of individuals who do not consent to the research. The research team mitigated these biases by returning to each community for at least six days and at all times of the day (to avoid excluding individuals who worked). The research team solicited community member respondents in the early morning (before work), at midday (during lunch breaks), and in the evening (after work). In addition, the study aimed to ensure balanced representation between males and females (due to possible different perspectives) and geographic representation (to avoid only sampling easily accessible or wealthy respondents). Theoretical saturation (i.e., no new priorities were mentioned with subsequent

17 interviews or discussion) was achieved in each community using this sampling approach.

Respondent ages, ranging from 18 to 82, were also balanced. While age may influence a respondent’s priorities (Jenkins and Curtis 2005), respondent sex and role are thought to be more influential (Freudenberger 2011), but this should be considered for future research. Eight to twelve community members participated in all three methods in all communities. Two to four WSHG members participated in all three methods in the five communities with WSHGs. Leaders and system operators were always interviewed if they existed in a community, but their availability for participation in photovoice or focus groups was limited for some of the communities due to lack of respondent time or willingness to participate (Table 2.1 and Table E2†). Non-participation of leaders and operators was not found to have a large impact on overall respondent demographics for photovoice or focus groups. Overall, nearly equal participation of males and females was achieved across all methods in all communities (Table 2.1).

Table 2.1. Percent participation by respondent sex and community roles for each priority identification method in all 20 communities.

*Mixed interviews included at least one male and one female. Mixed interviews occurred when one community member was solicited for an interview and then another community member of a different sex requested to join the same interview.

Priority Identification Methods and Data Collection Three methods were used to identify community and sanitation priorities: interviews, photovoice, and focus groups. All data collection complied with human subject research requirements as stipulated by the Institutional Review Board (IRB). The data collection protocol

18 (#16-0026) was approved by the IRB at the University of Colorado Boulder in January 2016

(renewed January 2018), with the review and agreement of academics and sanitation implementers in India experienced with research in slum resettlements. Informed consent was obtained prior to the inclusion of all respondents in this research. The first author conducted 232 interviews, 171 photovoice follow-up interviews, and 20 focus groups (for a total of 189 focus group respondents) in 20 communities, with the assistance of two translators who were fluent in the local respondents’ languages (Kannada and Tamil). Data collection from community members was continued until theoretical saturation was reached for each method. Time required and resources used for each method, in addition to observations and reflections on data collection, were recorded daily in the field.

Interviews. Ten to fifteen respondents were interviewed in each community (Table E2†).

Interviews were semi-structured and used primarily open-ended questions (Yin 2009) to elucidate both community and sanitation priorities in each interview. Sample interview questions included:

What is important to you and your community? What benefits do you find important from your sanitation system? What would you change to improve your existing sanitation system? In each community, the community leader and system operator were interviewed once each. For community members and WSHG members, additional respondents were interviewed until theoretical saturation was reached for each role. On average, this occurred after interviewing nine community members and three WSHG members.

Photovoice. Eight to ten respondents completed photovoice in each community (Table

E2†). The first author was the facilitator of the photovoice activity, with the assistance of the local translators. While most photovoice studies train respondents in a group at once (Catalani and

Minkler 2010), all respondents in this study preferred to have individual trainings. The photovoice

19 training included basic skills on digital camera use and photography and a discussion on the ethics, consent of subjects, and safety of photography in the community (Wang et al. 2004). Respondents were briefed on how and why to obtain consent from their subjects prior to photographing individuals or private property. Issues of sharing embarrassing information and depicting subjects in a false light were mitigated by having respondents discuss their photos in individual follow-up interviews (instead of in group discussions) and by sharing only the identified priorities (so that photos were anonymous and confidential) (Wang and Redwood-Jones 2001). Each participant had approximately 24 hours to take photos that responded to the prompts used to identify both community and sanitation priorities, which included: Please take pictures of problems and important priorities in your community and Please take pictures of challenges, benefits, and other important considerations for a sanitation system in your community.

The photos were then printed for use during photovoice follow-up interviews, where individual respondents were engaged in the visual analysis and discussion of the photos they took.

In addition to facilitating a discussion of photos with respondents, several photovoice studies have also performed visual analysis of the photos independent from respondents (Catalani and

Minkler 2010; Wang et al. 2004). In this research, priorities identified by photovoice came exclusively from the qualitative analysis of the photovoice follow-up interviews, which was the same method of analysis for interviews and focus groups, because this study aimed to elicit priorities solely from individual respondents. During the photovoice follow-up interviews, each respondent was encouraged to describe their photos and to reflect on what priorities they represented and why, using questions such as: How does this photograph reflect your community’s priorities for sanitation? Ten out of 17 community leaders, five of seven system operators, and two to four WSHG members per community consented to participate in photovoice. For

20 community members, additional respondents were recruited until theoretical saturation was reached, which occurred after seven participants, on average. For WSHG members, two to four respondents were included in the five communities with WSHGs.

Focus Groups. To be consistent with common focus group practice of sanitation implementers (Kamble 2014; Watkins et al. 2012), one mixed-sex, mixed-role focus group of eight to twelve community members was assembled in each community (Table E2†). Each focus group had nearly equal representation from males and females, who were mostly community members, and some focus groups had participation from a community leader and a system manager(s) (Table

E2†). Community leaders and operators did not always participate in focus groups due to availability constraints, which emulated real-world experiences of sanitation implementers.

Respondents brainstormed a list of their community priorities and a list of their sanitation priorities, in response to prompts such as: What is important to your community? and What are the most important priorities for sanitation in your community? Discussions concluded when the group reached consensus, or all focus group participants agreed that there were no additional priorities to include. Focus groups reached consensus after 1.3 hours, on average, and yielded final lists of community priorities and sanitation priorities.

Methodological Considerations. Researchers and practitioners employing priority identification methods must recognize how their own backgrounds and identities may bias the data collection processes. Careful measures were taken at every stage of data collection and analysis to mitigate biases. The authors recognized the potential biases of the first author—a female and community outsider—collecting data and sought to reduce these biases by hiring two local, male translators (one translator was from Karnataka and interpreted for the communities in Karnataka; the other translator was from Tamil Nadu and interpreted for the communities in Tamil Nadu).

21 Both had experience conducting field work in peri-urban slum resettlement communities in southern India and received extensive training on all data collection methods. All three data collection methods were piloted with both translators, after which the data collection protocol and questions were adjusted before starting the formal study. There are limitations with being an outsider; however, this role allowed the research team to mirror limitations experienced by sanitation implementers, who, in many cases, are also outsiders to the communities in which they work.

To further mitigate potential biases, such as social desirability bias (the tendency to conceal true opinions to present oneself favorably (Podsakoff et al. 2003)), confirmation bias (the tendency to seek information that supports a preconceived notion (Nickerson 1998)), and authority bias

(when the perspective of an authority figure is considered more important solely based on their position (Milgram and Gudehus 1978)), all questions and prompts were designed to be open-ended and never asked about specific issues. Steps were taken to ensure that respondents felt comfortable to voice candid opinions during data collection; for example, respondents were asked their preferred interview location and time. In addition, to help build rapport, each interview, focus group, and photovoice follow-up interview began with an in-depth introduction to the research, the first author, and the translator, and started with questions that asked the respondents to talk generally about their life and community. Finally, the first author and translators continually reflected on their influence on the data collection through written observations and discussions at the end of each fieldwork day to adjust data collection practices to make respondents more comfortable sharing.

Data Analysis Priority Identification and Ranking. Interviews, photovoice follow-up interviews, and focus group discussions were audio recorded, transcribed, and imported into QSR NVivo, a

22 qualitative coding software (QSR International 2015). Transcripts were coded for community and sanitation priorities. A coding dictionary was initially developed based on the emergent topics described by respondents as being important (Saldana 2009). For example, a respondent said,

“[We] are facing problems with antisocial elements. They come and misuse it, they cut off the wires, and sometimes they put the bottles into the toilet. The toilets get blocked, and it is very important that we stop these antisocial elements.” Many respondents in this community also used the phrase “stop antisocial elements”, so Stop Antisocial Elements at Toilets was identified as a sanitation priority. The coding dictionary was then refined by affinity grouping similar topics to develop common priority definitions for each topic (Saldana 2009). All coded data for each priority was double-checked to ensure that it only included problems that adversely affected community members and topics that were explicitly discussed as being important. Sanitation priorities related to all aspects of the sanitation service chain, including the user interface, treatment system, planning, and maintenance.

Finally, all transcripts were coded by two independent coders and Cohen’s Kappa coefficient, which measures the level of agreement between coders, was calculated (Bazeley and

Jackson 2013). The Kappa coefficient was 0.64, which reflects acceptable agreement between coders (Bazeley and Jackson 2013). The final two lists of community and sanitation priorities, for each community, were internally validated using member-checking, where respondents were asked to confirm that priorities identified were representative of their priorities (Birt et al. 2016).

These unranked lists of unique community and sanitation priorities were then evaluated using the AHP (see Section 2†) (Saaty 2008) to identify the most important priorities (i.e., the five priorities with the highest ranking, for both community and sanitation priorities; most multi-criteria decision-making methodologies compare the tradeoffs between five or fewer criteria (Mendoza

23 and Martins 2006; Wang et al. 2009)). In each community, an AHP group, which was a second focus group with eight to twelve community members (Table E3†), was asked to perform pairwise comparisons for both lists of community and sanitation priorities to provide the data needed to rank the priorities in each list. To compare how well each method identified different types of priorities, the defined priorities were classified as either physical or abstract. Physical priorities were visible or addressable with infrastructure (e.g., biogas could be addressed by implementing a digester). In contrast, abstract priorities were not visual and usually required a social or organizational intervention (e.g., women’s empowerment could be addressed by using sanitation system income for micro-loans to women).

Methods Comparison. Interviews, photovoice, and focus group results were analyzed to determine method effectiveness for identifying the greatest number of unique priorities. Since there were multiple respondents for some community roles (i.e., multiple community and WSHG members), the effectiveness of a given role and method was evaluated two ways: (1) based on the total number of unique priorities identified by all respondents in that role, and (2) based on the average number of unique priorities identified by one respondent in that role. By comparing the total and average number of unique priorities identified, the effectiveness of speaking to one individual, versus speaking to multiple individuals, was evaluated. Methods were compared using one-way analyses of variance (ANOVA) at a 95% confidence level (α=0.05). In all comparisons, the null hypothesis was that the population means, from which each sample was drawn, were equal.

Methods were also compared based on ease of implementation, time required, resources used, and number of respondents needed.

24 Results and Discussion Most Effective Method Overall, conducting interviews was the most effective method and facilitating a focus group was the least effective method (Figure 2.2). For each community, nine to fifteen unique priorities were identified for both community and sanitation priorities (Table E4†). Interviews identified more sanitation priorities than photovoice (p=0.015, Table E5(a)†). Specifically, interviews identified the most sanitation priorities in 12 out of 20 communities (Figure 2.2a). In those communities, interviews may have been more effective due to their open-ended nature in contrast to the task-oriented nature of photovoice. Interviews were equally as effective as photovoice in three communities. Photovoice identified one more sanitation priority than interviews in four communities; in Communities 6, 10, and 16 case information does not indicate a reason for this difference, but in Community 18, case knowledge indicates that the difference is likely due to two active members of the WSHG participating in photovoice who were more knowledgeable than other community roles about sanitation priorities but had similar knowledge regarding community priorities. Photovoice identified two more priorities than interviews in

Community 1, which is likely because photovoice was conducted on the day of a festival, so community members had more time at home to photograph priorities.

Figure 2.2. Results of the percent difference between methods for identifying the greatest number of (a) unique sanitation priorities and (b) unique community priorities. Comparison of interviews (10 to 15 respondents per community), photovoice (8 to 10 respondents per community), and focus groups (one focus group with 8 to 12 respondents per community) for identifying sanitation (a) and community (b) priorities. Results are normalized to the photovoice results in each community to show the percent difference between photovoice and interviews (blue) and between photovoice and focus groups (yellow). Absent bars indicate that the method performed equally to photovoice. *Results are arranged by trend, not by increasing community number.

Interviews identified the greatest number of unique community priorities in ten communities and identified the same number as photovoice in nine communities (Figure 2.2b). In one community, Community 20, photovoice elicited one additional unique community priority over interviews. This higher photovoice performance is likely due to the participation of two

WSHG members, who were highly active in the community and excited to participate in photovoice. These two photovoice respondents photographed about two more community

26 priorities than the other respondents and uniquely identified Education as a priority. This shows the potential effectiveness of photovoice, but this result was not representative; respondents in other communities with similarly active WSHGs (Communities 11 and 18) identified the same number of community priorities with photovoice as with interviews, and interviews were overall more effective. On average, there was not a large difference between the effectiveness of interviews and photovoice for identifying community priorities (p=0.095, Table E5(a)†).

Focus groups identified fewer community priorities and sanitation priorities than interviews and photovoice (p-values<0.015, Table E5(b-c)†). There was only one instance for each priority category where a focus group’s performance was as effective as another method. In

Community 4, the focus group and interviews identified the same number of sanitation priorities, which was more than those identified with photovoice (Figure 2.2a). This is likely because that focus group’s discussion lasted 2.5 hours, which was more than an hour longer than the average focus group time of 1.3 hours (Table 2.2). Also, that focus group spent more time discussing sanitation priorities than community priorities, likely because the focus group was conducted within sight of their failed sanitation system while most other focus groups were conducted in a community hall. In Community 5, the focus group and photovoice respondents identified a similar number of community priorities, but both identified one fewer community priority than interviews

(Figure 2.2b). On average, compared to interviews, a focus group identified 3.5 fewer community priorities and 3.5 fewer sanitation priorities.

Physical or Abstract Priorities. Physical priorities ranged from drainage to toilet cleanliness to water reuse, and abstract priorities ranged from education to no smells to privacy.

For physical priorities, interviews and photovoice performed similarly (p-values>0.094, Table

E6†), demonstrating that both interviews and photovoice were effective at identifying physical

27 priorities. On average, using ± standard deviation to show variation, interviews identified 96%

±6% of community and 92% ±12% of sanitation physical priorities; photovoice identified 95%

±7% of community and 82% ±17% of sanitation physical priorities. Focus groups identified fewer physical priorities than interviews or photovoice (p-values<0.033, Table E6†), but, based upon the results above, this is likely a reflection of the relative ineffectiveness of focus groups rather than a difference based on priority type.

For abstract priorities, interviews identified more unique community and sanitation priorities than photovoice or focus groups (Table E7†). Interviews identified 96% ±9% of community and 99% ±4% of sanitation abstract priorities, while photovoice identified only 81%

±21% of community and 80% ±18% of sanitation abstract priorities. Interviews identified more sanitation priorities than photovoice, likely because sanitation priorities consisted of more abstract priorities (61% on average) than physical priorities and because photovoice did not identify abstract priorities as well (photovoice identified 80% of abstract priorities while interviews identified 99%). This difference is less pronounced when looking at community priorities, likely because more community priorities were physical (72% on average). Photovoice was expected to identify fewer abstract priorities since abstract concepts can be more difficult to photograph than physical concepts and since it requires that respondents identify an artifact or scene that symbolizes a non-visual concept.(Wang and Burris 1997) For example, photovoice only identified 60% of

Community 4’s abstract priorities while interviews captured 100%; priorities not photographed included Safety and Dignity, Low Operation and Maintenance Demands, and Low Cost. Despite efforts taken by the researchers to encourage photovoice respondents to include photographs of artifacts that symbolize abstract concepts (e.g., a picture of a child with books could represent education), photovoice failed to identify as many abstract priorities as interviews. The worst

28 interview performance was in Community 3, where interviews only captured 80% the community’s abstract priorities, but photovoice also only captured 80%.

Most Important Priorities. The greatest variation between methods was observed within the top five ranked priorities (Figure E1†). All priorities that were not ranked in the top five had an AHP relative importance of less than 0.10, and there was not a substantial variation between the three methods for identifying those priorities. Therefore, the top five community and top five sanitation priorities were considered as the “most important” priorities. Overall, interviews were more effective than photovoice (p-values<0.037, Table E7(a)†) and focus groups (p-values<0.018,

Table E7(b)†) for identifying both the most important community and most important sanitation priorities. The next most effective method was photovoice. In addition, in all communities, interviews and photovoice identified the most important priority, ranked first according to the AHP results (Figure E1†). For community priorities, though, photovoice was more similar to the least effective method, focus groups, and identified only slightly more most important community priorities than focus groups, on average (p=0.057, Table E7(c)†).

Discussion of Methods’ Effectiveness Overall, interviews were the most effective method for identifying (i) a comprehensive list of unranked priorities and (ii) the most important priorities, for both community and sanitation priorities. These results support the continued use of interviewing, which is one of the most commonly used qualitative priority identification methods.(Kamble 2014; Raymond et al. 2009)

However, interviewing was the most time-intensive method; with ten to fifteen respondents for each method, interviews took more than twice as much time as photovoice and more than eight times as much time as a focus group (Table 2.2 and Figure E2†). This longer time may be a main reason for the effectiveness of interviews. Similarly, focus groups required the least amount of

29 time in all communities and never identified more priorities than the other two methods.

Photovoice required less time than interviews and more time than focus groups, but, relative to interviews, identified the same or more sanitation priorities in seven communities and the same or more community priorities in ten communities.

Table 2.2. Comparison of the average time and the resources needed for interviews, photovoice, and focus groups in all 20 communities. Estimates are for 8 to 15 respondents and do not include recruitment of respondents and introductions to the research, which was required for all methods. The 24-hour period that respondents had for photovoice is not included in the data collection time estimates. Figure E2† shows, for each method, the number of priorities identified and amount of time required by community. Average Data Collection Time Method Resources Required (and Range) in Hours Interviews 10.6 (8.5 to 12.8) Tape recorder, audio transcriptions, qualitative coding software 8-10 digital cameras, photo printing capabilities, tape recorder, Photovoice 4.5 (4.0 to 6.0) audio transcriptions, qualitative coding software Focus Groups 1.3 (0.8 to 2.5) Tape recorder, audio transcriptions, qualitative coding software

Photovoice performed nearly as well as interviews while requiring less data collection time.

In many instances, photovoice performed comparably or slightly less effectively than interviews, but photovoice still identified reasonably complete lists of priorities (e.g., on average, photovoice identified 94% ±10% of the community priorities identified by interviews and 89% ±15% of the sanitation priorities). Main considerations for using photovoice include equipment and time requirements outside of follow-up interviews. The time spent during an average photovoice follow-up interview was about 0.4 ±0.06 hours shorter than an average semi-structured interview.

Photovoice interviews were expected to be shorter because respondents had additional time to reflect and document their thoughts (at least 24 hours), but this did not result in more unique priorities than interviews in most communities. However, the total amount of time to complete the entire photovoice method is longer for both respondents and the interviewer. The total time for an interviewer includes the time to train respondents, pass out cameras, wait for respondents to take

30 the pictures, collect cameras, print the pictures, and conduct the follow-up interviews. The total time for a participant must also include the time spent taking pictures. Despite the longer time commitment, many respondents voiced a willingness to complete the photovoice exercise because of an excitement to use a digital camera, but this is expected to vary according to context. Another main consideration is photovoice equipment because photovoice requires cameras and the ability to print pictures.

Photovoice and interviews could be employed to complement each other and triangulate results, when possible, to collect more comprehensive data. For instance, theoretical saturation was reached with photovoice with fewer respondents (8 to 10) than with interviews (10 to 15), and photovoice could reduce time spent interviewing. In addition, photovoice may be a useful tool to document the extent to which a priority affects a community, whereas interviews may provide more comprehensive lists of the priorities overall. For example, many photovoice respondents chose to photograph every instance of a given problem in their community; one community member photographed over ninety clogged drains to document every location where drainage was an issue. Finally, photovoice may be a useful method to increase stakeholder engagement and build rapport in communities. Respondents who participated in photovoice were more willing to participate in subsequent interviews for a future study; these findings are consistent with other studies that used photovoice (Bisung et al. 2015b; Kaminsky 2016).

While focus groups are used commonly to reduce data collection time from the same number of respondents (Stewart and Shamdasani 2014), focus groups failed to identify comprehensive lists of community and sanitation priorities. Despite current practice of sanitation implementers to tend to use focus groups to save time,(Kamble 2014) focus groups may not be effective enough to justify the reduced time. The authors followed common focus group practice

31 of conducting mixed-sex, mixed-role focus groups (Kamble 2014), and found that some individual perspectives were suppressed during focus groups, rendering focus groups ineffective at identifying the most priorities. For example, Community 2 focus group respondents said they did not voice Move Household Toilets Outside as a priority because the leader was present during the focus group discussions, but other community members voiced this priority when in private (i.e., during interviews and photovoice with the female first author and male translator). In Community

5, female respondents identified Sanitary Napkin Disposal as a priority in interviews and photovoice with the female first author and male translator; however, the focus group respondents did not identify that priority. It is expected that the suppression of perspectives during focus groups was due to the public nature of the focus group and presence of other focus group participants, which validates previously found focus group limitations (Carey 1994). Conducting multiple focus groups in a single community and separating respondents based on sex, role, or social network could improve the effectiveness of focus groups (Carey 1994; Krueger and Casey 2014), but research is still needed to systematically evaluate this approach in the context of sanitation. Finally, unless multiple focus groups are conducted within a community, theoretical saturation cannot be reached in this method in the same way as interviews and photovoice (i.e., adding additional respondents until no new priorities are mentioned), which remains another limitation of the focus group method.

Impact of Community Role and Participant Sex on Method Effectiveness Respondent’s Community Role Community Members. Community members were the most effective at identifying the greatest number of unique priorities (Table 2.3(c-e)) and the greatest number of most important priorities (Table E7(c-e)†), for both community and sanitation priorities; this was true for both interviews and photovoice. The one exception was that WSHG members identified more of the

32 most important sanitation priorities than community members, when using photovoice (p=0.000,

Table E7(3)†). In all 20 communities, the total response from all (8 to 12) participating community members identified on average 35% ±8% more unique priorities than the average community member response, for both community priorities (Figure E4(a)†) and sanitation priorities (Figure

E4(b)†). However, a single community member still provided more information than any other role (Table 2.3(c-e) and Figure E3†). This was even true in the six communities that had all three roles; a single community member, on average, identified more priorities than both the leader and system manager (Figure 2.3). The only exception was that the average response from a single

WSHG member was better than the average community member when using photovoice to identify sanitation priorities (p=0.026, Table 2.3(e)).

Figure 2.3. Results of the number of unique community priorities identified in interviews by each community role (community leader, community members, system manager). Communities presented are the six communities (out of 20) that had three roles (leader, members, and manager; where a community had either an operator or WSHG). Results are normalized to community leaders and reflect the percent difference between the number of unique community priorities identified by the community leader and members of each comparative role. Total Responses for Community Members and WSHG Members reflect results from all respondents in the role until theoretical saturation was reached (n=7 to 12 and n=3 to 5, respectively). Complete results of the

33 comparison between roles for all communities are presented in Figure E3(a)†. Results for sanitation priorities are presented in Figures E3(b) and E5†. CL = Community leader; CM = Community member; OP = Operator; WSHG = Women’s Self-Help Group. *Community number not in numerical order.

Table 2.3. Results of a comparison of the most effective role from all 20 communities based on identification of the greatest number of unique priorities. Results are for comparisons between (a) leader and operator, (b) leader and WSHG member, (c) community member and leader, (d) community member and operator, (e) community member and WSHG member, (f) WSHG member and operator. For Community Member (Member) and WSHG Member (WSHG), the results are for the average response for one individual. The p-values from one-way ANOVA are in parentheses. Bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). Respondent Role (a) (b) (c) (d) (e) (f) Priority Method Leader Leader Member Member Member WSHG Category vs. vs. vs. vs. vs. vs. Operator WSHG Leader Operator WSHG Operator

Community Leader WSHG Member Member Member WSHG Priorities (0.914)* (0.042) (0.000) (0.012) (0.003) (0.030)

Interviews

Sanitation Leader WSHG Member Member Member WSHG Priorities (0.932)* (0.009) (0.001) (0.004) (0.000) (0.006)

Community Leader WSHG Member Member Member WSHG Priorities (0.378)* (0.141)* (0.044) (0.006) (0.005) (0.090)*

Photovoice

Sanitation Leader WSHG Member Member WSHG WSHG Priorities (0.738)* (0.036) (0.036) (0.029) (0.026) (0.010)

Community Leaders. While community leaders are most often solicited for sanitation priority identification (Kapiriri and Norheim 2002), they identified fewer community priorities and sanitation priorities than the average community member (p-values<0.044, Table 2.3(c)). Only 17 of the 20 communities had a leader, and the average community member identified more priorities in 16 of these 17 communities (Figure E3†). The one exception was Community 3 where the leader identified two additional community priorities (25% more) and one more sanitation priority (10%

34 more) when compared to the average response of a single community member; however, the total response of 11 community members identified more priorities than the one community leader

(Figure E3†). In all communities, the leaders were elected annually (usually based on their relationships to previous leaders (e.g., sons of past leaders) rather than their community service or skillset), and they most often organized festivals and managed intra-community conflict but rarely initiated or were involved with community projects. Community leaders were also males who mostly worked outside the community and spent minimal time inside the community. While the role of a community leader may not provide greater insight than the average community member, interacting with community leaders is often necessary for obtaining permission for data collection in a community and encouraging other community roles to participate.

Sanitation System Manager. A difference between the types of sanitation system manager was observed; although no community had both an operator and a WSHG for direct comparison, operators identified fewer priorities than WSHG members, on average (Table 2.3).

Male operators were the least effective of all community roles at identifying priorities, in both interviews and photovoice (Table 2.3(a,d,f)). Operators identified significantly fewer unique and fewer most important community and sanitation priorities than the average community member

(p-values<0.029, Table 2.3(d)). Operators also identified fewer community priorities and fewer of the most important community and sanitation priorities than community leaders, but the difference was not significant (p-values>0.378, Table 2.3(a)). Only three of seven operators received training and only four of seven operators received a monthly salary, which required them to hold at least one additional job. In five of these communities, perhaps due to lack of salary incentives, the operators did not perform adequate maintenance for the sanitation system. Additionally, operators were only present at the sanitation system when maintenance was needed) and had only been in

35 their role for a short time, on average three years; these relatively minimal interactions with the sanitation system could further explain their limited ability to identify priorities. The poor performance of the operator compared with the other community roles suggests that, like community leaders, operators may not be sufficient spokespersons.

In contrast, managing the sanitation system was the primary focus for the WSHG members who spent most of the day at the system, performing maintenance and socializing and had maintained the system for longer, on average seven years. The average WSHG member identified more community and sanitation priorities in interviews and more sanitation priorities in photovoice than operators and leaders (p-values<0.042, Table 2.3(b,f)), but fewer community and sanitation priorities in interviews and community priorities in photovoice than the average community member (p-values<0.005, Table 2.3(f)).

In three of the five communities with WSHGs (Communities 11, 18, and 20), the average

WSHG member response identified more sanitation priorities than the combined community member responses (Figure E3(b)†). Case knowledge suggests that WSHG members, who have a sense of ownership for a project or responsibility for the community, tend to identify more priorities in both interviews and photovoice. In Communities 11, 18, and 20, WSHGs played an active role in community life by providing micro-loans to women and families and employing

WSHG members as caretakers for sanitation facilities. Communities 6 and 9 previously had similar

WSHGs, but after the failure of the sanitation system, the WSHGs declined significantly in membership, activity, interest, and motivation, which may explain why these WSHGs identified fewer priorities. Community 20 also had a failed sanitation system but an active WSHG so the effectiveness of WSHG members to identify priorities may not only be based on the current state of the sanitation system. While members of active WSHGs identified more priorities than

36 community members, the results are limited by the small sample size (only 5 communities with

WSHGs). When active WSHGs are present, WSHG members could also be a useful source of information and should be included when possible. Future work should seek to validate the effectiveness of active WSHGs for priority identification, as compared with inactive WSHGs and other community roles.

Community Member Sex In general, female community members identified slightly more unique priorities than male community members (p-values from 0.000 to 0.387, Table E9†). For interviews, both the average female and combined female responses identified more community priorities and more of the most important community priorities than the average male and combined male responses (p- values=0.000). The average female also identified more unique sanitation priorities and more of the most important sanitation priorities in interviews than the average male (p-values<0.039). For photovoice, the average female identified more unique community priorities than the average male

(p=0.049), and the combined female responses identified more sanitation priorities than all combined male responses (p=0.015).

Hofstede’s cultural dimensions present some insight into differences based on respondent sex: India has a high score for power distance, which is the idea that individuals (particularly castes and males/females) are unequal (Hofstede 1998, 2012). Previous studies have recognized differences in the beliefs and perspectives of males and females (Ryan, M et al. 2001; Schwartz,

S. and Rubel, T. 2005), and most priority assessment tools recommend including both sexes

(Watkins et al. 2012). Additionally, research suggests that males and females have different experiences with sanitation systems (O’Reilly 2004) and thus might identify different priorities, even though they might identify relatively similar numbers of priorities, as found in this study.

Another potential reason for a difference in male and female responses is that in most of the

37 selected cases, males were the primary wage-earners and sought work outside of their community.

Females were the primary caregivers and homemakers and spent significantly more time in their community, interacting with neighbors and having more interactions with the full sanitation service chain, such as cleaning a household’s toilet or passing by the treatment system to collect water, than males; this could partially explain why females identified more sanitation priorities than males. Interviews with female community members lasted an average of 10 minutes longer and identified an average of 2.5 more priorities compared to interviews with male community members.

Most Effective Method for Each Role and Sex A comparison between interviews and photovoice was performed to determine if either method was more effective when used with a specific community role or respondent sex. For community members, interviews and photovoice were similarly effective for identifying community priorities (all p-values>0.062) and interviews were more effective than photovoice for identifying sanitation priorities (p-values<0.006) (Table E10†). Interviews were more effective than photovoice at identifying the most important priorities when evaluated for the combined community members response (p-values<0.026), but no difference was found between the methods for the average community member (p-values>0.150). For community leaders, operators, and WSHG members, there were not large differences between interviews and photovoice (p- values>0.236). In general, within each sex, results indicate that photovoice and interviews provide similar results for identifying community and sanitation priorities.

Conclusion Findings show that interviews are the most effective method for identifying priorities, relative to the other three methods, when used in isolation. Interviews consistently identified (i) the most unique community and sanitation priorities, (ii) the greatest number of most important

38 community and sanitation priorities, and (iii) the most abstract priorities. While interviews are time-intensive in data collection and analysis, the comprehensive lists of priorities identified by interviews motivate the continued use of interviews as a priority identification tool in the sanitation field. Photovoice performed slightly less effectively compared to interviews but was less time consuming during data collection and reached theoretical saturation quicker. Photovoice could be used in conjunction with interviews to reduce overall interview time and as a tool to encourage community participation in project planning and implementation. In contrast, focus groups were found to be the least effective method for identifying both community and sanitation priorities and should not be used in isolation for identifying sanitation priorities, as the method is limited by the potential for censorship of perspectives due to group dynamics. To achieve the most comprehensive lists of priorities, sanitation implementers should use multiple methods and triangulate the results.

Findings also demonstrate that the most priorities are identified by including multiple community members, especially females. In contrast with common practices of sanitation implementers, community leaders were found to be relatively ineffective at voicing comprehensive lists of community and sanitation priorities, so they should not be the sole target of a priority assessment. Case knowledge suggests that WSHG members, particularly of active WSHGs, may provide unique insight for priorities and should be included in priority assessments because they interact with sanitation systems the most within a community and more than male operators. To obtain the most comprehensive information, as many community roles and sexes as possible should be included in data collection.

Overall, interviews were the most effective method, relative to photovoice and focus groups, for priority identification due to their effectiveness for identifying priorities across all

39 roles, sexes, priority categories, and priority types. Priority identification is, however, even more effective when multiple methods and multiple respondents with different perspectives and experiences can be used and results triangulated. Practically, results are beneficial for environmental engineers, non-governmental organizations, government agencies, sanitation implementers, and development professionals. This information can help to streamline method selection and possibly justify expanding data collection, provide decision-making support, and manage limited data collection time and resources efficiently. Ultimately, knowledge of how to better identify priorities could improve how well projects address priorities, increase system use and acceptance, and expand positive public and environmental health benefits.

Conflicts of Interest

There are no conflicts to declare.

Acknowledgements

This work was completed with the financial support of the Mortenson Center in Engineering with

Developing Communities at the University of Colorado Boulder. We would like to recognize the

4S team at Eawag-Sandec, CDD Society, and BORDA for their assistance with this research. We would also like to thank our research assistants, Vijay Kumar and Sridhar Selvaraj. Finally, we thank all of the communities and respondents in this research for their time and support.

Chapter 2 References Alkire, S. (2005). Valuing Freedoms: Sen’s Capability Approach and Poverty Reduction. Oxford University Press, Oxford. Allotey, P., Reidpath, D., Kouamé, A., and Cummins, R. (2003). “The DALY, context and the determinants of the severity of disease: an exploratory comparison of paraplegia in Australia and Cameroon.” Social Science & Medicine, 57(5), 949–958. Alsaawi, A. (2014). “A critical review of qualitative interviews.” European Journal of Business and Social Sciences, 3(4), 149–156. Badowski, N., Castro, C. M., Montgomery, M., Pickering, A. J., Mamuya, S., and Davis, J. (2011). “Understanding Household Behavioral Risk Factors for Diarrheal Disease in Dar es

40 Salaam: A Photovoice Community Assessment.” Journal of Environmental and Public Health, 2011, 1–10. Bazeley, P., and Jackson, K. (2013). Qualitative Data Analysis with NVivo. Sage Publications. Birt, L., Scott, S., Cavers, D., Campbell, C., and Walter, F. (2016). “Member checking: A tool to enhance trustworthiness or merely a nod to validation?” Qualitative Health Research, 26(13), 1802–1811. Bisung, E., Elliot, S. J., Abudho, B., Karanja, D. M., and Schuster-Wallace, C. J. (2015a). “Using Photovoice as a Community Based Participatory Research Tool for Changing Water, Sanitation, and Hygiene Behaviors in Usoma, Kenya.” BioMed Research International, 8(2), 1–10. Bisung, E., Elliott, S. J., Abudho, B., Schuster-Wallace, C. J., and Karanja, D. M. (2015b). “Dreaming of toilets: Using photovoice to explore knowledge, attitudes and practices around water–health linkages in rural Kenya.” Health & Place, 31, 208–215. Breslin, E. D. (2003). Demand response approach in practice: why sustainability remains elusive. WaterAid, Mozambique. Brown, A. (2010). “Empowerment and Gender Equality for Water and Sanitation in Rural India: Two Case Studies.” Master of Science, Oregon State University, Corvallis, Oregon. Carey, J. M. M. A. (1994). “The Group Effect in Focus Groups: Planning, Implementing, and Interpreting Focus Group Research.” Critical Issues in Qualitative Research Methods, Morse, Janice M., ed., SAGE, Thousand Oaks, California. Catalani, C., and Minkler, M. (2010). “Photovoice: A Review of the Literature in Health and Public Health.” Health Education & Behavior, 37(3), 424–451. Freudenberger, K. S. (2011). Rapid Rural Appraisal and Participatory Rural Appraisal: A Manual for CRS field Workers and Partners. Manual, Catholic Relief Services, Baltimore, Maryland. Harrell, M. C., and Bradley, L. M. A. (2009). Data Collection Methods: Semi-Structured Interviews and Focus Groups. Rand Corporation for the National Defense Research Institute, Arlington, VA. Hofstede, G. (1998). “Attitudes, values and organizational culture: Disentangling the concepts.” Organization studies, 19(3), 477–493. Hofstede, G. (2012). “India.” Cultural Dimensions, (Aug. 7, 2017). Jenkins, M. W., and Curtis, V. (2005). “Achieving the ‘good life’: Why some people want latrines in rural Benin.” Social Science & Medicine, 61(11), 2446–2459. Kamble, S. M. (2014). “Participatory rural appraisal: a tool for inclusive growth and participatory development a case study of Village Marale, MS, India.” Int Res J of Soc Sci, 3(3), 48–50. Kaminsky, J. (2016). “Institutionalizing infrastructure: photo-elicitation of cultural-cognitive knowledge of development.” Construction Management and Economics, 33(11–12), 1–15. Kamyotra, J. S., and Bhardwaj, R. M. (2011). “Municipal wastewater management in India.” India Infrastructure Report, 299. Kapiriri, L., and Norheim, O. F. (2002). “Whose priorities count? Comparison of community- identified health problems and Burden-of-Disease-assessed health priorities in a district in Uganda.” Health Expectations, 5(1), 55–62. Krueger, R. A., and Casey, M. A. (2014). Focus Groups: A Practical Guide for Applied Research. SAGE Publications.

41 Leder, S. (2007). “Gender issues in water and sanitation programmes: Lessons from India.” South Asian Water Studies, (A. A. Cronin, P. K. Mehta, and A. Prakash, eds.), 5(4), 59–62. López, E. D. S., Eng, E., Randall-David, E., and Robinson, N. (2005). “Quality-of-Life Concerns of African American Breast Cancer Survivors Within Rural North Carolina: Blending the Techniques of Photovoice and Grounded Theory.” Qualitative Health Research, 15(1), 99– 115. McNall, M., and Foster-Fishman, P. G. (2007). “Methods of Rapid Evaluation, Assessment, and Appraisal.” American Journal of Evaluation, 28(2), 151–168. Mendoza, G. A., and Martins, H. (2006). “Multi-criteria decision analysis in natural resource management: A critical review of methods and new modelling paradigms.” Forest Ecology and Management, 230(1–3), 1–22. Milgram, S., and Gudehus, C. (1978). Obedience to authority. Ziff-Davis Publishing Company. Murphy, H. M., McBean, E. A., and Farahbakhsh, K. (2009). “Appropriate technology – A comprehensive approach for water and sanitation in the developing world.” Technology in Society, 31(2), 158–167. Nickerson, R. S. (1998). “Confirmation bias: A ubiquitous phenomenon in many guises.” Review of general psychology, 2(2), 175. O’Reilly, K. (2004). “Developing Contradictions: Women’s Participation as a Site of Struggle Within an Indian NGO∗.” The Professional Geographer, 56(2), 174–184. Platteau, J.-P., and Abraham, A. (2002). “Participatory Development in the Presence of Endogenous Community Imperfections.” Journal of Development Studies, 39(2), 104–136. Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., and Podsakoff, N. P. (2003). “Common method biases in behavioral research: A critical review of the literature and recommended remedies.” Journal of Applied Psychology, 88(5), 879–903. QSR International. (2015). NVivo qualitative data analysis software. QSR International Pty Ltd. Rao, V., and Ibanez, A. M. (2003). The social impact of social funds in Jamaica: A mixed-methods analysis of participation, targeting, and collective action in community-driven development. World Bank Publications. Raymond, C. M., Bryan, B. A., MacDonald, D. H., Cast, A., Strathearn, S., Grandgirard, A., and Kalivas, T. (2009). “Mapping community values for natural capital and ecosystem services.” Ecological Economics, 68(5), 1301–1315. Ryan, M, Da, Scott, Reeves, C., Bate, A., van Teijlingen, Er, Russell, E., Napper, M., and Robb, C. (2001). “Eliciting public preferences for healthcare: a systematic review of techniques.” Health technology assessment (Winchester, England), 5(5), 1–186. Saaty, R. W. (1987). “The analytic heirarchy process - what it is and how it is used.” Mathematical Modeling, 9(3–5), 161–176. Saaty, T. L. (2008). “Decision making with the analytic hierarchy process.” International Journal of Services Sciences, 1(1), 83–98. Saldana, J. (2009). The Coding Manual for Qualitative Researchers. Sage, London. Sara, J., and Katz, T. (1998). Making Rural Water Supply Sustainable: Report on the Impact of Project Rules. Water and Sanitation Program, World Bank, Washington, D.C. Scheyvens, R., and Storey, D. (Eds.). (2003). Development Fieldwork: A Practical Guide. Sage Publications, London. Schwartz, S., and Rubel, T. (2005). “Sex differences in value priorities: Cross-cultural and multimethod studies.” Journal of Personality and Social Psychology, 89(6), 1010–1028.

42 Starkl, M., Brunner, N., and Stenström, T.-A. (2013). “Why Do Water and Sanitation Systems for the Poor Still Fail? Policy Analysis in Economically Advanced Developing Countries.” Environmental Science & Technology, 47, 6102–6110. Stewart, D. W., and Shamdasani, P. N. (2014). Focus Groups: Theory and Practice. SAGE Publications, Thousand Oaks, California. Stokes, D., and Bergin, R. (2006). “Methodology or ‘methodolatry’? An evaluation of focus groups and depth interviews.” Qualitative Market Research: An International Journal, 9(1), 26–37. Tilley, E., Ulrich, L., Lüthi, C., Reymond, P., and Zurbrügg, C. (2014). Compendium of Sanitation Systems and Technologies. Eawag. Uddin, S. M. N., Muhandiki, V. S., Sakai, A., Al Mamun, A., and Hridi, S. M. (2014). “Socio- cultural acceptance of appropriate technology: Identifying and prioritizing barriers for widespread use of the urine diversion toilets in rural Muslim communities of Bangladesh.” Technology in Society, 38, 32–39. UN-HABITAT. (2010). Chapter 1: Development Context and the Millennium Agenda. The Challenge of Slums: Global Report on Human Settlements 2003, United Nations, Geneva, Switzerland. United Nations. (2015). Sustainable Development Goals. United Nations. Wang, C., and Burris, M. A. (1997). “Photovoice: Concept, Methodology, and Use for Participatory Needs Assessment.” Health, Education & Behavior, 24(3), 369–387. Wang, C. C., Morrel-Samuels, S., Hutchison, P. M., Bell, L., and Pestronk, R. M. (2004). “Flint photovoice: Community building among youths, adults, and policymakers.” American journal of public health, 94(6), 911–913. Wang, C. C., and Redwood-Jones, Y. A. (2001). “Photovoice ethics: Perspectives from Flint photovoice.” Health education & behavior, 28(5), 560–572. Wang, J.-J., Jing, Y.-Y., Zhang, C.-F., and Zhao, J.-H. (2009). “Review on multi-criteria decision analysis aid in sustainable energy decision-making.” Renewable and Sustainable Energy Reviews, 13(9), 2263–2278. Wankhade, K. (2015). “Urban sanitation in India: key shifts in the national policy frame.” Environment and Urbanization, 27(2), 555–572. Watkins, R., West Meiers, M., and Visser, Y. (2012). A Guide to Assessing Needs: Essential Tools for Collecting Information, Making Decisions, and Achieving Development Results. The World Bank, Washington, D.C. Wood, B. T., Dougill, A. J., Quinn, C. H., and Stringer, L. C. (2016). “Exploring Power and Procedural Justice Within Climate Compatible Development Project Design: Whose Priorities Are Being Considered?” The Journal of Environment & Development, 25(4), 363–395. Yin, R. K. (2009). Case study research: design and methods. Applied social research methods, Sage Publications, Los Angeles, Calif.

43 Chapter 3: Priority Addressment Protocol: Understanding the ability and

potential of sanitation systems to address priorities

This chapter is the published journal article: Davis, A.; Javernick-Will, A.; Cook, S. Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities. Environmental Science & Technology 2019, 53 (1), 401–411. https://doi.org/10.1021/acs.est.8b04761.

Abstract Art

Keywords: Resource-limited communities, priority assessment, sanitation failure, resource recovery

44 Abstract Sanitation acceptance is unlikely if user priorities are not addressed. However, sanitation systems are commonly implemented, especially in resource-limited communities, without incorporating local context. Understanding sanitation systems’ abilities to address different priorities will further inform technology selection processes. Therefore, a protocol was created to identify priorities and measure how well sanitation systems address them, based upon their importance to a community. This protocol was applied to 20 community-based sanitation systems in India. Overall, 52 sanitation and 40 community priorities were identified; most, along with their relative importance, were case-specific and not yet identified in literature. Existing sanitation systems poorly addressed priorities. Nonfunctional systems addressed the fewest, but, if use and maintenance were improved, they had the potential to address priorities almost as well as functional systems. Resource recovery systems addressed the most priorities, but there was usually minimal benefit to adding all three resources to an existing system; biogas and water had greater potential to address more priorities than compost. This priority addressment protocol can help identify the most appropriate technologies and strategies to improve technology development and success.

45 Introduction The United Nations’ Sustainable Development Agenda encourages an increased focus on sanitation (United Nations 2015b); however, sanitation still receives little attention from many governments and foreign assistance programs, and 60% of the global population lacks access to functional or adequate sanitation (WHO and UNICEF 2017). Access is limited by many factors, especially high failure rates; 70% of sanitation systems fail within two years (Davis 2015). This issue is most common in resource-limited communities where sanitation use and acceptance is less likely to occur if user priorities are not addressed (Freudenberger 2011; WHO 2004).

Therefore, identifying priorities can improve success (Harris-Lovett et al. 2015; Roma and

Jeffrey 2010; Starkl et al. 2013b) and provide a strategy to increase universal sanitation access.

However, sanitation systems are commonly implemented in resource-limited communities without incorporating local context, usually because implementers lack the resources or expertise to effectively assess priorities (Murphy et al. 2009). When priorities can be assessed, demand- responsive assessments are recommended but are resource-intensive (Chambers 1981). So, implementers tend to use supply-driven (Breslin 2003) or limited (Platteau and Abraham 2002) assessments, which commonly result in insufficient data. In addition, it is recognized that culture influences priorities (Murphy et al. 2009), but it is unclear to what extent communities with similar cultures share values. Also, most assessments focus on community-level needs (Lloyd-Jones and

Rakodi 2014), but it is unknown if priority assessments should be context-specific or if overall community priorities can be translated to different projects, such as sanitation or energy. Finally, the relative importance of identified priorities is usually not assessed (Sperling et al. 2016), but because not all priorities may be addressable, especially when resources are limited, processes to identify a community’s most important priorities can focus implementation efforts. Overall, there

46 is a need to evaluate the usefulness of different types of priorities and focus on important priorities to maximize data quality while minimizing data collection requirements.

Further complicating the issue of failure, the ability of sanitation technologies to address priorities is not well known. In most sanitation monitoring efforts, implementers evaluate outcomes (e.g., functionality (Mercy Corps 2006), health (Crawford and Bryce 2003)), usually without direct comparisons to communities’ priorities. Finally, there is a growing effort to measure the social sustainability of sanitation. Other research has proposed social indicators such as

“acceptance” (Balkema et al. 2002; Kalbar et al. 2012a) and “appropriateness to local context”

(Lennartsson et al. 2009; McConville and Mihelcic 2007; Starkl et al. 2015), but many of these indicators lack consensus, empirical validation, or clear measurement methods. Technology selection and monitoring processes could be better informed if implementers could analyze and quantify how well existing systems, and potential new sanitation technologies, address priorities.

For example, sanitation systems that recover energy, water, or nutrients from wastewater (i.e., resource recovery systems) have been introduced as an option to increase priority addressment

(Guest et al. 2009b). Some studies have analyzed potential benefits from resource recovery, such as by evaluating the ability of these technologies to meet Sustainable Development Goals at a large-scale (Trimmer et al. 2017) or to offset costs (Mwirigi et al. 2014; Orner and Mihelcic 2018), but most do not evaluate potential benefits within the context of a community’s goals and priorities.

Also, given culture-specific, and possibly case-specific, priorities and given the many types of resource recovery technologies available, it can also be important to evaluate which resources (e.g., biogas) can address the most priorities.

Therefore, this research created a “priority addressment protocol” that identifies priorities and measures how well sanitation systems address them, based upon their importance to each

47 community (Figure 3.1). This protocol was applied in 20 resource-limited communities (i.e., cases) in India with community-based sanitation systems to: (1) identify sanitation and community priorities; (2) evaluate the ability of existing sanitation systems to address priorities; and (3) analyze the potential of different conventional and resource recovery technologies to improve priority addressment. The results and new protocol can help identify the most appropriate sanitation systems and design improvements that can encourage greater sanitation acceptance, use, and success.

Figure 3.1. Overview of the priority addressment protocol used to identify priorities and evaluate the current ability and potential of different sanitation systems to address priorities. Purple indicates priority identification, blue indicates priority ranking, and green indicates priority addressment.

49 Methods Case Descriptions & Priority Identification Twenty peri-urban, low-income, slum resettlement cases in southern India were selected

(Table F1), detailed in Davis et al. (2018) Cases were resettled by government agencies to peri- urban areas to improve living conditions for India’s lowest caste, but these resettlements still often lacked important infrastructure such as water or electricity. Each case had one centralized sanitation system. The systems served 800 to 1000 users, were implemented by external organizations, and were maintained by communities. Cases were selected to ensure comparable demographics, size, geography, and income but different sanitation technologies and functional statuses (Table F1). Priority identification used three qualitative data collection methods: interviews, focus groups, and photovoice. In photovoice, participants are usually given one to three days to take pictures in response to a prompt. In this research, participants had 24 hours to photograph their sanitation and community priorities; they then described the photos in a follow- up interview (Wang and Burris 1997). Data collection aimed to capture a representative cross- section of case perspectives (balanced gender, age, and geographic representation and multiple community roles) and used door-to-door sampling in the morning, midday, and evening to include a range of lifestyles (e.g., employed and unemployed individuals) (Davis et al. 2018b). Data collection concluded when theoretical saturation was achieved (i.e., when no additional priorities were identified by subsequent participants) for each case. In total, 232 interviews, 171 photovoice follow-up interviews, and 20 focus groups (with 189 focus group participants) were completed across all 20 cases. Data was coded inductively (Saldana 2009) (see Tables F2 and F3 for coding dictionaries) and triangulated between the three methods to identify one unordered list of sanitation priorities and one unordered list of community priorities per case. Priorities were then ranked using the Analytical Hierarchy Process (AHP) in each case by an additional focus group (called the AHP

50 group). Each priority was assigned an AHP weight, which represents a priority’s relative importance (a case’s sanitation AHP weights sum to one). These methods are detailed in Davis et al. (2018) and all data collection followed protocol #16-0026 (approved by University of Colorado

Boulder IRB in January 2016).

Priority Addressment A priority addressment protocol was created to quantify the current ability and the potential of different sanitation systems to address priorities (Figure 3.1) and was evaluated in three scenarios (Table 3.1). The first, called the Current Scenario, evaluated how well a case’s existing sanitation system—under existing use, maintenance, and performance conditions—addressed their sanitation and community priorities. The second, a hypothetical scenario called the Intended

Design Scenario, evaluated how well an existing sanitation system would address priorities if the system was functional. The third, also a hypothetical scenario, called the Added Resource

Recovery Scenario, evaluated how well a sanitation system with resource recovery would address priorities if the system was functional and designed to recover biogas, water, and compost.

Table 3.1. Description of Current, Intended Design, and Added Resource Recovery Scenarios. Scenario Technology Type System Status Functional or nonfunctional (based on existing Current Existing system and current use, operation, and maintenance) Functional; made hypothetical changes to Intended Existing existing system (from Current Scenario) such Design that it became functional Resource Recovery; made hypothetical Added Functional; made hypothetical changes to changes to existing system such that it had Resource existing system (from Current Scenario) such full resource recovery (biogas, water, Recovery that it became functional compost) capability

In the Current Scenario, there were five functional conventional, five nonfunctional conventional, five functional resource recovery, and five nonfunctional resource recovery systems

(Table F1). Functional status was defined using three criteria: the sanitation system was (1) compliant with local regulations (Central Pollution Control Board 2017) for pH and chemical and

51 biochemical oxygen demands, determined from effluent water quality tests; (2) adequately maintained, determined by triangulating data from observations, documentation (e.g., maintenance manuals), and system manager interviews; and (3) used continuously by more than 75% of the intended population, determined by triangulating data from observations, documentation (e.g., monitoring reports), and community member and system manager interviews; 75% was selected because health benefits increase significantly when the majority of a case’s population is using toilets (Haller et al. 2007; Hutton 2013) and all cases with regulatory-compliant systems had greater than 75% use (Figure F1).

Priority addressment data was collected using interviews, technical evaluations, and observations. An average of six interviews, with community members and system managers, were conducted per case (Table F4), focusing on how often the community typically experienced problems with each priority. Technical evaluations assessed the sanitation system’s construction quality, effluent water quality, odor, and cleanliness. Researchers observed system use and maintenance along with community life, such as status of roads and housing; on average, seven hours of observations were completed per case. Interview transcripts and observation notes were deductively coded (Saldana 2009) for design features and ongoing problems by two independent coders. Based on this information, each priority was assigned an “addressed value” to characterize the ability of the sanitation system to either always (1), usually (0.67), occasionally (0.33), or never

(0) address that priority (Figure 3.1).

In the Intended Design Scenario, each addressed value was re-evaluated based on this scenario’s hypothetical changes (Table 3.1) and the priority (Tables F5 and F7). In summary, an addressed value: (i) remained or increased to always addressed (i.e., 1.0) for priorities that could be completely addressed if the existing sanitation system was functional (e.g., Toilet Cleanliness);

52 (ii) increased by one addressment level (e.g., from occasionally, 0.33, to usually, 0.67, addressed) for priorities that could be partially addressed by sanitation but may also be influenced by other non-sanitation factors (e.g., Health & Hygiene); or (iii) remained unchanged from the Current

Scenario for priorities that could only be addressed with non-sanitation related changes or sanitation system design changes (e.g., Individual Septic Tanks).

In the Added Resource Recovery Scenario, each addressed value was also re-evaluated, based on this scenario’s hypothetical changes (Table 3.1) and the priority (Tables F6 and F8). In summary, an addressed value: (i) remained or increased to always addressed for priorities that could be completely addressed if the existing sanitation system was functional and recovered biogas, water, and compost; (ii) increased by one addressment level for priorities that could be partially addressed by resource recovery but may also be influenced by other non-sanitation factors; (iii) decreased by one addressment level for priorities that could be negatively affected by adding resource recovery but may also be influenced by other non-sanitation factors; or (iv) remained unchanged from the Intended Design Scenario for priorities that could only be addressed with non-sanitation related changes or other sanitation system design changes.

Finally, a “total case score” was calculated for each case’s sanitation and community priorities (Figure 3.1). Each priority’s addressed value was multiplied by its AHP weight. Then the weighted, addressed values were summed. The maximum total case score is 1, which means that all priorities are always addressed; the minimum is 0, which means that no priorities are ever addressed. Higher scores mean that more priorities and/or more of the most important priorities were addressed. To compare total case scores, one-way analyses of variance were performed at a

95% confidence level (α=0.05) (Judd et al. 2009).

53 Case 17’s Total Case Sanitation Scores. Case 17 is a representative example. It had 12 sanitation priorities with AHP weights ranging from 0.010 to 0.193 (Figure 3.2a). With an existing nonfunctional conventional sanitation system in the Current Scenario: No Open Defecation (AHP weight=0.010) was always addressed (i.e., addressed value=1.0) because 88% of the community used the sanitation system; Toilet Cleanliness (AHP weight=0.123) was usually addressed (i.e., addressed value=0.67) because households reported cleanliness issues on average twice a week;

Good Quality Construction (AHP weight=0.113) and Low Cost (AHP weight=0.035) were both occasionally addressed (i.e., addressed value=0.33) because the sewers were incorrectly constructed, which increased O&M costs by approximately 50%; the other eight were never addressed (i.e., addressed value=0). The AHP weighted sum of the 12 addressed values resulted in a Current Scenario total case sanitation score of 0.15 (Figure 3.2a), which was low because most priorities were not addressed by the existing system, especially those priorities most important to the case (e.g., Better Sanitation Planning and Water Reuse, with the highest and second highest

AHP weights of 0.193 and 0.187, respectively).

Figure 3.2. Case 17 presents a representative example for how the priority addressment protocol was applied to determine the extent that sanitation priorities (a) and community priorities (b) were addressed in each scenario. Colors represent how well each priority was addressed or could be addressed in the Current (blue), Intended Design (orange), and Added Resource Recovery (grey) scenarios. A wedge’s height (color fill) represents a priority’s addressed value for each scenario, where zero is never addressed, 0.33 is occasionally addressed, 0.67 is usually addressed, and 1.0 is always addressed. Stacked wedges for each priority are additive, even between scenarios; for a given priority, stacked colors indicate that subsequent scenarios improved priority addressment. For example, the community priority Jobs & Income has an AHP weight of 0.173 (wedge width) and addressed values of 0.33 for the Current Scenario (blue), 0.67 for the Intended Design Scenario (stacked orange), and 1.0 for the Added Resource Recovery Scenario (stacked grey). The community priority Good Sanitation System has an AHP weight of 0.027 and addressed values of zero for the Current Scenario (no blue shown since it is unaddressed), 1.0 for the Intended Design Scenario (orange), and 1.0 for the Added Resource Recovery Scenario (no added grey because no change from the Intended Design Scenario). The grey arrow and * for Low O&M Demands indicates where the addressed value decreased from 0.67 in the Intended Design Scenario to 0.33 in the Added Resource Recovery Scenario.

In the Intended Design Scenario, the following hypothetical changes were made to Case

17’s existing sanitation system such that it was maintained and used as intended by the original

55 design: (i) a skilled operator performed all maintenance; (ii) sewer blockages were removed and grading was improved; (iii) treatment tank material quality was improved; (iv) sewer and treatment tank access lids were replaced; and (v) the fence around the treatment system was repaired.

Accordingly, the system was then classified as functional because: (i) it became fully used (i.e., increasing use from 88% to 100%); (ii) it became properly maintained (i.e., increasing maintenance tasks performed from 11% to 100%); and (iii) it started to meet effluent water quality regulations. The hypothetical changes resulted in No Open Defecation remaining always addressed and six more sanitation priorities becoming always addressed: Toilet Cleanliness, Good O&M,

Good Quality Construction, Functioning Treatment System, No Smells, and Security for Treatment

System. Low O&M Demands and Low Cost changed from occasionally to usually addressed because the well-constructed, functional system would require less emergency maintenance. Better

Sanitation Planning became occasionally addressed since good planning is positively correlated with sanitation system functionality (Schmitt et al. 2017). Biogas and Water Reuse remained never addressed because the system was not designed to recover these resources. The Intended Design

Scenario total case sanitation score was 0.62. The higher score showed that improving maintenance of the existing sanitation system could help address many of Case 17’s most important sanitation priorities.

In the Added Resource Recovery Scenario, the following hypothetical changes were made to the existing system: (i) the system became functional (same changes as in Intended Design

Scenario), and (ii) the existing system’s design was altered to include resource recovery technologies for biogas (digester), water (toilet flushing and irrigation piping), and compost (onsite static piles). The seven sanitation priorities that were always addressed in the Intended Design

Scenario remained always addressed. Three more became always addressed: Water Reuse and

56 Biogas because the system’s design was changed from having no to full (water, biogas, compost) resource recovery capability; Low Cost because income from the sale of biogas, water, and compost could offset O&M costs and because sludge production would likely decrease since desludging is the largest contributor to sanitation O&M costs (Eales et al. 2013). Only one other sanitation priority’s addressed value changed in this scenario: Low O&M Demands decreased from usually to occasionally addressed because resource recovery systems have more complex O&M needs, such as frequent biogas digester monitoring (Eales et al. 2013; Tilley et al. 2014b). The

Added Resource Recovery Scenario total case sanitation score was 0.84; this scenario had the highest score because recovering biogas, water, and compost could address more sanitation priorities.

Case 17’s Total Case Community Score. There were 11 community priorities with AHP weights ranging from 0.013 to 0.232 (Figure 3.2b). In the Current Scenario, no community priorities were always addressed; Electricity and Drainage were usually addressed; Jobs &

Income, Government Support, Potable Drinking Water, Solid Waste Management, and Water

Supply were occasionally addressed; and Community Cleanliness, Health & Healthcare, Fence

Around Lake, and Good Sanitation System were never addressed. The Current Scenario total case community score was 0.19. In the Intended Design Scenario, Good Sanitation System became always addressed because this system was hypothetically functional. Community Cleanliness and

Health & Healthcare went from never to occasionally addressed because a functional sanitation system could partially reduce lake pollution and exposure to fecal pathogens (Mara et al. 2010).

Jobs & Income became usually addressed because a functional sanitation system could provide economic benefits of reduced health costs (WSP 2012), missed workdays (Hutton et al. 2004), and increased employment (Ajayi 2006). The other five addressed values remained unchanged. The

57 Intended Design Scenario total case community score was 0.39. In the Added Resource Recovery

Scenario, Electricity and Jobs & Income increased by one addressment level from the Intended

Design Scenario because biogas could be used for electricity generation and because the recovered resources could be sold or used to reduce expenditures on energy, water, and fertilizer. Water

Supply changed from occasionally to usually addressed because recycled water could reduce scarcity but not meet all water demands. The Added Resource Recovery Scenario total case community score was 0.48.

Results and Discussion Each case’s AHP group decided that the final unordered priority lists were valid and comprehensive. In nine cases, one sanitation or community priority was voiced by only one participant, but each AHP group confirmed that it was important. For example, in Case 7, the male operator was the only participant to voice Security for Treatment System, but the AHP group members shared this value. In eight cases, one AHP group member wanted to remove a priority, but the AHP group always chose to keep the full list. For example, during Case 14’s AHP group, a 20-year-old community member wanted to remove Water Savings from the list, but older community members remembered earlier droughts and reinforced its importance.

Sanitation Priorities On average, each case identified 12 sanitation priorities. Overall, 52 different sanitation priorities were identified (Table 3.2). Only six were very common (i.e., shared by 10 or more cases): No Smells, Functioning Treatment System, Toilet Cleanliness, No Open Defecation, Safety

& Dignity, and Water Reuse. Additionally, 12 cases expressed at least one economic-related priority, either: Low Cost, Income Generation, Micro-Loan Program, Jobs for Women, or

Women’s Empowerment. Many are common to other peri-urban communities, who often express priorities related to system performance, maintenance, access, safety, and cost (Bapat and Agarwal

58 2003; Burra et al. 2003; Lloyd-Jones and Rakodi 2014; Nawab et al. 2006) and a willingness to adopt resource recovery (e.g., water reuse) systems (Diener et al. 2014; Gupta 2015). The following less-common priorities have also been previously identified in the literature: Good O&M

(Novotný et al. 2018), Good Quality Construction (Novotný et al. 2018), Water Supply at Toilets

(Burra et al. 2003), Reduced Waiting Time (Bapat and Agarwal 2003), Comfortable (Nawab et al.

2006), Child-Friendly Toilets (Burra et al. 2003), Privacy (Jordan and Wagner 1993), Biogas

(Diener et al. 2014), Low Cost (Bapat and Agarwal 2003), Compost (Diener et al. 2014), Health

& Hygiene (Lloyd-Jones and Rakodi 2014), and Community Involvement in Planning (Jordan and

Wagner 1993).

Several sanitation priorities were shared by at least two cases but were identified for different reasons. For example, Move Toilets Away from Kitchen/Prayer Room was expressed in

Cases 1, 7, and 15 because of cleanliness concerns and in Case 2 because of space concerns.

Western Toilets was identified in Cases 4, 5, and 7 because the elderly struggled with squat-plate toilets and in Cases 6, 9, and 18 because community members wanted “modern” facilities. Repair

System Damage was stated in Case 3 because of broken access covers, in Case 10 because of cracked pipes, and in Case 12 because of broken doors on toilet stalls. Due to smells and blockages in small, local sewer systems, Cases 1, 3, 8, 10, 16, and 17 wanted Individual Septic Tanks while

Cases 9 and 19 wanted a Direct Municipal Sewer Connection instead of a local treatment system.

Additionally, cases ranked the shared priorities differently. The importance of the six most commonly expressed priorities varied between cases (Figure F2); for example, the rank of Toilet

Cleanliness ranged from first (AHP weight=0.22) to ninth (AHP weight=0.04) among the cases.

Also, 12 of the 20 cases had different priorities ranked first. The highest-ranked priorities were often those that cases perceived to have the greatest potential to benefit or disrupt their sanitation

59 systems. For example, Toilet Cleanliness was ranked first (AHP weight=0.22) in Case 14 because most community members stopped using the toilets due to poor maintenance and uncleanliness. In contrast, Case 11 ranked Water Supply at Toilets first (AHP weight=0.26) because they wanted to keep their constantly available water supply, which was important for users. This shows that the priority assessments identified comprehensive lists of sanitation priorities and not only outstanding problems.

Many sanitation priorities were also case-specific; with many not yet identified in the literature. Although cases had major similarities (i.e., resource-limited slum resettlements in southern India with centralized treatment systems), 18 priorities were expressed only once (i.e., by one case) (Table 3.2), including Water Savings, Jobs for Women, Child-Friendly Toilets, and Stop

Antisocial Elements. This is likely because valuation of sanitation is informed by community-level culture (Kaminsky 2015; Uddin et al. 2014). For example, Case 20 valued Income Generation for the whole community’s benefit while Case 18 specifically valued Jobs for Women because both males and females recognized women’s employment as beneficial. Additionally, priorities were case-specific due to different sanitation experiences. For example, Case 1 valued Treat Kitchen

Greywater because community members had previously lived in a community with this capability.

Case 10’s operator was negligent, so they valued Community O&M Training for community members. Overall, finding case-specific priorities and AHP weights across cases with similar characteristics shows the need to complete context-specific assessments in every case.

60 Table 3.2. The 52 sanitation priorities identified across all 20 cases, and the number of cases that shared each priority. Priority definitions are in Tables F2 and F3. Priorities are grouped by Performance/O&M, Use & Access, Benefits, and Planning & Design for readability. The most commonly expressed priorities (shared by 10 or more cases) are highlighted in yellow and bolded. Priorities that could not be addressed by at least one case’s existing system design are highlighted in grey and italicized. R denotes priorities that were influenced by resource recovery in the Added Resource Recovery Scenario.

Priorities Related to Priorities Related to Priorities Related to Priorities Related to Performance/O&M Use & Access Benefits Planning & Design # of # of # of # of Priorities Priorities Priorities Priorities Cases Cases Cases Cases R No Smells 16 No Open Defecation 14 Water Reuse 10 Treatment System Far Away 7 R Functioning Treatment System 15 Safety & Dignity 11 Biogas 8 Western Toilets 7 R Toilet Cleanliness 15 Water Supply at ToiletsR 9 Low Cost 8 Individual Septic Tanks 6 R Good O&M 9 Visual Aesthetics 5 Compost 6 Increase Sewer Pipe Size 5 No Sewer Blockages 9 Reduced Waiting Time 4 Health & Hygiene 1 Move Toilets Away from Kitchen/Prayer Rooms 4 R Good Quality Construction 7 Comfortable 3 Income Generation 1 Bathing Facilities at Toilets 2 R R Low O&M Demands 5 Multi-Use Area (Park) 3 Jobs for Women 1 Central Location 2 R Security for Treatment System 4 Sanitary Napkin Disposal 3 Micro-Loan Program 1 Direct Municipal Sewer Connection 2 R Repair System Damage 3 Open 24 Hours 2 Water Savings 1 Move Manholes to Grade 2 R Government Support for O&M 2 Privacy 2 Women's Empowerment 1 Better Sanitation Planning 1 Stop Shower Drain Clogging 2 Child-Friendly Toilets 1 Community Involvement in Sanitation Planning 1 Community O&M Training 1 Easy to Use 1 Shade for Sanitation Caretakers 1 R Efficient & Functional Treatment System Pumps 1 Lights in Toilets 1 Treat Kitchen Greywater 1 Treatment of Wastewater 1 Stop Antisocial Elements 1 Toilets in All Houses 1

61 Current Scenario. Overall, the existing sanitation systems did not address priorities well

(Figure 3.3). In all 20 cases, regardless of technology or status, no system always addressed all sanitation priorities (all scores were less than 1.0). The average Current Scenario total case sanitation score was 0.42. Since sanitation priorities were not assessed prior to system implementation in 17 cases (Figure 3.3), it is likely that the poor addressment occurred because priorities were unknown. Additionally, some were unaddressable, usually due to the system’s design or nonfunctionality. Seventeen of the 20 cases had at least one unaddressable priority, with

13 cases having at least one in their top five most important priorities, usually ranked first. Fourteen priorities required significant planning and design changes that were too expensive or complex

(Table 3.2). For example, in Cases 1 and 8, septic tanks were not installed because the government would only pay for a centralized treatment system; and in Cases 3, 10, 16, and 19, septic tanks were not installed due to groundwater contamination risk. Western toilets were too expensive in all seven cases (Cases 1, 4, 5, 6, 7, 9, 18). In Case 16, Treatment System Far Away could not be addressed because a cultural heritage site protected their desired location. Additionally, some cases valued resource recovery (Compost, Water Reuse, Biogas), but those priorities were not addressed due to high costs (Cases 1, 3, 15), insufficient space (Cases 9, 20), or inadequate implementer design knowledge (Case 8).

Total case sanitation scores were lower for nonfunctional (average=0.18) than functional

(average=0.66) sanitation systems because functional systems addressed more sanitation priorities

(p=0.000). Case knowledge indicates that unaddressed priorities may be both a cause and an effect of system nonfunctionality. To help understand this relationship, all cases were re-analyzed as hypothetically functional in the Intended Design Scenario. In addition, of the Current Scenario functional systems, total case sanitation scores were lower for conventional (average=0.50) than

62 for resource recovery (average=0.82) systems, so the Added Resource Recovery Scenario explored the potential of resource recovery

Intended Design Scenario. Overall, if the existing sanitation systems could function according to their design then they could address more sanitation priorities (Figure 3.3). With an average score of 0.75, all cases had a higher total case sanitation score in the Intended Design

Scenario than in the Current Scenario (Table F9). Systems that were nonfunctional in the Current

Scenario had large score increases, on average by 475% (Table F10). For example, Case 12’s total case sanitation score increased from 0.35 to 0.82 because the existing nonfunctional resource recovery system hypothetically changed from: (i) 32% to 100% used, so income from user fees increased such that Micro-Loan Program became usually addressed; (ii) 7% to 100% maintained, so an operator performed maintenance and repaired the fence, such that Water Supply at Toilets,

Biogas, Toilet Cleanliness, Security for Treatment System, and Visual Aesthetics became always addressed; and (iii) 0% to 100% of regulations met, so No Smells and Functioning Treatment

System became always addressed. Systems that were already functional in the Current Scenario had smaller score increases, on average by 26%. For example, Case 11’s total case sanitation score increased from 0.82 to 0.98 in the Intended Design Scenario because the existing functional resource recovery system hypothetically changed from 92% to 100% maintained, so small cracks in the digester were fixed such that Biogas became always addressed. No total case sanitation score was 1.0 in the Intended Design Scenario because no case had a system that, even when functional, could always address all of their sanitation priorities.

Added Resource Recovery Scenario. Hypothetically functional, full resource recovery systems resulted in the highest average total case sanitation score of 0.81. More sanitation priorities were addressed in the Added Resource Recovery Scenario than in the Current Scenario (p=0.000),

63 but priority addressment was similar between the Added Resource Recovery and Intended Design scenarios (p=0.325) (Table F9). This suggests that there may be limited benefits to adding resource recovery to address sanitation priorities. In addition, although fewer resource recovery-related priorities were expressed by cases with existing conventional systems (average=2) than cases with existing resource recovery systems (average=4), conventional systems’ scores increased more in this scenario because their resource recovery-related priorities could be met by adding or changing technologies.

For the 10 cases that already had some resource recovery, total case sanitation scores from the Intended Design to the Added Resource Recovery Scenario did not change in four cases (Cases

5, 9, 11, 14), increased in four cases (Cases 12, 13, 18, 20, by an average of 6%), and decreased in two cases (Cases 2, 6, both by 2%) (Figure 3.3). Systems with existing resource recovery technologies were already well-aligned with each case’s priorities, so adding greater resource recovery capability in this scenario had minimal benefits. For example, in the Intended Design

Scenario, Case 11’s functional DEWATS system (with a biogas digester and onsite irrigation) produced biogas, which was sold as cooking fuel, and recycled water, which irrigated a profitable vegetable farm. Therefore, no further resource recovery was needed to address Case 11’s priorities.

However, for four of these 10 cases, additional resources needed to be recovered to address priorities. For example, Cases 12, 18, and 20 valued Compost, which was never addressed by their existing DEWATS design that only recovered biogas and water. While compost recovery would increase O&M tasks, total case sanitation scores increased when compost recovery was hypothetically added because Case 12 valued Compost more than Low O&M Demands (AHP weights of 0.05 and 0.01, respectively), and Cases 18 and 20 did not value Low O&M Demands.

In contrast, total case sanitation scores decreased for Cases 2 and 6 due to the hypothetical addition

64 of compost recovery; compost could not address any priorities in either case, and it instead decreased the addressment of Low O&M Demands due to the increase in composting O&M tasks.

The total case sanitation scores for the other 10 cases, which had existing conventional systems that were hypothetically changed to include full resource recovery, did not change in two cases (Cases 7, 16), increased in seven cases (Cases 1, 3, 8, 10, 15, 17, 19, by an average of 25%), and decreased in one case (Case 4 by 0.2%) from the Intended Design Scenario to the Added

Resource Recovery Scenario. The two cases with unchanged scores expressed no resource recovery-related sanitation priorities. For seven cases, though, resource recovery systems better addressed priorities than conventional systems. For example, Water Reuse in Cases 1, 3, and 17 was only addressed in the Added Resource Recovery Scenario since these cases’ existing systems were not designed to recycle water. Low Cost in Cases 3, 4, 10, 17, and 19 became better addressed because resource recovery provides opportunities for cost savings (e.g., reducing expenses due to recycling water) and income generation (e.g., selling biogas). Finally, in Case 4, adding resource recovery resulted in a trade-off because it could reduce costs while also increasing technological complexity and maintenance burdens; since Low O&M Demands (AHP weight=0.08) was valued more than Low Cost (AHP weight=0.05), the total case sanitation score decreased.

An important consideration is that some communities reject resource recovery because the benefits do not outweigh the additional maintenance costs (Diener et al. 2014; Nawab et al. 2006).

This is highlighted with the three cases that had total case sanitation scores that decreased in this scenario. Additionally, there can be psychological and cultural barriers towards adopting resource recovery systems; for example, composting toilets are sometimes viewed as a sign of poverty

(Nawab et al. 2006) and water and fertilizer from sanitation systems can be considered unclean

(Mariwah and Drangert 2011; Mwirigi et al. 2014; Walekhwa et al. 2009). While adding resource

65 recovery increased the scores for the majority of cases, all scenario’s total case sanitation scores were still less than one (Figure 3.3), demonstrating that there are still opportunities to improve service delivery, technology development, and priority alignment.

Figure 3.3. Comparison of the ability of the Current Scenario (blue) and potential of the Intended Design (orange) and Added Resource Recovery (grey) Scenarios to address sanitation priorities. Stacked columns are additive between scenarios. The cases are grouped based on each case’s Current Scenario system status (functional or nonfunctional) and technology type (conventional or resource recovery). An arrow indicates when a case’s score decreased (by 0.03 or less) between the Intended Design and Added Resource Recovery Scenarios, which happened in three cases (4, 6, 2), and the black horizontal line shows the decreased score value. A * next to the case number indicates that a limited prior priority assessment (on only community priorities) was conducted by implementers; ** indicates that a comprehensive prior priority assessment (on community and sanitation priorities) was conducted by implementers.

Community Priorities On average, each case identified 13 community priorities; a total of 40 different community priorities were identified from the 20 cases (Table F11). The 12 most commonly expressed priorities have been previously identified: Community Cleanliness (Lloyd-Jones and Rakodi

2014), Good Sanitation System (Hubbard et al. 2011), Drainage (Bapat and Agarwal 2003), Solid

66 Waste Management (Lloyd-Jones and Rakodi 2014), Jobs & Income (Lloyd-Jones and Rakodi

2014), Water Supply (Jordan and Wagner 1993), Education (Lloyd-Jones and Rakodi 2014),

Health & Healthcare (Lloyd-Jones and Rakodi 2014), House Improvements (Lloyd-Jones and

Rakodi 2014), Potable Drinking Water (Jordan and Wagner 1993), Government Support (Lloyd-

Jones and Rakodi 2014), and Cooking Fuel (Bapat and Agarwal 2003). The importance of these priorities also varied greatly between cases (Figure F3). The less-common community priorities were often case-specific, with 16 expressed by one case apiece, including Micro-Loans for Women,

Land Ownership, and Library (Table F11), and were not yet identified in the literature. Priorities were likely different between cases due to differences in existing infrastructure and services. For example, only Case 1 valued Graveyard since they were the only case without one. Overall, only

11 community priorities were related to sanitation or resource recovery: Community Cleanliness,

Good Sanitation System, Jobs & Income, Water Supply, Health & Healthcare, Cooking Fuel,

Eradicating Pests, Electricity, Overall Community Development, Micro-Loans for Women, and

Women’s Empowerment.

Current Scenario. Community priorities, more than sanitation, were poorly addressed in the Current Scenario. The average total case community score was 0.31 (minimum=0.17; maximum=0.55) (Figure F4). The low scores were expected since 29 community priorities were unrelated to sanitation (e.g., Road Improvements). Also, all priorities related to sanitation or resource recovery, except Good Sanitation System, were influenced by multiple factors and could not be fully addressed by a sanitation system alone (e.g., Health & Healthcare). Similar to sanitation, more community priorities were addressed by functional than nonfunctional sanitation systems (score averages of 0.37 and 0.25, respectively).

67 Intended Design Scenario. Hypothetical functionality had minimal benefits. The total case community scores remained low across all 20 cases (average=0.44) in this scenario because many community priorities do not relate to sanitation and most sanitation systems are not designed to meet multiple infrastructure or social needs. However, hypothetically changing the 10 existing nonfunctional systems to become functional allowed them to address more community priorities

(scores increased by 98% on average) (Table F12); they could then always address Good

Sanitation System and better address Community Cleanliness, Health & Healthcare, Eradicating

Pests, Jobs & Income, and Overall Community Development. Only two of the 10 cases with existing functional systems had total case community scores change in this scenario (Cases 2, 11).

The scores increased because hypothetical digester repairs resulted in biogas production and sales; therefore, Electricity, Cooking Fuel, and Jobs & Income could be better addressed.

The protocol’s score evaluation should be updated as more research on possible impacts becomes available and should be as context-specific as possible. There are mixed findings on the amount of benefits from sanitation (some studies show a range of benefits (Haller et al. 2007;

Hutton et al. 2004; Mara et al. 2010) while some find no impacts (Aboud and Yousafzai 2018;

Tofail et al. 2018)), and not all benefits can be realized in certain contexts (e.g., a market for biogas needs to be available for it to have an economic benefit). Due to currently limited information, two main assumptions were used. First, since the uncertainty analysis demonstrated that the value of a single priority’s addressment score alone was relatively uninfluential on the total case score

(Figure F5), an assumption of maximum potential benefits from sanitation, as found in literature, was used (e.g., functional sanitation systems and resource recovery positively influence health and income). Second, it was assumed that all priorities associated with a hypothetical change would be simultaneously improved. This large improvement would be unlikely (e.g., since there is a

68 limited amount of biogas that can be produced from small sanitation systems), but this second assumption was used because it is uncertain which priority, over other more and less important priorities, would be addressed and because total case scores and results were insensitive to this calculation approach (i.e., the scores were equal or decreased by only 2% ± 6% when only one versus multiple related priorities’ addressed values were changed) (Figure F5).

Added Resource Recovery Scenario. In this scenario, the average total case community score was 0.49, the highest of the three scenarios. However, this scenario’s individual case scores were not significantly different from the Intended Design Scenario’s scores (p=0.212) (Table F13); there was no change in scores for three cases and less than a 20% change for the other 17 cases

(Figure F4). Scores increased in all 10 cases that had conventional systems hypothetically changed to resource recovery systems (scores increased 19% on average) and in seven cases that already recovered one resource and were hypothetically changed to recover two additional resources

(scores increased 8% on average). The scores did not change for three cases because their priorities were well-addressed by the existing designs, which already included two or three types of resource recovery. Although these scores did not change, there could be a benefit from additional resource recovery not quantified with the “always addressed” definition, such as the benefit of reducing resource scarcity. Overall, the minimal changes in scores suggest that there might be diminishing benefits of adding multiple types of resource recovery. Of the three resources, biogas and water had greater potential to address more sanitation and community priorities than compost. For example, Biogas, valued in nine cases, could partially address Electricity in eight cases and

Cooking Fuel in 10 cases. Water Reuse, valued in 10 cases, could partially address Water Supply in 17 cases. Compost was valued in six cases, but since all cases lacked agricultural opportunities, no community priorities could be addressed by compost.

69 Implications and Priority Assessment Importance Overall, low total case scores for the Current Scenario highlight that existing systems poorly address priorities because priorities were unknown or unaddressable or because systems were nonfunctional. Score increases in the hypothetical scenarios demonstrate that improvements to sanitation technology design and service delivery could address more priorities. This shows the need to evaluate why systems are failing using a systematic approach. Additionally, most priorities and rankings were case-specific. Only nine (Cases 4, 6, 8, 9, 11, 15, 18, 19, 20) had their priorities evaluated prior to sanitation implementation, with only three (Cases 11, 18, 20) having both sanitation and community priorities assessed using a diversity of community perspectives (Davis et al. 2018b). This shows the importance of implementers conducting assessments in each case.

Prior priority assessments, though, were not correlated with system status, system technology, or total case scores within the 20 cases (Figures 3.2 and F5). Knowing priorities is still expected to be important (Murphy et al. 2009; Wood et al. 2016), so this lack of a trend may be because assessments focused exclusively on problems. For example, many of the prior evaluations were limited to community priorities. In this study, it was found that community priorities mostly reflected a snapshot of current problems; 85% of community priorities were current problems

(unaddressed). This is likely because cases lacked access to basic infrastructure and services. In contrast, sanitation priorities included problems and existing capabilities that cases wanted to keep; only 55% of sanitation priorities were problems (unaddressed). Consequently, assessing overall community priorities may miss existing services and infrastructure that could be a problem in the future and must be maintained. Therefore, priority assessments should be context-specific and encourage participants to identify priorities that are not just current problems so that both short- term and long-term needs can be considered.

70 Further, no case’s prior assessment identified the most important priorities. While ranking priorities can have limitations (Hino and Imai 2018), knowing their relative importance can identify interventions that maximize incentives for a case to use and maintain a sanitation system.

Also, it helps to quantify addressment, which can be used to compare sanitation systems’ social sustainability. While many frameworks include indicators for social sustainability, such as acceptance (Molinos-Senante et al. 2014; Palme et al. 2005), satisfaction (Iribarnegaray et al. 2012;

Plakas et al. 2016), appropriateness to local context (Lennartsson et al. 2009), and cultural sensitivity (Balkema et al. 2002), these frameworks call for method development to measure these social indicators, do not define the indicators (e.g., do not state how to measure them), or state that indicators should be adapted to local context without providing that guidance. This study’s priority addressment protocol uses concepts from existing social indicators, such as acceptance and satisfaction, to identify specific ways that sanitation systems could be improved to increase social sustainability. The protocol does this by combining context-specific priorities (i.e., individual indicators determined by communities themselves) into one quantitative indicator. This protocol and resulting social sustainability indicator (i.e., total case score) was used to evaluate sanitation; it can also be used to evaluate a diverse range of engineering systems (e.g., drinking water, energy).

Researchers and implementers can use this study’s results and priority addressment protocol to elucidate which technologies and strategies minimize tradeoffs and meet the most priorities long- term.

Acknowledgements We thank our assistants, Vijay Kumar and Sridhar Selvaraj, and all participants in this research for their indispensable time and support.

71 Funding Sources This work was completed with financial support from the Mortenson Center in Engineering for Developing Communities.

Chapter 3 References Aboud, F. E., and Yousafzai, A. K. (2018). “Health and nutrition interventions for infant development.” The Lancet Child & Adolescent Health, 2(4), 231–233. Ajayi, R. (2006). “Women’s participation in self-help community development projects in Ndokwa agricultural zone of Delta State, Nigeria.” Community Development Journal, 41(2), 189–209. Balkema, A. J., Preisig, H. A., Otterpohl, R., and Lambert, F. (2002). “Indicators for the sustainability assessment of wastewater treatment systems.” Urban Water, 4(2), 153–161. Bapat, M., and Agarwal, I. (2003). “Our needs, our priorities; women and men from the slums in Mumbai and Pune talk about their needs for water and sanitation.” Environment and Urbanization, 15(2), 71–86. Breslin, E. D. (2003). Demand response approach in practice: why sustainability remains elusive. WaterAid, Mozambique. Burra, S., Patel, S., and Kerr, T. (2003). “Community-designed, built and managed toilet blocks in Indian cities.” Environment and Urbanization, 15(2), 11–32. Central Pollution Control Board. (2017). “Environmental Standards: Water Quality Criteria.” Central Pollution Control Board, (May 17, 2016). Chambers, R. (1981). “Rapid rural appraisal: rationale and repertoire.” Public Administration and Development, 1, 95–106. Crawford, P., and Bryce, P. (2003). “Project monitoring and evaluation: a method for enhancing the efficiency and effectiveness of aid project implementation.” International Journal of Project Management, 21(5), 363–373. Davis, A., Javernick-Will, A., and Cook, S. (2018). “A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems.” Environmental Science: Water Research & Technology, 4, 1451– 1463. Davis, S. (2015). “Statistics on Sanitation Failures.” Improve International. Diener, S., Semiyaga, S., Niwagaba, C. B., Muspratt, A. M., Gning, J. B., Mbgur, M., Ennin, J. E., Zurbrugg, C., and Strande, L. (2014). “A value proposition: Resource recovery from faecal sludge? Can it be the driver for improved sanitation?” Resources, Conservation and Recycling, 88, 32–38. Eales, K., Blackett, I., Siregar, R., and Febriani, E. (2013). Review of Community-Managed Decentralized Wastewater Treatment Systems in Indonesia. WSP. Freudenberger, K. S. (2011). Rapid Rural Appraisal and Participatory Rural Appraisal: A Manual for CRS field Workers and Partners. Manual, Catholic Relief Services, Baltimore, Maryland. Guest, J. S., Skerlos, S. J., Barnard, J. L., Beck, M. B., Daigger, G. T., Hilger, H., Jackson, S. J., Karvazy, K., Kelly, L., Macpherson, L., Mihelcic, J. R., Pramanik, A., Raskin, L., Van Loosdrecht, M. C. M., Yeh, D., and Love, N. G. (2009). “A New Planning and Design

72 Paradigm to Achieve Sustainable Resource Recovery from Wastewater.” Environmental Science & Technology, 43(16), 6126–6130. Gupta, A. (2015). “EcoSan: Sustainable solution to sanitation problem in rural India.” Water, sanitation and hygiene services beyond 2015: Improving access and sustainability, Loughborough, England. Haller, L., Hutton, G., and Bartram, J. (2007). “Estimating the costs and health benefits of water and sanitation improvements at global level.” Journal of Water and Health, 5(4), 467– 480. Harris-Lovett, S. R., Binz, C., Sedlak, D. L., Kiparsky, M., and Truffer, B. (2015). “Beyond User Acceptance: A Legitimacy Framework for Potable Water Reuse in California.” Environmental Science & Technology, 49(13), 7552–7561. Hino, A., and Imai, R. (2018). “Ranking and Rating: Neglected Biases in Factor Analysis of Postmaterialist Values.” International Journal of Public Opinion Research, 1–14. Hubbard, B., Sarisky, J., Gelting, R., Baffigo, V., Seminario, R., and Centurion, C. (2011). “A community demand-driven approach toward sustainable water and sanitation infrastructure development.” International Journal of Hygiene and Environmental Health, 214(4), 326–334. Hutton, G. (2013). “Global costs and benefits of reaching universal coverage of sanitation and drinking-water supply.” Journal of Water and Health, 11(1), 1–12. Hutton, G., Haller, L., and Water, S. (2004). Evaluation of the costs and benefits of water and sanitation improvements at the global level. World Health Organization, Geneva, Switzerland. Iribarnegaray, M. A., Copa, F. R., Gatto D’Andrea, M. L., Arredondo, M. F., Cabral, J. D., Correa, J. J., Liberal, V. I., and Seghezzo, L. (2012). “A comprehensive index to assess the sustainability of water and sanitation management systems.” Journal of Water, Sanitation and Hygiene for Development, 2(3), 205. Jordan, S., and Wagner, F. (1993). “Meeting women’s needs and priorities for water and sanitation in cities.” Environment and Urbanization, 5(2), 135–1445. Judd, C. M., McClelland, G. H., and Ryan, C. S. (2009). Data Analysis: A Model Comparison Approach. Routledge, New York, NY. Kalbar, P. P., Karmakar, S., and Asolekar, S. R. (2012). “Technology assessment for wastewater treatment using multiple-attribute decision-making.” Technology in Society, 34(4), 295– 302. Kaminsky, J. A. (2015). “Cultured Construction: Global Evidence of the Impact of National Values on Sanitation Infrastructure Choice.” Environmental Science & Technology, 49(12), 7134–7141. Lennartsson, M., Kvarnström, E., Lundberg, T., Buenfil, J., and Sawyer, R. (2009). Comparing sanitation systems using sustainability criteria. Stockholm Environment Institute, SEI, Stockholm. Lloyd-Jones, T., and Rakodi, C. (2014). Urban Livelihoods: A People-Centred Approach to Reducing Poverty. Routledge, London, UK. Mara, D., Lane, J., Scott, B., and Trouba, D. (2010). “Sanitation and Health.” PLoS Medicine, 7(11), 1–7. Mariwah, S., and Drangert, J.-O. (2011). “Community perceptions of human excreta as fertilizer in peri-urban agriculture in Ghana.” Waste Management & Research, 29(8), 815–822.

73 McConville, J. R., and Mihelcic, J. R. (2007). “Adapting Life-Cycle Thinking Tools to Evaluate Project Sustainability in International Water and Sanitation Development Work.” Environmental Engineering Science, 24(7), 937–948. Mercy Corps. (2006). Design, Monitoring & Evaluation (DM&E) Guidebook. Mercy Corps. Molinos-Senante, M., Gómez, T., Garrido-Baserba, M., Caballero, R., and Sala-Garrido, R. (2014). “Assessing the sustainability of small wastewater treatment systems: A composite indicator approach.” Science of The Total Environment, 497–498, 607–617. Murphy, H. M., McBean, E. A., and Farahbakhsh, K. (2009). “Appropriate technology – A comprehensive approach for water and sanitation in the developing world.” Technology in Society, 31(2), 158–167. Mwirigi, J., Balana, B. B., Mugisha, J., Walekhwa, P., Melamu, R., Nakami, S., and Makenzi, P. (2014). “Socio-economic hurdles to widespread adoption of small-scale biogas digesters in Sub-Saharan Africa: A review.” Biomass and Bioenergy, 70, 17–25. Nawab, B., Nyborg, I. L. P., Esser, K. B., and Jenssen, P. D. (2006). “Cultural preferences in designing ecological sanitation systems in North West Frontier Province, Pakistan.” Journal of Environmental Psychology, 26(3), 236–246. Novotný, J., Hasman, J., and Lepič, M. (2018). “Contextual factors and motivations affecting rural community sanitation in low- and middle-income countries: A systematic review.” International Journal of Hygiene and Environmental Health, 221(2), 121–133. Orner, K. D., and Mihelcic, J. R. (2018). “A review of sanitation technologies to achieve multiple sustainable development goals that promote resource recovery.” Environmental Science: Water Research & Technology, 4, 16–32. Palme, U., Lundin, M., Tillman, A.-M., and Molander, S. (2005). “Sustainable development indicators for wastewater systems – researchers and indicator users in a co-operative case study.” Resources, Conservation and Recycling, 43(3), 293–311. Plakas, K. V., Georgiadis, A. A., and Karabelas, A. J. (2016). “Sustainability assessment of tertiary wastewater treatment technologies: a multi-criteria analysis.” Water Science and Technology, 73(7), 1532–1540. Platteau, J.-P., and Abraham, A. (2002). “Participatory Development in the Presence of Endogenous Community Imperfections.” Journal of Development Studies, 39(2), 104– 136. Roma, E., and Jeffrey, P. (2010). “Evaluation of community participation in the implementation of community-based sanitation systems: a case study from Indonesia.” Water Science & Technology, 62(5), 1028–1036. Saldana, J. (2009). The Coding Manual for Qualitative Researchers. Sage, London. Schmitt, R. J. P., Morgenroth, E., and Larsen, T. A. (2017). “Robust planning of sanitation services in urban informal settlements: An analytical framework.” Water Research, 110, 297–312. Sperling, J., Romero-Lankao, P., and Beig, G. (2016). “Exploring citizen infrastructure and environmental priorities in Mumbai, India.” Environmental Science & Policy, 60, 19–27. Starkl, M., Brunner, N., Feil, M., and Hauser, A. (2015). “Ensuring Sustainability of Non- Networked Sanitation Technologies: An Approach to Standardization.” Environmental Science & Technology, 49(11), 6411–6418. Starkl, M., Brunner, N., and Stenström, T.-A. (2013). “Why Do Water and Sanitation Systems for the Poor Still Fail? Policy Analysis in Economically Advanced Developing Countries.” Environmental Science & Technology, 47, 6102–6110.

74 Tilley, E., Ulrich, L., Lüthi, C., Reymond, P., and Zurbrügg, C. (2014). Compendium of Sanitation Systems and Technologies. Eawag. Tofail, F., Fernald, L. C., Das, K. K., Rahman, M., Ahmed, T., Jannat, K. K., Unicomb, L., Arnold, B. F., Ashraf, S., Winch, P. J., Kariger, P., Stewart, C. P., Colford, J. M., and Luby, S. P. (2018). “Effect of water quality, sanitation, hand washing, and nutritional interventions on child development in rural Bangladesh (WASH Benefits Bangladesh): a cluster-randomised controlled trial.” The Lancet Child & Adolescent Health, 2(4), 255– 268. Trimmer, J. T., Cusick, R. D., and Guest, J. S. (2017). “Amplifying Progress toward Multiple Development Goals through Resource Recovery from Sanitation.” Environmental Science & Technology, 51(18), 10765–10776. Uddin, S. M. N., Muhandiki, V. S., Sakai, A., Al Mamun, A., and Hridi, S. M. (2014). “Socio- cultural acceptance of appropriate technology: Identifying and prioritizing barriers for widespread use of the urine diversion toilets in rural Muslim communities of Bangladesh.” Technology in Society, 38, 32–39. United Nations. (2015). Sustainable Development Goals. United Nations. Walekhwa, P. N., Mugisha, J., and Drake, L. (2009). “Biogas energy from family-sized digesters in Uganda: Critical factors and policy implications.” Energy Policy, 37(7), 2754–2762. Wang, C., and Burris, M. A. (1997). “Photovoice: Concept, Methodology, and Use for Participatory Needs Assessment.” Health, Education & Behavior, 24(3), 369–387. WHO. (2004). “Rapid Needs Assessment for Water, Sanitation and Hygiene.” WHO, and UNICEF. (2017). Progress on Drinking Water, Sanitation and Hygiene: Update and SDG Baselines. Joint Monitoring Program, Geneva. Wood, B. T., Dougill, A. J., Quinn, C. H., and Stringer, L. C. (2016). “Exploring Power and Procedural Justice Within Climate Compatible Development Project Design: Whose Priorities Are Being Considered?” The Journal of Environment & Development, 25(4), 363–395. WSP. (2012). “Economic Assessment of Sanitation Interventions in Cambodia.” The World Bank.

75 Chapter 4: The use of qualitative comparative analysis to identify

pathways to successful and failed sanitation systems

This chapter is the published journal article: Davis, A., Javernick-Will, A., and Cook, S. (2019). “The use of qualitative comparative analysis to identify pathways to successful and failed sanitation systems”. Science of the Total Environment. https://doi.org/10.1016/j.scitotenv.2019.01.291.

Abstract Art

Highlights

 Sanitation failure is not systematically understood

 Qualitative comparative analysis elucidated pathways for success and failure

 Successful systems needed O&M resources, local engagement, and community buy-in

 Failed systems lacked municipal engagement, addressed priorities, and O&M support

 Systems must be implemented with a holistic view to best use limited resources

Keywords

Sanitation, success, failure, qualitative comparative analysis, developing communities

76 Abstract Sanitation systems globally fail at high rates. Researchers and practitioners attribute the causes of both sanitation success and failure to numerous factors that include technical and non- technical issues. A comprehensive understanding of what leads to sanitation failure and how to achieve sanitation success is imperative to prioritize the use of limited resources. To determine which combinations of causal conditions led to successful and failed sanitation systems, we applied fuzzy-set qualitative comparative analysis to 20 cases in Karnataka and Tamil Nadu, India with small-scale sanitation systems. Two pathways led to successful sanitation systems, and four pathways led to failed sanitation systems. All successful systems required Sufficient O&M Funds, a Clear O&M Plan, and Technical Support in addition to either Addressed Sanitation Priorities and Community Participation in Planning or Behavior Change Education and Municipality

Involved in Planning. All failed systems had Lack of Municipality in Planning, Unaddressed

Sanitation Priorities, and No Technical Support. Most failed systems also had No Clear O&M

Plan, Poor Construction Quality, Lack of Community Participation in Planning, and Insufficient

O&M Funds. Two failed cases had unique pathways because Government Barriers permanently disrupted use and maintenance. Overall, implementing organizations who initiate sanitation projects in resource-limited communities should ensure that (1) communities have adequate technical and financial resources for maintenance; (2) community and municipality stakeholders are engaged in planning and know their maintenance responsibilities; and (3) appropriate technologies are selected that meet community needs and achieve community buy-in.

77 Introduction Sanitation systems fail at high rates, with up to 70% of sanitation systems in resource- limited communities failing within two years of construction (WHO and UNICEF 2017).

Sanitation failure is a major problem because it leads to human and environmental health risks

(WHO and UNICEF 2017). Despite the importance of and need for universal access to safe sanitation, sanitation systems continue to fail. Therefore, there is a need to comprehensively and systematically understand why systems are still failing to avoid negative outcomes and achieve sanitation success.

Previous research has identified many factors that influence sanitation success and failure.

For example, sanitation failure has been attributed to supply-driven approaches (Starkl et al.

2013b), lack of maintenance (Katukiza et al. 2010), faulty designs (Sujaritpong and Nitivattananon

2009), high operation and maintenance (O&M) costs (Cronin et al. 2014), lack of ongoing support

(Eales et al. 2013), lack of user acceptance (Bao et al. 2013), inadequate technical knowledge

(Kaminsky and Javernick-Will 2012), or inappropriate technologies (Murphy et al. 2009).

However, persistent sanitation failure suggests that these factors may not fully explain the causes of failure and/or these factors could be better addressed in the sanitation sector. Sanitation success has been attributed to community participation (Roma and Jeffrey 2010), user satisfaction

(Seymour 2014), affordability (Mwirigi et al. 2014), appropriate technologies (Black 1998;

Bouabid and Louis 2015; Murphy et al. 2009; Palaniappan et al. 2008), maintenance support

(Sansom 2011), and low maintenance complexity (Brikké and Bredero 2003). Despite this knowledge, studies have found contrary results for the same factor. For example, in a study of sanitation in India, Battacharyya (2015) found that community participation, good quality construction, and water supply were each positively correlated with success, but Banerjee (2013) found that these same factors are present in failed systems in India. Also, in a literature review by

78 Mansuri (2004), community participation had either positive, negative, or no impacts on sanitation outcomes; thus, those authors suggested that participation may not be the only driver of success or failure. Overall, analyzing factors in isolation has been insufficient to characterize the complex causes of sanitation success and failure.

While most research has analyzed individual factors associated with success or failure of sanitation systems, a few studies have focused on identifying combinations of factors. For example, one study found that poor quality construction of school toilets could be overcome by the simultaneous presence of multiple other factors, such as a maintenance plan, a sanitation champion, and government support (Chatterley et al. 2014). Another study, of rural water supply systems, found that success could still be achieved, despite the absence of post-construction support, if good financial management and community participation were both present (Marks et al. 2018). Comprehensive evaluations that identify which combinations of factors, including the integration of social, institutional, technical, and economic factors (Tilley et al. 2014a; Törnqvist et al. 2008; WaterAid 2011a), are needed to better understand sanitation systems’ outcomes.

A common method to holistically examine how factors combine together to produce an outcome is qualitative comparative analysis (QCA). QCA has been used extensively in the social sciences (Fischer and Maggetti 2017; Nair and Howlett 2015) and is being used increasingly in engineering (Jordan et al. 2016; Kunz et al. 2015; Opdyke et al. 2018). QCA results in both in- depth case knowledge and generalizable results (Ragin 1987) by using set theory and Boolean algebra to analyze combinations of factors (i.e., pathways) that lead to an outcome of interest

(Ragin 2008). There can be multiple pathways identified for a given outcome, allowing for a better understanding of the complex causes of success and failure and to highlight alternative pathways to success. QCA has been successfully used to evaluate pathways to success in resource-limited

79 communities for water supply systems (Marks et al. 2018) and for management of school toilets

(Chatterley et al. 2014), and this method may be useful to investigate sanitation systems to improve success and universal access.

To this end, this study used fuzzy-set QCA (fsQCA) to investigate the causes of success or failure of 20 small, community-based sanitation systems in India. Specifically, we aimed to determine which factors, in combination, led to sanitation system success or failure in order to provide recommendations for implementing organizations, municipalities, and communities to improve sanitation use, maintenance, and performance. The participating communities were in

India, which has the world’s fastest growing population and faces significant challenges for sanitation: more than 50% of sanitation systems in India have failed (Chaplin 2011) and 60% of the population lacks access to safely managed sanitation (WHO and UNICEF 2017). These issues of failure and lack of sanitation access are also present globally, thus it is important to understand strategies to reduce failure. We selected ten successful and ten failed systems and completed in- depth case studies in each community. Next, we used fsQCA to determine combinations of factors that led to success or failure of sanitation systems (Figure 4.1). Finally, by comparing the resulting pathways, we identified holistic strategies that lead to successful sanitation systems. Identifying sanitation success pathways can guide implementing organizations, communities, and governments to focus their limited resources to avoid failure-prone scenarios and promote success.

Figure 4.1. Overview of data collection and analysis methods used to identify pathways to success and failure of sanitation systems.

80 Methods Research Context The 20 communities (i.e., cases) in Karnataka or Tamil Nadu, India were selected to ensure variability between outcomes and factors (i.e., causal conditions). Specific case details and additional methods details are included in the Supporting Information (SI). Major variations between cases included sanitation technologies, implementing organizations, and current system outcomes (SI Table G1). To reduce the total number of causal conditions analyzed, we ensured each case had the following similarities: had one small-scale sanitation system, which was implemented by an external non-governmental organization (i.e., implementing organization) between 2008 and 2010; served 800 to 1000 users; and required community members help maintain the system. Each case was a peri-urban slum resettlement in southern India where local municipalities were responsible for infrastructure and service provision and most residents were from India’s lowest caste, were employed as day laborers, and had low incomes (5,000 rupees

($70)/household/month, on average).

Data Collection We collected extensive empirical evidence from interviews, documentation, and sanitation system evaluations and observations to thoroughly understand the causal mechanisms for each sanitation system’s outcome, to generate case knowledge (Figure 4.1). We collected data from

June to August 2016 and from January to May 2017.

Interviews. Semi-structured interviews were conducted with community members, community leaders, sanitation system operators (a male operator or women’s self-help group

(WSHG) members), implementing organizations, and local municipalities. Interviews explored system use, maintenance, and history; technology selection; and stakeholder roles in planning and maintenance. Examples of interview questions included: Can you describe how the sanitation system was planned? and What support does your organization provide to the community

81 regarding the sanitation system? In total, 507 participants were interviewed (SI Table G2).

Interviews with community members aimed to achieve balanced gender, age, and geographical representation. Participants were selected using door-to-door sampling in the morning, midday, and evening to capture perspectives from domestic, employed, and unemployed individuals.

Interviews concluded in each case when theoretical saturation was reached (i.e., when no new themes or topics were mentioned in a subsequent interview). Interviews with community members, leaders, and sanitation system operators were conducted using local translators who had experience with sanitation fieldwork and were trained to follow the Institutional Review Board-approved protocol (#16-0026). Interviews with implementing organizations and municipalities were conducted in English.

Documentation. We collected documentation from implementing organizations and municipalities that included: standard operating procedures describing planning and implementation strategies; feasibility studies describing decision-making and project goals; planning meeting notes summarizing stakeholder roles in technology selection, construction, and maintenance; detailed project reports describing final system designs, cost information, and material quantities and quality; and sanitation system monitoring and evaluation reports describing historical performance data, system damage, number and type of maintenance tasks performed, user fees collected (where applicable), resource recovery profits, and ongoing technical and financial assistance from implementing organizations or municipalities.

Sanitation System Evaluations and Observations. Sanitation system performance was evaluated using the three regulated parameters for domestic wastewater in India: chemical oxygen demand (COD), biochemical oxygen demand (BOD), and pH (SI Table G3) (Central Pollution

Control Board 2017). Influent and effluent wastewater samples were taken from fourteen of the

82 twenty cases; for the remaining six cases, effluent samples could not be taken because of system damage or lack of wastewater. We also evaluated construction quality based on damage, design errors, and material type. To estimate the percentage of community members who were using the sanitation system, we triangulated data on frequency of use from interviews and researcher observations of the number of individuals using the toilets for two hours in the morning and evening on two separate days, the amount of open defecation at community-reported open defecation sites, the wastewater levels in the treatment tanks, and the cleanliness of at least one- third of a case’s toilets, selected randomly, to help evaluate proper use or misuse.

Data Analysis Interview transcriptions, observation notes, and documentation were uploaded into QSR

NVivo, a qualitative coding software (QSR International 2015). Qualitative data were coded using both deductive and inductive methods (Saldana 2009). In deductive coding, researchers use theory to hypothesize important themes related to system success or failure. For example, the theme

Community Participation in Planning was identified prior to the start of coding because literature states that community involvement may increase willingness to use and maintain the sanitation system (Chatterley et al. 2013; Eales et al. 2013; Marks et al. 2014). In inductive coding, researchers remain open to new themes related to success or failure that emerge from the case knowledge. For example, the theme Formal Sanitation System Handover was identified during coding because participants described how handover was important to reinforce O&M responsibility. To ensure internal validity, a coding dictionary was developed iteratively between two coders. Inter-coder agreement was measured using Cohen’s Kappa coefficient (Bazeley and

Jackson 2013); the final coefficient was 0.59, which reflects acceptable coding agreement.

Conflicting statements between participants were resolved by triangulating answers with documentation and observations (Basurto and Speer 2012).

83 Fuzzy-set Qualitative Comparative Analysis A variant of QCA that uses fuzzy logic is fsQCA (Ragin 2008), which is useful when cases do not dichotomously fall fully in or fully out of a set, such as when implementing organizations involve community members in planning to varying extents. In fsQCA, fuzzy sets for the causal conditions and outcomes were defined through an iterative process called calibration, which ensures that fuzzy set definitions provide a consistent measure for meaningful differences between cases (Basurto and Speer 2012).

Domain and Causal Conditions Identification. Causal conditions are factors hypothesized to influence an outcome, identified from theory or case knowledge. In total, we analyzed nine causal conditions for all 20 cases (Table 4.1). First, we identified an initial list of causal conditions from literature (SI Table G4). For example, Eales et al. (2013) asserts that sanitation systems with Technical Support are more likely to meet regulations, based on case studies from decentralized sanitation systems in Indonesia. Second, we identified which of those causal conditions were domain conditions, which are conditions that do not vary across the cases and are therefore removed from the analysis (Ragin, 2008). The domain conditions included: regulations (Hawkins et al., 2013), sanitation system age (Sabogal et al., 2014), system size

(Brikké, 2000), technology complexity (Brikké, 2000), community socio-economic status

(Mwirigi et al., 2009), culture (Mwirigi et al., 2009), capital costs (Eales et al., 2013), community financial contributions to capital costs (Marks and Davis, 2012), and community participation in construction activities (Roma and Jeffrey, 2010). Third, we identified additional causal conditions from case knowledge. For example, interviews uncovered Government Barriers because some municipalities had taken deliberate actions to disrupt a sanitation system. Finally, the full list of causal conditions was evaluated to remove conditions if they had low necessity scores (i.e., less than 0.3, a conventional cutoff for condition inclusion (Opdyke et al., 2018)), and case knowledge

84 indicated that the condition was not an important driver of success or failure; if they were too similar to another condition (i.e., those conditions were combined into one); and if they were found to be least-important during the QCA minimization process (discussed below).

Table 4.1. Causal conditions hypothesized to influence sanitation system success and failure. Causal Conditions* Definition Source The sanitation system is an appropriate Black, 1998; Hacker and technology that addresses a majority of Kaminsky, 2017; Murphy et Addressed Sanitation the community’s (most important) al., 2009; Palaniappan et al., Priorities sanitation priorities; quantified using 2008; Seymour, 2014; Case the priority addressment protocol Knowledge (Davis et al., 2019). Behavior change theory is used to Mosler, 2012; Rosenquist, teach community members the benefits Behavior Change Education 2005; Wegelin-Schuringa, of sanitation and to reduce open 2000; Case Knowledge defecation. All required maintenance tasks are Brikké and Bredero, 2003; known, and all stakeholders agree on Clear O&M Plan Chatterley et al., 2014; Case whose responsibility it is to perform Knowledge and finance each task. Community members are regularly and Battacharyya, 2015; Black, meaningfully involved in planning, 1998; Bouabid and Louis, Community Participation in which includes attending meetings and 2015; Mansuri, 2004; Roma Planning helping to make decisions such as site and Jeffrey, 2010; selection and appropriate technology Palaniappan et al., 2008; Case selection. Knowledge The sanitation system is well- Chatterley et al., 2014, 2013; Construction Quality constructed based on high material Case Knowledge quality and correct implementation. The local municipality has not taken Absence of Government deliberate actions that prevent or Case Knowledge Barriers disrupt sanitation system use, maintenance, or performance. The local municipality is regularly and meaningfully involved in planning, Bouabid and Louis, 2015; Municipality Involved in which includes attending meetings and Harris et al., 2011; Kooy and Planning helping to make decisions such as site Harris, 2012; Sansom, 2011; selection and appropriate technology Case Knowledge selection. Funds are available from user fees or Bouabid and Louis, 2015; implementing organization or Sufficient Funds for O&M Eales et al., 2013; Starkl et municipality assistance equal to or in al., 2013; Case Knowledge excess of the system’s O&M costs. Bouabid and Louis, 2015; Adequate technical capacity for Chatterley et al., 2014; Eales maintenance is available through a et al., 2013; IDECK, 2015; Technical Support skilled operator and external Kooy and Harris, 2012; maintenance assistance. Sakthivel et al., 2014; Tilley et al., 2014; Case Knowledge *The presence of causal conditions is hypothesized for success, and the absence is hypothesized for failure.

85 Causal Condition Calibration. The complete calibration guide for all causal conditions is included in the SI (Table G5, Figures G1 and G2). Seven of the nine causal conditions had mostly qualitative data, so we calibrated these conditions indirectly (i.e., set membership is defined qualitatively, based on case knowledge and theory (Basurto and Speer, 2012)). First, we defined the anchor points for in-set membership (fuzzy set score of 1), out-of-set membership (fuzzy set score of 0), and the crossover point (fuzzy set score of 0.5) for each causal condition based on theory. Next, we adjusted these definitions until meaningful differences between the 20 cases were accurately reflected by the calibrations. For example, out-of-set membership for Community

Participation in Planning was when community members were entirely uninvolved in planning and learned of the project only after construction began; we added “community members attended exposure visits” to the in-set membership definition because case knowledge indicated that exposure visits (i.e., where nearby successful sanitation systems were visited to learn about technology options and O&M needs) differentiated cases with in-set membership from cases with partial membership. The remaining two of the nine causal conditions were calibrated directly (i.e., set membership is defined by continuously normalizing raw quantitative data within anchor points

(0, 0.5, 1) (Ragin, 2008)). For Addressed Sanitation Priorities, raw data for each case was a quantitative priority addressment score, which reflects the extent to which priorities were addressed based on importance (i.e., how appropriate the technology was to the local context)

(Davis et al., 2019) (SI Figure G1). For Sufficient O&M Funds, raw data for each case was the amount of available funds as a percentage of monthly O&M costs (SI Figure G2).

Outcome Classification and Calibration. Each case’s sanitation system was classified as being either successful or failed. Success was defined as the presence of three criteria (Davis et al.,

2019, 2018): (1) the system is used by at least 75% of the community; (2) at least 90% of

86 maintenance tasks are performed correctly and on time; and (3) the system complies with local regulations for pH, COD, and BOD. Cases were classified as failed if they did not meet at least one of the success criteria. Use was directly calibrated (SI Figure G3). In-set membership was defined as more than 75% of the system’s target population using the system correctly, daily, and exclusively (i.e., no open defecation) (Andres et al., 2014; Harris et al., 2017) while out-of-set membership was defined as less than 25% using it correctly, daily, and exclusively. Maintenance was also directly calibrated (SI Figure G4). In-set membership was defined as at least 90% of the total required maintenance tasks were completed correctly and on time (Brikké, 2000; Eales et al.,

2013) while out-of-set membership was defined as less than 25% completed correctly and on time.

Performance was indirectly calibrated using a three-value fuzzy set (SI Table G6). In-set membership was defined as complying with all applicable pH, BOD, and COD regulations while out-of-set membership was defined as failing to comply with all three regulations; an intermediate value of 0.3 was defined as a system failing to comply with only one regulation. For the fsQCA, success outcome scores were determined by taking the minimum of the fuzzy set scores for use, maintenance, and performance (SI Table G7). The outcome of failure was analyzed using the negated (i.e., absence) of the success outcome scores.

Pathway Identification and Interpretation. Fuzzy set scores for all conditions and outcomes were assigned for every case and summarized in a QCA truth table (Table 4.2). We used the software fs/QCA (Ragin 2013) to minimize the truth table and to calculate, using Boolean algebra and fuzzy logic (Ragin 2008), pathways to success and to failure. Minimization was performed by comparing all possible combinations of causal conditions in a stepwise process to remove least-important causal conditions and identify the simplest combinations of causal conditions needed to produce an outcome. Each pathway is a combination of causal conditions

87 that results in an outcome. To interpret each pathway’s validity, we used four main QCA metrics.

Consistency evaluates each pathway’s reliability; it is the fraction of cases that exhibit the same pathway and outcome, and fractions above 0.8 are required for a pathway to be “consistent” (Ragin

2006). Necessity evaluates how commonly a causal condition is present with an outcome; it is calculated using the same process as consistency, and fractions above 0.9 are required for a causal condition to be “necessary” (Ragin 2008). Coverage helps evaluate the generalizability of findings; of cases with the same outcome, it is the fraction explained by the same pathway (Rihoux and

Ragin 2009) such that higher coverage indicates that that pathway explains more cases. Sufficiency evaluates how commonly a causal condition results in a positive outcome; it is calculated the same way as coverage, and fractions above 0.8 are required for a causal condition to be “sufficient”

(Ragin 2008).

88 Table 4.2. The truth table summarizes the fuzzy scores for each causal condition and the outcome for all 20 cases.

Causal Conditions Outcomes

Case Case

Use

Plan

Success)

Failure Failure

Quality Success

Change Change

Support

Barriers

Planning

Behavior Behavior

Priorities

Sufficient Sufficient

Technical Technical [min(Use,

Sanitation Sanitation Education

Addressed Addressed

Involved in Involved

in Planning

Community Community

Clear O&M Clear O&M

Government Government

O&M Funds O&M

Municipality Municipality

Maintenance Performance

Participation Participation Construction

Maintenance, Maintenance, Performance]

1 0.22 0.67 1 0.33 1 0 1 1 0.67 0.96 1 1 0.96 0.04 2 0.82 0.33 0.67 0.67 1 0 0.33 1 1 0.92 1 1 0.92 0.08 3 0.62 0.67 0.67 0.33 0.67 0 1 0.88 0.67 0.92 1 0.98 0.92 0.08 4 0.03 0.33 0 0.67 0.33 0 0 0.51 0 0 0 0 0 1 5 0.04 0 0 0 0.33 0 0.33 0.27 0 0 0.3 0.01 0 1 6 0.11 1 0.33 0.33 0.67 0 0 0.21 0.33 0.07 0 0.43 0 1 7 0.96 0.67 1 0.33 1 0 1 1 0.67 1 1 1 1 0 8 0.7 0.67 1 0.33 0.67 0 1 1 1 1 1 0.99 0.99 0.01 9 0.24 1 0.67 0.67 0.67 1 0 0.07 0.33 0.14 0 0.98 0 1 10 0 0 0 0.67 0.33 0 0 0.88 0 0 0.3 0 0 1 11 0.97 0.67 1 0.67 1 0 0.67 1 1 1 1 1 1 0 12 0.33 0.33 0.33 0.33 0.67 0 0.33 0.32 0.33 0.07 0 0.06 0 1 13 0.96 1 1 1 1 0 0.33 0.99 1 1 1 1 1 0 14 0.98 0.67 1 0.67 1 0 0.33 1 0.67 1 1 1 1 0 15 0.66 0.67 1 0.33 0.67 0 1 1 0.67 0.92 1 0.99 0.92 0.08 16 0.01 0 0 0.67 0.33 0 0 0.5 0 0.05 0.3 0.06 0.05 0.95 17 0.02 0.33 0 0.33 0 0 0 0.5 0 0.14 0 0.99 0 1 18 0.98 1 1 1 1 0 0.67 1 1 1 1 1 1 0 19 0.03 0 0 0 0 0 0 0.43 0 0.5 0 0.98 0 1 20 0.01 1 0.67 1 0.33 1 0 1 0 0 0 0 0 1 Note: Scores greater than 0.5 indicate membership in the set for the condition or outcome; scores less than 0.5 indicate non-membership in the set for the condition or outcome. Success outcome scores greater than 0.5 are considered to be successful cases; success outcome scores less than 0.5 are considered to be failed cases

89 Results Two pathways to success were identified (Figure 4.2a) and demonstrate alternative ways to achieve the same outcome of success. Together, they described all ten successful cases. Each success pathway had five causal conditions, of which three were shared: Sufficient O&M Funds,

Clear O&M Plan, and Technical Support. The first success pathway, which described five of the ten successful cases, also had Addressed Sanitation Priorities and Community Participation in

Planning causal conditions. The second success pathway, which described seven of the ten successful cases, also had Behavior Change Education and Municipality Involved in Planning causal conditions. Two cases, Cases 13 and 18, had all eight causal conditions present and were thus described by both pathways. Both success pathways highlight the importance of leveraging local knowledge to incentivize system buy-in and of establishing adequate resources and clear responsibilities for maintenance.

Four pathways led to failure (Figure 4.2b). All four, which together described all ten failed cases, shared three common causal conditions: Lack of Municipality in Planning, No Technical

Support, and Unaddressed Sanitation Priorities. The first failure pathway described six of the ten failed cases and also included No Clear O&M Plan and Poor Construction Quality; the second described four of the ten failed cases and also included No Clear O&M Plan, Lack of Community

Participation in Planning, and Insufficient O&M Funds; the third described one unique case (Case

20) and also included Government Barriers and Poor Construction Quality; the fourth described another unique case (Case 9) and also included Government Barriers and Insufficient O&M Funds.

Cases 5 and 19 were described by both the first and second failure pathways. Overall, the failure pathways each highlight that failed systems were unable to overcome inadequate maintenance resources, especially when municipalities and communities were uninvolved in planning.

90 While the high consistency and coverage of the success and failure pathways highlight the strength of the results, the number of cases and the focused context limit our ability to evaluate and consider all possible combinations of causal conditions that could influence sanitation outcomes. Additionally, pathways represent the combinations of conditions that together were sufficient to produce the outcome of success or failure. Causal conditions in a pathway are presented in order of decreasing necessity scores; the order is not chronological. Finally, the results demonstrate alternative combinations of conditions. For a given outcome, each pathway was sufficient to produce that outcome; one is not necessarily better than the either.

Figure 4.2. Results of the combinations of conditions that led to (a) success and (b) failure. Bold conditions have necessity scores greater than or equal to 0.90. Bold case numbers indicate cases that are uniquely explained by one combination of conditions; Underlined case numbers indicate cases that are explained by more than one combination of conditions. The causal conditions in the pathways are not chronological or linear; causal conditions are presented based on necessity scores, with the exception of Government Barriers in failure pathways 3 and 4, which is presented first to differentiate these two pathways. Case numbering matches cases from Davis et al., 2019, 2018.

92 Discussion The Importance of Operation & Maintenance for Success Adequate managerial, technical, and financial resources for maintenance were imperative for success, while their absence was a major driver of failure. All successful cases had a Clear

O&M Plan, where technical and financial responsibility for each well-specified maintenance task was assigned and agreed upon by the implementing organization and community, and eight of the ten failed cases lacked an O&M plan. Implementing organizations should focus their efforts on reducing O&M vulnerabilities by creating clear and comprehensive O&M plans, bolstering operator skills and reliability, and securing ongoing financial, technical, and managerial support for the duration of the systems’ lifetimes.

Clear responsibility meant that all successful cases also had Technical Support, where technical maintenance assistance from the implementing organization or municipality and a skilled operator were both regularly present. Skilled operators were essential for daily maintenance and recognizing larger system issues that would require technical assistance (e.g., pump failure). Two systems were operated by well-organized community WSHGs; six had male operators from the community; two were operated by male employees from the municipality or implementing organization. In the successful cases, the WSHGs and male operators from the communities did not previously have experience as operators but received effective training that provided them the knowledge and skills to diagnose problems, perform corrective action, and complete regular maintenance tasks correctly. The male operators from the implementing organizations or municipalities were professional operators with formal education in sanitation (e.g., sanitary engineering degrees). Technical maintenance assistance was important to train new operators and assist with maintenance during operator changes and transitions. Overall, technical maintenance assistance provided accountability and redundancy to community maintenance efforts, which

93 aligns with existing literature that found that communities are often unable to adequately perform maintenance independently (Bouabid and Louis 2015; Chowns 2015; Marks et al. 2014). Since sanitation system size and the number of required O&M tasks were similar across all 20 cases, both conditions are domain conditions and their influence on success or failure could not be analyzed (Ragin 2008).

When technical support was not provided, systems failed. All ten failed cases had No

Technical Support, where no technical maintenance assistance was provided by the implementing organization or the municipality, and operators did not exist (four cases) or were unskilled (four cases). As a result, communities were not prepared to take responsibility for the sanitation system and unable to perform essential maintenance like desludging, subsequently leading to poor system performance. Additionally, eight failed cases had No Clear O&M Plan (failure pathways 1 and 2,

Figure 4.2b), where technical and financial responsibilities were not clearly assigned, and communities or municipalities claimed that maintenance was not their responsibility.

All successful cases also had Sufficient O&M Funds, where O&M funds came from user fees and/or funds provided by the implementing organization or municipality. Five cases (Cases

7, 11, 13, 15, and 18) generated equal to or more than 100% of O&M costs; four cases (Cases 2,

3, 8, and 14) generated an average of 67% of O&M costs; one case (Case 1) did not generate income. Income was generated primarily through user fees, and in four of the ten successful cases, additional O&M income was generated by selling resources recovered from the sanitation system: vegetables grown using recycled water (Case 11), compost sales (Cases 13 and 14), and biogas for cooking fuel (Cases 13 and 18). In these four cases, this income provided additional benefits beyond sanitation. For example, excess system income was lent as micro-loans to women (Cases

11 and 18) or used to improve other community infrastructure (e.g., community hall (Case 18)).

94 These non-sanitation benefits may have contributed to these cases’ strong commitment to using and maintaining the systems. All successful cases also had external financial assistance, where the implementing organization (Case 2) or the municipality (Cases 1, 3, 7, 8, 11, 13, 14, 15, and 18) agreed to and consistently paid for large O&M costs (e.g., desludging) and/or covered all costs that were not covered by system income generation. Case knowledge indicates that successful systems still required this financial redundancy and external assistance (i.e., subsidies) particularly when large, unexpected O&M costs arose or when community members irregularly paid user fees.

Literature supports the finding that adequate O&M funds are essential for success and are particularly important to be secured prior to system implementation (Bouabid and Louis 2015;

Eales et al. 2013; Palaniappan et al. 2008; Starkl et al. 2013b). While these systems were successful, there is still progress to be made towards sustainability. Additionally, system capital costs from all 20 cases were almost entirely subsidized (i.e., negligible community contributions) by the implementing organization (ten cases), municipality (six cases), or by both (four cases); since capital costs and community contributions did not vary across the 20 cases, the effects of these conditions were not analyzed.

In contrast, five of the ten failed cases had Insufficient O&M Funds (failure pathways 2 and 4), where income generation and financial assistance were less than 50% of O&M costs. As a result, operators were un-paid or under-paid, and maintenance tasks that significantly impacted performance (e.g., desludging) were not done. The lack of funds were because (i) community members did not pay user fees, because they could not afford to, did not see the benefit of sanitation, or were unsatisfied (e.g., “why should I pay to use the toilet when the toilet is never clean?”); (ii) there were no funds provided by the implementing organization because they did not intend to support maintenance long-term; or (iii) there were no funds provided by the municipality,

95 because the implementing organizations did not include the municipality in planning or the municipality did not have the resources to pay for sanitation. Four failed systems (Cases 4, 10, 16, and 17) generated 71% of O&M costs on average but received no external financial assistance and were therefore unable to pay for costly maintenance (e.g., desludging). One failed system (Case

20) generated over 100% of O&M costs, but Government Barriers ultimately disrupted maintenance (described further below). Overall, implementing organizations must ensure that adequate funds are available for O&M long-term, especially to incentivize operator retention.

The Importance of Local Stakeholder Engagement and Community Buy-In for Success Local Stakeholder Engagement. Engagement of local stakeholders was important in both pathways to leverage local knowledge for appropriate technology selection and define maintenance responsibilities. For successful cases, implementing organizations either involved the community or the municipality in planning; two successful cases (Cases 13, 18) had both involved.

Local stakeholder engagement also better supported long-term relationships with the implementing organizations, which helped ensure ongoing technical and financial assistance. The success pathways also demonstrate that local engagement needed to be combined with Addressed

Sanitation Priorities if the community was engaged or with Behavior Change Education if the municipality was engaged to help achieve community acceptance and willingness to use and maintain the system. All failed cases lacked the engagement of local stakeholders, which meant that implementing organizations were unfamiliar with local norms and sanitation priorities and thus implemented inappropriate systems or ineffective management strategies. Local stakeholder engagement is considered essential for community buy-in (Roma and Jeffrey 2010), appropriate technology selection (Palaniappan et al. 2008), and ongoing maintenance (Battacharyya 2015).

One particularly effective approach of local engagement was Community Participation in

Planning. In the first success pathway, implementing organizations recognized the importance of

96 engaging the community in planning, especially when the municipality did not have the resources to participate in planning. This helped to create a strong O&M plan and encourage system use and maintenance. Community Participation in Planning positively influenced success when community members had moderate to high citizen power, meaning they were well-informed of the project, attended meetings, co-identified project goals, and provided input for how the sanitation system should be designed and managed. Case knowledge further indicates that Community

Participation in Planning in all cases in the first success pathway increased community buy-in and the likelihood that priorities were incorporated into appropriate technology selection and design.

In contrast, Lack of Community Participation in Planning in the second failure pathway meant community members were uninformed of the sanitation project and were not bought into the goals of sanitation system and that the sanitation technologies may have been inappropriate for the local context. A common theme expressed by these failed cases, as summarized by one community member, was: “the first time we learned about the sanitation system was when the contractor came and started building.” Implementing organizations that did not engage the community typically thought that community members lacked the technical skills to be involved in the planning of a sanitation system. This exacerbated unclear O&M responsibilities, left the communities unprepared to independently perform maintenance, and meant that sanitation priorities were unknown and therefore unaddressed.

However, Lack of Community Participation in Planning did not always contribute to failure if the municipality was involved. Five successful cases (1, 3, 7, 8, and 15) in the second success pathway had limited community participation in planning where communities were only informed of the project and had no decision-making input, but the municipality was involved and dedicated to the sanitation system’s success. Implementing organizations recognized the

97 importance of Municipality Involved in Planning because a national policy in India places the responsibility for the provision and management of sanitation infrastructure on the local municipality (Ministry of Urban Development 2008). Thus, implementing organizations made significant efforts to accommodate the limited time and resources of municipalities. One implementing organization’s engineer stated, “We know that government is important. We must do more than just ask for permission for the project to be successful. We must ask them for help in the planning.” Therefore, the municipalities in the second success pathway attended planning meetings, helped define maintenance responsibilities, made financial contributions for capital costs, provided oversight during construction, assisted with training of operators, and committed to providing financial and technical maintenance assistance.

In all failed cases, there was a Lack of Municipality in Planning. In three cases, the municipality was not informed of the project and thus was not given a chance to be involved in planning. In seven cases, the municipality was asked to participate in planning but was uninterested, unable, or unwilling to be involved. Community members and implementing organizations still expected the municipality to support maintenance despite the municipality never agreeing to do so, and as a result, failed cases struggled with uncertain O&M responsibilities. For example, in Case 16, formal handover to the municipality did not occur, so the municipality stated,

“We cannot interfere with a project until it no longer belongs to the [implementing organization].”

Since most implementing organizations did not intend to provide long-term technical or financial assistance and because municipalities also did not commit these resources, all failed cases had No

Technical Support, eight had No Clear O&M Plan, seven had Poor Quality Construction, five had

Insufficient O&M Funds, and two had Government Barriers. Engaging the municipality in the early planning stages could have potentially mitigated these challenges. Implementing

98 organizations need to prioritize engagement of municipalities, and municipalities need to budget time and resources to provide technical and financial assistance for sanitation.

Community Buy-In. Beyond local engagement, efforts to incentivize community buy-in were also important in both success pathways. In the first success pathway, community members cited Addressed Sanitation Priorities as an important reason why they continued to use and maintain the sanitation system. In these five successful cases, community members were more involved in decision-making and implementing organizations were familiar with the communities, so more priorities were addressed by the sanitation systems. In successful Cases 11 and 18, implementing organizations also conducted extensive priority assessments that focused on identifying sanitation-specific priorities, which informed technology selection and design; as a result, priorities were extremely well-addressed in these two cases. In successful Cases 2, 13, and

14, implementing organizations did not conduct priority assessments, but priorities were still well- addressed because those organizations were already familiar with the communities (through previous projects such as water supply). In contrast, all ten failed systems had Unaddressed

Sanitation Priorities, likely due to a combination of the lack of local stakeholder engagement, the lack of priority assessments, and poor-quality construction (i.e., implemented systems did not match intended designs). Case 20 was the only failed case where the implementing organization identified sanitation priorities; all other implementing organizations for failed cases either did not conduct a priority assessment or only identified priorities unspecific to sanitation. In all ten failed cases, priorities did not influence decision-making and system design, so community buy-in did not occur, suggesting that technologies may not have been appropriate to the local context

(Bouabid and Louis 2015; Palaniappan et al. 2008).

99 However, unaddressed priorities did not always lead to failure; community buy-in could be achieved alternatively through Behavior Change Education. In the second success pathway, community members did not participate meaningfully in planning and their sanitation priorities were not well-addressed, but Behavior Change Education was used to convince community members of the value of sanitation. Through educational activities that communicated sanitation health benefits such as seminars, street plays, community mapping to identify open defecation sites, or games, community members accepted and used a system that still left some of their priorities unaddressed. For example, priorities were poorly-addressed in successful Case 1, and the community preferred decentralized, individual septic tanks with toilets placed farther from their kitchens, but community members often stated, “We still use the toilets because we know it keeps us from getting sick”. Beyond behavior change and addressed priorities, community participation in construction has been theorized to influence buy-in (Roma and Jeffrey 2010); however, all 20 cases had minimal or no participation in construction, so this condition was not explored in the analysis. Together, the two success pathways demonstrate alternative ways to engage local stakeholders and encourage community buy-in to lead to success, while the failure pathways highlight negative consequences of inadequate stakeholder engagement and buy-in.

Unique Pathways to Failure: Cases 9 and 20 Two failed cases (9 and 20) were each described by a unique pathway, the third and fourth failure pathways, which shared the three common failure conditions of Lack of Municipality in

Planning, Unaddressed Sanitation Priorities, and No Technical Support. These two pathways diverged from the other two failure pathways because both had Government Barriers, which included deliberate actions from the local municipality that permanently disrupted a sanitation system’s use, maintenance, and performance.

100 Case 20, described by failure pathway 3, had strong potential to be successful. Initially,

Case 20’s sanitation system generated over three times the amount of O&M costs from user fees along with sales of recycled water for use in construction, biogas that was used to run a tea shop, and biogas that was used to heat bathing water. After four years of operation, the municipality revoked the land lease and took control of the system to gain access to the income generation. The municipality did not allow the WSHG to continue to manage the system and instead neglected it and eventually locked the toilets. Since the WSHG has struggled to regain ownership of the system, this Government Barrier resulted in permanent system failure (i.e., closure).

Case 9, described by failure pathway 4, also had the potential to be successful. The WSHG managing the system occasionally struggled to perform major maintenance, like desludging or repairs from weather damage due to No Technical Support and Insufficient O&M Funds but performed enough maintenance to keep the system functional. However, three years after implementation, the municipality resettled another community nearby and overloaded the system by connecting 120 new household toilets to the original sanitation system because the municipality was uninvolved in planning and did not understand the system’s intended design. Case 9 did not have sufficient funds to rectify the resulting issues of tank and sewer damage. Also, the municipality was unwilling to repair or expand the system. So, the Government Barriers contributed to permanent failure. While difficult to anticipate, these disruptions could be mitigated if implementing organizations engage municipalities in planning and if communities and implementing organizations build strong, long-term relationships with municipalities.

A Comparison of Top-Down and Bottom-Up Implementation Approaches Comparing the two success pathways shows that both top-down (organization- and municipality-driven) and bottom-up (community-driven) approaches were able to achieve success.

The first success pathway had cases where the implementing organizations were able to employ a

101 bottom-up approach. In these cases, implementing organizations engaged the community in planning and decision-making and addressed local priorities in the design and implementation of the sanitation system. Notably, bottom-up strategies were only effective as long as communities were not expected to maintain sanitation systems without any technical or financial assistance.

Despite a strong call in the sanitation sector to move away from top-down implementation strategies (Breslin 2003; Gabe et al. 2009; Hubbard et al. 2011), most implementing organizations relied on top-down strategies (five successful and eight failed cases). The second success pathway demonstrates that systems implemented with top-down approaches can still achieve success even if there is a lack of community participation and priority assessments. Local engagement and community buy-in were still essential but were achieved with alternative strategies of engaging the municipality and incentivizing community use through extensive behavior change education efforts. In successful top-down approaches, municipalities took ownership of sanitation service delivery and prioritized long-term technical and financial assistance, demonstrating that municipalities are important implementation partners. Top-down strategies resulted in failure

(eight cases) when behavior change education was absent and neither communities nor municipalities were engaged.

Finally, successful Cases 13 and 18 employed a combination of top-down and bottom-up strategies in planning. Each implementing organization initiated a sanitation project and each community was actively engaged in determining the sanitation system’s technology and design/ability (resource recovery) that would best address their priorities. The communities also had strong relationships with the municipalities, so community members played an active role in engaging the local municipality. This cooperation provided redundant technical and financial resources, which both supported the sanitation system and had additional benefits. For example,

102 both cases generated the most income from sales of resource recovery products, and because the municipalities financially supported O&M, the communities used the extra income to improve solid waste management (Case 13) and water supply (Case 18). In turn, these extra benefits further strengthened community buy-in and the municipalities’ understanding of the importance of sanitation. Overall, regardless of the implementation approach, implementing organizations should aim to increase community acceptance and satisfaction for sanitation systems by improving the quality of service delivered, addressing local priorities, and ensuring systems receive maintenance support.

Conclusion Two pathways led to successful sanitation systems, and four pathways led to failed sanitation systems. All successful systems required Sufficient O&M Funds, Clear O&M Plan, and

Technical Support in addition to either Addressed Sanitation Priorities and Community

Participation in Planning or Behavior Change Education and Municipality Involved in Planning.

Overall, the pathways to success demonstrate the importance of involving municipalities and communities in all project phases to ensure that appropriate technologies are selected that match the local context. Either pathway can lead to success, therefore implementing organizations should focus on the pathway that best aligns with available resources, expertise, and context-specific needs. Since adequate O&M funds, clear O&M plans, and technical support were in both success pathways, we recommend that implementing organizations should prioritize creating a clear O&M plan that identifies reliable and trained operators, establishes long-term technical support, and secures sufficient local O&M funds to ensure long-term use and maintenance. Failed systems lacked many of these important conditions for success, and these results demonstrate that there are many complex causes of sanitation failure. All failed systems had Lack of Municipality in

Planning, Unaddressed Sanitation Priorities, and No Technical Support. Most failed systems also

103 had No Clear O&M Plan, Poor Construction Quality, Lack of Community Participation in

Planning, and Insufficient O&M Funds. Two failed cases had unique pathways because

Government Barriers permanently disrupted use and maintenance. The pathways to failure emphasize the importance of engaging local stakeholders—especially the local municipality, as well as establishing clear mechanisms for ongoing technical, managerial, and financial support for resource-limited communities.

Acknowledgements We thank our research assistants, Vijay Kumar and Sridhar Selvaraj, and all participants in this research for their indispensable time and support.

Funding This work was completed with financial support from the Mortenson Center in Engineering for Developing Communities at the University of Colorado Boulder.

Conflicts of Interest The authors have no conflicts of interest.

Chapter 4 References Andres, L. A., Briceno, B., Chase, C., and Echenique, J. A. (2014). Sanitation and Externalities: Evidence from Early Childhood Health in Rural India. Policy Research Working Paper, The World Bank, South Asia Region. Banerjee, A. (2013). Legal and policy framework for wastewater treatment and reuse in India: A background review. GIZ-India. Bao, P. N., Aramaki, T., and Hanaki, K. (2013). “Assessment of stakeholders’ preferences towards sustainable sanitation scenarios: Assessment on sanitation scenarios.” Water and Environment Journal, 27(1), 58–70. Basurto, X., and Speer, J. (2012). “Structuring the Calibration of Qualitative Data as Sets for Qualitative Comparative Analysis (QCA).” Field Methods, 24(2), 155–174. Battacharyya, U. (2015). “Community participation in water and sanitation services: role of Panchayati Raj in India.” Water, sanitation and hygiene services beyond 2015: Improving access and sustainability, Loughborough, England. Bazeley, P., and Jackson, K. (2013). Qualitative Data Analysis with NVivo. Sage Publications. Black, M. (1998). Learning What Works: A 20 Year Retrospective View on International Water and Sanitation Cooperation. UNDP - World Bank Water and Sanitation Program, Washington, D.C.

104 Bouabid, A., and Louis, G. E. (2015). “Capacity factor analysis for evaluating water and sanitation infrastructure choices for developing communities.” Journal of Environmental Management, 161, 335–343. Breslin, E. D. (2003). Demand response approach in practice: why sustainability remains elusive. WaterAid, Mozambique. Brikké, F. (2000). “Operation and Maintenance of rural water supply and sanitation systems.” A training package for managers and planners. IRC International Water and Sanitation Center, WHO. Brikké, F., and Bredero, M. (2003). Linking technology choice with operation and maintenance in the context of community water supply and sanitation. WHO and IRC WASH, Geneva, Switzerland. Central Pollution Control Board. (2017). “Environmental Standards: Water Quality Criteria.” Central Pollution Control Board, (May 17, 2016). Chaplin, S. E. (2011). “Indian cities, sanitation and the state: the politics of the failure to provide.” Environment and Urbanization, 23(1), 57–70. Chatterley, C., Javernick-Will, A., Linden, K. G., Alam, K., and Bottinelli, L. (2014). “A qualitative comparative analysis of well-managed school sanitation in Bangladesh.” BMC Public Health, 14(6), 1–14. Chatterley, C., Linden, K. G., and Javernick-Will, A. (2013). “Identifying pathways to continued maintenance of school sanitation in Belize.” Journal of Water, Sanitation and Hygiene for Development, 3(3), 411–422. Chowns, E. (2015). “Is Community Management an Efficient and Effective Model of Public Service Delivery? Lessons from the Rural Water Supply Sector in Malawi: Is Community Management Efficient and Effective?” Public Administration and Development, 35(4), 263–276. Cronin, A. A., Ohikata, M., and Kumar, M. (2014). “Social and economic cost-benefit analysis of sanitation in Odisha State, India.” Journal of Water, Sanitation and Hygiene for Development, 4(3), 521. Davis, A., Javernick-Will, A., and Cook, S. (2018). “A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems.” Environmental Science: Water Research & Technology, 4, 1451– 1463. Davis, A., Javernick-Will, A., and Cook, S. (2019). “Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities.” Environmental Science & Technology, 53(1), 401–411. Eales, K., Blackett, I., Siregar, R., and Febriani, E. (2013). Review of Community-Managed Decentralized Wastewater Treatment Systems in Indonesia. WSP. Fischer, M., and Maggetti, M. (2017). “Qualitative Comparative Analysis and the Study of Policy Processes.” Journal of Comparative Policy Analysis: Research and Practice, 19(4), 345– 361. Gabe, J., Trowsdale, S., and Vale, R. (2009). “Achieving integrated urban water management: planning top-down or bottom-up?” Water Science and Technology, 59(10), 1999–2008. Hacker, M. E., and Kaminsky, J. A. (2017). “Cultural preferences for the methods and motivation of sanitation infrastructure development.” Journal of Water Sanitation and Hygiene for Development, 7(3), 407–415.

105 Harris, D., Kooy, M., and Jones, L. (2011). Analysing the governance and political economy of water and sanitation service delivery. Overseas Development Institute, London, UK. Harris, M., Alzua, M. L., Osbert, N., and Pickering, A. (2017). “Community-Level Sanitation Coverage More Strongly Associated with Child Growth and Household Drinking Water Quality than Access to a Private Toilet in Rural Mali.” Environmental Science & Technology, 51(12), 7219–7227. Hawkins, P., Blackett, I., and Heymans, C. (2013). Poor-inclusive urban sanitation: An overview. Targeting the Urban Poor and Improving Services in Small Towns, WSP, The World Bank, Washington, D.C. Hubbard, B., Sarisky, J., Gelting, R., Baffigo, V., Seminario, R., and Centurion, C. (2011). “A community demand-driven approach toward sustainable water and sanitation infrastructure development.” International Journal of Hygiene and Environmental Health, 214(4), 326– 334. IDECK. (2015). Strengthening the operation & maintenance sector for servicing decentralised urban sanitation infrastructure in Bangalore: Landscape synthesis report. CDD Society and BORDA, Bangalore, India. Jordan, E., Javernick-Will, A., and Tierney, K. (2016). “Post-tsunami recovery in Tamil Nadu, India: combined social and infrastructural outcomes.” Natural Hazards, 84(2), 1327–1347. Kaminsky, J., and Javernick-Will, A. (2012). “Causes for Sustainable Maintenance and Operation of On-Site Sanitation Systems.” Proceedings of the 2012 Construction Research Congress, 21–23. Katukiza, A. Y., Ronteltap, M., Oleja, A., Niwagaba, C. B., Kansiime, F., and Lens, P. N. L. (2010). “Selection of sustainable sanitation technologies for urban slums — A case of Bwaise III in Kampala, Uganda.” Science of The Total Environment, 409(1), 52–62. Kooy, M., and Harris, D. (2012). Political economy analysis for water, sanitation and hygiene (WASH) service delivery. Project Briefing, Overseas Development Institute, London, UK. Kunz, N. C., Fischer, M., Ingold, K., and Hering, J. G. (2015). “Why Do Some Water Utilities Recycle More than Others? A Qualitative Comparative Analysis in New South Wales, Australia.” Environmental Science & Technology, 49(14), 8287–8296. Mansuri, G. (2004). “Community-Based and -Driven Development: A Critical Review.” The World Bank Research Observer, 19(1), 1–39. Marks, S. J., and Davis, J. (2012). “Does User Participation Lead to Sense of Ownership for Rural Water Systems? Evidence from Kenya.” World Development, 40(8), 1569–1576. Marks, S. J., Komives, K., and Davis, J. (2014). “Community participation and water supply sustainability evidence from handpump projects in rural Ghana.” Journal of Planning Education and Research, 34(3), 276–286. Marks, S. J., Kumpel, E., Guo, J., Bartram, J., and Davis, J. (2018). “Pathways to sustainability: A fuzzy-set qualitative comparative analysis of rural water supply programs.” Journal of Cleaner Production, 205, 789–798. Ministry of Urban Development. (2008). National Urban Sanitation Policy. Government of India, New Delhi, India. Mosler, H.-J. (2012). “A systematic approach to behavior change interventions for the water and sanitation sector in developing countries: a conceptual model, a review, and a guideline.” International Journal of Environmental Health Research, 22(5), 431–449.

106 Murphy, H. M., McBean, E. A., and Farahbakhsh, K. (2009). “Appropriate technology – A comprehensive approach for water and sanitation in the developing world.” Technology in Society, 31(2), 158–167. Mwirigi, J., Balana, B. B., Mugisha, J., Walekhwa, P., Melamu, R., Nakami, S., and Makenzi, P. (2014). “Socio-economic hurdles to widespread adoption of small-scale biogas digesters in Sub-Saharan Africa: A review.” Biomass and Bioenergy, 70, 17–25. Mwirigi, J. W., Makenzi, P. M., and Ochola, W. O. (2009). “Socio-economic constraints to adoption and sustainability of biogas technology by farmers in Nakuru Districts, Kenya.” Energy for Sustainable Development, 13(2), 106–115. Nair, S., and Howlett, M. (2015). “Scaling up of Policy Experiments and Pilots: A Qualitative Comparative Analysis and Lessons for the Water Sector.” Water Resources Management, 29(14), 4945–4961. Opdyke, A., Javernick-Will, A., and Koschmann, M. (2018). “A Comparative Analysis of Coordination, Participation, and Training in Post-Disaster Shelter Projects.” Sustainability, 10(11), 4241. Palaniappan, M., Lang, M., and Gleick, P. H. (2008). A Review of Decsion-Making Support Tools in the Water, Sanitation, and Hygiene Sector. The Pacific Institute, Oakland, California. QSR International. (2015). NVivo qualitative data analysis software. QSR International Pty Ltd. Ragin, C. C. (1987). The comparative method: moving beyond qualitative and quantitative strategies. University of California Press, Berkeley, Calif. Ragin, C. C. (2006). “Set Relations in Social Research: Evaluating Their Consistency and Coverage.” Political Analysis, 14(3), 291–310. Ragin, C. C. (2008). Redesigning social inquiry: fuzzy sets and beyond. University of Chicago Press, Chicago. Ragin, C. C. (2013). Fuzzy-Set/Qualitative Comparative Analysis. Department of Sociology, University of Arizona, Tuscon Arizona. Rihoux, B., and Ragin, C. C. (2009). Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques. Sage Publications, Inc. Roma, E., and Jeffrey, P. (2010). “Evaluation of community participation in the implementation of community-based sanitation systems: a case study from Indonesia.” Water Science & Technology, 62(5), 1028–1036. Rosenquist, L. E. D. (2005). “A psychosocial analysis of the human-sanitation nexus.” Journal of Environmental Psychology, 25(3), 335–346. Sabogal, R. I., Medlin, E., Aquino, G., and Gelting, R. J. (2014). “Sustainability of water, sanitation and hygiene interventions in Central America.” Journal of Water, Sanitation and Hygiene for Development, 4(1), 89–99. Sakthivel, S. R., Seshadri, A., Rahman, A., and Chariar, V. M. (2014). “Standardisation of Design and Maintenance of DEWATS Plants in India.” India Institute of Technology Technical Paper Series. Saldana, J. (2009). The Coding Manual for Qualitative Researchers. Sage, London. Sansom, K. (2011). “Complementary roles? NGO-Government relations for community-based sanitation in South Asia.” Public Administration and Development, 31(4), 282–293. Seymour, Z. (2014). “Sanitation in developing countries: a systematic review of user preferences and motivations.” Journal of Water, Sanitation and Hygiene for Development, 4(4), 681– 691.

107 Starkl, M., Brunner, N., and Stenström, T.-A. (2013). “Why Do Water and Sanitation Systems for the Poor Still Fail? Policy Analysis in Economically Advanced Developing Countries.” Environmental Science & Technology, 47, 6102–6110. Sujaritpong, S., and Nitivattananon, V. (2009). “Factors influencing wastewater management performance: Case study of housing estates in suburban Bangkok, Thailand.” Journal of Environmental Management, 90(1), 455–465. Tilley, E., Strande, L., Lüthi, C., Mosler, H.-J., Udert, K. M., Gebauer, H., and Hering, J. G. (2014). “Looking beyond Technology: An Integrated Approach to Water, Sanitation and Hygiene in Low Income Countries.” Environmental Science & Technology, 48(17), 9965–9970. Törnqvist, R., Norström, A., Kärrman, E., and Malmqvist, P.-A. (2008). “A framework for planning of sustainable water and sanitation systems in peri-urban areas.” Water Science & Technology, 58(3), 563–570. WaterAid. (2011). Sustainability framework. Framework, WaterAid, London, UK. Wegelin-Schuringa, M. (2000). “Public awareness and mobilisation for ecosanitation.” IRC International Water and Sanitation Center, Delft, Netherlands, IRC International Water and Sanitation Center, Delft, Netherlands, Bonn, Germany. WHO, and UNICEF. (2017). Progress on Drinking Water, Sanitation and Hygiene: Update and SDG Baselines. Joint Monitoring Program, Geneva.

108 Chapter 5: Analyzing sanitation sustainability assessment frameworks to

identify ways to improve how sanitation sustainability is measured in

resource-limited communities

A final version of this chapter will be submitted to Environmental Science & Technology for publication as a journal article.

Introduction Sanitation system failure rates are unacceptably high, and more than 60 percent of the global population lacks access to safe sanitation (WHO and UNICEF 2017). To enhance implemented systems’ abilities to provide safe sanitation services and to improve the sustainability of future systems, researchers and practitioners have developed numerous sanitation sustainability assessment frameworks, especially for resource-limited contexts (Balkema et al. 2002; Barnes et al. 2011; Schweitzer et al. 2014). While there are a plethora of frameworks, to address continued failure, there is a need to measure sustainability consistently and uniformly (Balkema et al. 2002;

Dillard et al. 2008) and better understand the capabilities, benefits, and drawbacks of existing frameworks (Schwemlein et al. 2016). Thus, knowledge of the essential and most effective elements of sustainability assessment frameworks could better inform sanitation planning and improve global sanitation success and sustainability.

Frameworks increasingly evaluate and compare multiple dimensions of sustainability, aligning with common definitions of sustainable sanitation systems that minimize social, economic, and environmental impacts (Flores et al. 2009; Katukiza et al. 2012; Muga and Mihelcic

2008; Palme et al. 2005; WaterAid 2011a). But many frameworks focus specifically on only one element (i.e., pillar) of sustainability. For example, numerous frameworks use life cycle assessment principles (ISO 1997) to evaluate environmental impacts (Chevalier and Meunier

109 2005; Fang et al. 2016; Gallego et al. 2008; Guest et al. 2009a; Rodriguez-Garcia et al. 2011;

Schoen et al. 2017); others use life cycle costing methodology (Hunkeler and Rebitzer 2003) to quantify economic impacts (Ashley et al. 2008; Brunner et al. 2018; Molinos-Senante et al. 2010;

Rebitzer et al. 2003; Van Minh and Hung 2011); and others assess social impacts (Davis et al.

2019a; Noga and Wolbring 2012; Sperling et al. 2016).

In addition, frameworks often have unique objectives and employ different methods to analyze and compare indicators. Most frameworks are used for two main purposes. Monitoring frameworks either qualitatively evaluate and discuss impacts and tradeoffs (CDC 2008; Murphy et al. 2009; Starkl et al. 2013a; Tilley et al. 2014a) or track a single system’s change over time

(UNICEF 2012, 2017). Comparison frameworks assess impacts in relative terms by comparing multiple systems to each other or to a baseline and are often used for technology selection (Gaulke et al. 2010; Lennartsson et al. 2009; Olschewski and Casey 2015; Wongburi and Park 2018).

Existing frameworks support decisions and data processing in different ways, which usually are qualitative, semi-quantitative, or quantitative with a single recommendation. Frameworks that use qualitative methods typically discuss each indicator separately to identify strengths and challenges within one system. Frameworks that use semi-quantitative methods typically evaluate whether a system’s impacts were better (+), equal to (0), or worse (-) than a baseline. Frameworks that use quantitative methods typically employ multi-criteria decision analysis (MCDA) to aggregate quantitative data into a single-score per sanitation system (Iribarnegaray et al. 2012; Kalbar et al.

2012a; Malekpour et al. 2013; Molinos-Senante et al. 2014; Murray et al. 2009).

Each framework also presents its own unique set of indicators, which are measurable factors that may influence or reflect an aspect of sustainability. Of the hundreds of indicators, some are extremely difficult to measure and infeasible for implementing organizations in resource-

110 limited contexts (e.g., Number of Diarrheal Diseases Annually). Some are too context-specific

(e.g., Post-Flood Latrine Repairs) or too general (e.g., Export of Problems in Time and Space).

Also, even for a single indicator, there is usually a lack of a consistent measurement approach (e.g.,

System Performance measured with varied effluent water quality criteria across frameworks). Due to limited assessment and monitoring resources, combined with the large number of sanitation systems and technologies, it is infeasible to collect data on all indicators. Presently, it is not clear which indicators are the most useful, which indicators best differentiate between sustainable and unsustainable systems, or how to determine the most optimal combination of useful indicators within resource limitations.

To this end, a diverse set of existing sanitation sustainability assessment frameworks intended for resource-limited contexts were selected and evaluated to help determine what types of data should be collected (i.e., indicators), if and how frameworks could be adapted for different contexts, how results should be presented and compared, and how sanitation sustainability should be defined. As part of the evaluation, these frameworks were applied to existing successful and failed sanitation systems to determine how well frameworks differentiate systems with distinct outcomes. This framework evaluation, including the identification of their unique and common abilities and potential limitations of measuring sanitation sustainability, provides suggestions for framework improvements, that may ultimately help to increase sanitation sustainability in resource-limited communities.

111 Methods Framework Selection and Description Multiple existing sanitation sustainability assessment frameworks were identified from the literature and by consulting water, sanitation, and hygiene (WASH) practitioners. Of these, frameworks were selected for further analysis based on several criteria. Each had to: take a holistic evaluation approach and include, at a minimum, social, economic, and environmental indicators

(i.e., the three pillars); have an explicit sanitation focus in resource-limited communities; and clearly define its objective (e.g., monitor community sanitation systems). Since the goal was to gain insight into sanitation sustainability instead of attempting to comprehensively evaluate all available frameworks, a subset of the frameworks that met these criteria was chosen to represent the diversity of frameworks available, including different types of purpose, indicators, and data processing approaches (e.g., qualitative versus quantitative). Ultimately, six sanitation sustainability frameworks were selected that captured a range of author-affiliated organizations

(e.g., researchers, practitioners), varied complexity (i.e., number of indicators), diversity of indicators (e.g., range of different social indicators), different data processing methods (e.g.,

MCDA, qualitative), and different intended applications (e.g., planning, post-implementation)

(Table 5.).

112 Table 5.1. Summary of the six sanitation sustainability frameworks selected.

Intended No. of Indicators Framework Name Objective Unit(s) of Analysis Data Processing Method Reference Project Phase Indicators Categorization TechSelect 1.0: Quantitative indicator Technology To evaluate and rank Small-scale measurements (scale: 0 to Assessment for Economic, (Kalbar et al. sanitation technology sanitation systems 1) aggregated into single- Wastewater Planning 12 Environmental, 2012a; b, alternatives to select (especially high- score using multi-criteria Treatment Using Social 2016) appropriate technologies rises in India) decision analysis with Multiple-Attribute expert weighting Decision-Making To help municipalities Semi-quantitative Economic, Stockholm evaluate the relative indicator measurements Onsite, small-scale, Environmental, Environment Institute sustainability of technologies (scale: --, -, 0, +, ++) used (Lennartsson Planning and municipal 34 Social, (SEI) Sustainability compared to an existing, to compare system to the et al. 2009) sanitation systems Health, Criteria conventional sewered “0 alternative” (i.e., Technical treatment system baseline) Semi-quantitative (Olschewski To provide a decision indicator measurements and Casey support tool to evaluate the (scale: + = high value, 2015; Skat Economic, Skat Foundation and sustainable application of a Planning or neutral or positive, Foundation Environmental, WaterAid Technology potential or implemented Post- supportive characteristics; 2013; Skat WASH technologies 18 Social, Applicability technology, considering the Implement- 0 = potential impact, could Foundation Institutional, Framework (TAF) roles of and impacts on ation become critical, needs and Rural Technical beneficiaries, implementers, follow up; - = low value, Water Supply and governments negative, critical, Network hindering characteristics) 2018) Assessing the Quantitative indicator Sustainability of To assess the global measurements (scale: 0 to Post- Economic, (Molinos- Small Wastewater sustainability of small-scale Small-scale 1) aggregated into single- Implement- 17 Environmental, Senante et al. Treatment Systems: A sanitation systems post- sanitation systems score using multi-criteria ation Social 2014, 2015) Composite Indicator implementation decision analysis with Approach expert weighting To monitor and evaluate the impacts of community-led Community-led Qualitative and Economic, total sanitation programs that Post- (UNICEF UNICEF total sanitation quantitative indicator Environmental, aim to achieve open Implement- 20 2012, 2014, Sustainability Checks latrine construction measurements; Qualitative Social, defecation free districts ation 2017) programs discussion of impacts Institutional through behavior change and latrine construction Centers for Disease To monitor and evaluate the Community water Control and American impacts of implemented supply systems, Qualitative and Red Cross integrated WASH Post- Economic, (CDC 2008; household latrines, quantitative indicator Sustainability of interventions following Implement- 11 Environmental, Sabogal et al. and community measurements; Qualitative water, sanitation and Hurricane Mitch in El ation Social 2014) hygiene education discussion of impacts hygiene interventions Salvador, Guatemala, programs (CDC/ARC WASH) Honduras, and Nicaragua

113 Framework Application and Adaptation Data collection and analysis (i.e., indicator measurements and scoring) followed the methods described by each framework. If a framework had an indicator that was ambiguous (i.e., no indication of how to measure the indicator), then literature and case data were used to define the indicator or to specify a metric and measurement scale (Table H2). For example, Capacity to

Pay from the SEI Sustainability Criteria was defined as “disposable income” with units of

“cost/person/year”; to assure the indicator was specific to sanitation sustainability it was measured as household monthly sanitation fee as a percent of average household monthly disposable income,

(where disposable income = monthly income – monthly expenses). In another example,

Participation from the TechSelect framework defines the range of a measurement scale from 0

(worst) to 100 (best); to allow for consistent definitions along this scale range, calibrations (i.e., quantitative differentiation between cases) previously generated from a qualitative comparative analysis study of sanitation success and failure were used (Davis et al. 2019b); systems were assigned a score of 0 out of 100 when there was complete non-participation of community members in planning and a score of 100 when there was high participation where community members regularly attended meetings and equally participated in all decisions. Full descriptions on indicator definitions and measurements are in Table H2.

Data Collection The various indicators and their measurements required a large range of data, so cases where extensive data and a detailed understanding of their sanitation systems’ function (success or failure) was already available and additional data could be collected were chosen. The cases meeting these criteria included 12 resource-limited communities in India with small-scale sanitation systems (Davis et al. 2019b; a). The communities selected were peri-urban, slum resettlements in Karnataka and Tamil Nadu, India. All had sanitation systems that served between

114 800 and 1000 individuals and were intended to be maintained by community operators. Ten had successful sanitation systems, where systems were (1) used daily, exclusively, and correctly by at least 75 percent of the intended population; (2) maintained correctly and on-time; and (3) complied with local regulations. Of these ten successful systems, five were conventional systems intended to solely contain and treat wastewater; five were resource recovery systems intended to produce and recover biogas, compost, or water (Table H1). Two additional communities that had failed sanitation systems (i.e., they were not being used or maintained and were not compliant with regulations) were also included to determine how frameworks differentiated between successful and failed systems. One was a conventional system and the other a resource recovery system.

These two failed systems were selected because their sanitation system levels of use, maintenance, and performance were significantly lower than all successful systems, providing a clear distinction between the failed and successful cases (Davis et al. 2019b). Only a limited number of failed cases were included due to the difficulty of data collection from failed systems but still helped determine how frameworks differentiated between successful and failed systems.

Initial data collection was conducted by the first author, with the help of translators/research assistants trained in sanitation fieldwork, from May – August 2016 and

January – May 2017, and is detailed in Davis et al., (2019b, 2019a). In summary, previous data collection included interviews, photovoice, and focus groups that explored local priorities for sanitation and sanitation systems’ ability to address priorities; interviews with community members, operators, implementing organizations, and municipalities that discussed major causes of system success or failure; and documentation, researcher observations, and technical system evaluations that assessed construction quality and system performance and validated interview data.

115 The existing dataset had to be supplemented with framework-specific measurements such as household expenses, willingness to pay, hygiene behaviors, open defecation free (ODF) verification processes, and availability of sanitation materials and financing. The additional data collection was conducted from December 2018 – February 2019 by one of the research assistants who participated in the previous data collection efforts. A survey questionnaire was created based on each framework’s specific details, using each framework’s exact survey questions when possible (Appendix C, Section C2). For example, for the SEI Sustainability Criteria, willingness to pay data was collected using the survey questions outlined in Fujita et al. (2005). The research assistant performed additional observations to evaluate new topics such as appropriate handwashing behavior. Newly collected documentation described ODF verification procedures and population growth. All data collection followed the approved Institutional Review Board protocol #16-0026.

Results & Discussion Indicators There were a total of 112 indicators from the six selected frameworks. The most common indicators were Reliability and Acceptability/Appropriateness, which were included in all six frameworks; Complexity, included in five (not CDC/ARC WASH); Investment Costs, included in four (not TAF or UNICEF Sustainability Checks); and O&M Costs, included in half (not TAF,

UNICEF Sustainability Checks, or CDC/ARC WASH). Indicators most commonly measured economic impacts and benefits, resource recovery benefits, system performance and its impacts, resource use, system use and acceptance, and technical and maintenance requirements (Table 5.2).

There were four unique indicators: Replicability, Material Use, and Compatibility with Existing

System (SEI Sustainability Criteria); and Water Infrastructure (CDC/ARC WASH). In addition to social, economic, and environmental categories, some frameworks included technical and health

116 (as subsets of environmental) and institutional (as a subset of social indicators) categories. Of all the indicators, environmental indicators were the most common (35), followed by social indicators

(27), and economic indicators (12) (Table H3).

117 Table 5.2. Summary of the main types of indicators included in the selected sanitation sustainability assessment frameworks. Table H2 presents a complete list of the individual indicators.

Framework Composite SEI UNICEF CDC/ Indicator Main Indicators Topics (# of indicators) TechSelect† Indicator Sustainability TAF Sustainability ARC Categories Approach Criteria Checks† WASH† Investment Costs (4) x x x x Economic O&M Costs (3) x x x Economic Impacts Capacity to Pay/Affordability (4) x x x Willingness to Pay* (2) x x Benefits Other System Benefits (6) x x x Health System Health & Hygiene (9) x x x Performance Odors* (2) x x & Impacts System Performance (12) x x x x x Global Warming (2) x x Resource Environmental Energy Use (2) x x Use & Land Use (2) x x x Impacts Material Use (4) x Other Water Infrastructure (1) x Reliability* (7) x x x x x x Technical & Design Life (3) x x x Maintenance Maintenance (2) x x Technical Aspects Complexity* (13) x x x x x Flexibility (3) x x Other Replicability (1) x Use (5) x x Use & Social Appropriateness/Acceptability/Satisfaction (12) x x x x x x Acceptance Education/Behavior Change (4) x x x External Support/Resources (5) x x x Institutional Institutional Legal (4) x x *Reliability and Complexity were categorized as social by the Composite Indicator Approach; Willingness to Pay was categorized as social by SEI Sustainability Criteria; System benefits were categorized as environmental (Potential for Reuse) and social (Local Development) by the Composite Indicator Approach and SEI Sustainability Criteria; Odors was categorized as social by the Composite Indicator Approach. †TechSelect, UNICEF Sustainability Checks, and CDC/ARC WASH frameworks did not explicitly categorize their indicators, so category was based on the most common category used by the other frameworks and from Balkema et al. (2002).

118 Despite some common indicators, indicators and their measurements were largely different between frameworks. Between the selected frameworks, there were 24 different indicator topics

(e.g., reliability, use, legal) (Table 5.2), and frameworks proposed different indicators to measure these topics. For example, five frameworks measured system performance. Two of these used global measurements, or indicators whose impacts are based on all supply chain resource use and emissions: TAF measured the Potential for Negative Impacts or Benefits for Natural Resources on a Larger Scale; TechSelect measured global Eutrophication Potential. The other three frameworks used local measurements: CDC/ARC WASH suggests a generic Water Quality

Results measurement; SEI Sustainability Checks uses Discharge levels of biochemical oxygen demand (BOD), nitrogen (N), and phosphorous (P); and the Composite Indicator Approach uses

Percent Removal of N, P, organic matter, and total suspended solids (TSS). While the last two measure similar aspects, how those pollutants and their removal are measured are very different; specifically, percent removal does not assure a certain water quality and can make sanitation system effluents with the same final water quality have very different indicator values since the removal is based on influent wastewater. To better align these varied indicator metrics, the most complete evaluation of system performance and its impacts may be to combine local and global approaches and measure effluent water quality based on local regulations as well as its global impact.

In addition to measuring different indicators, some frameworks had the same indicator but measured it differently. For example, for the Potential for Water Reuse indicator, the SEI

Sustainability Criteria focused on quality (defining this indicator as “the potential of technologies to achieve an effluent with enough quality to be reused”) and the Composite Indicator Approach focused on quantity (defining this indicator as “percent of the consumption of the system”). An

119 improvement to these frameworks could be to include both quantity and quality metrics. This is the approach used by the UNICEF Sustainability Checks for evaluating ODF; this framework measures both the Quality of Triggering Process and Quality of ODF Verification Process.

Measuring both allows for a more comprehensive understanding of the process’ breadth and impacts. These comprehensive indicator measurements are imperative to reflect the most important issues.

Similarly, frameworks used different measurements for the same economic indicator topics. UNICEF Sustainability Checks measures Affordability based on both self-reporting on ability to afford latrine construction and access to finance mechanisms. The SEI Sustainability

Criteria defines Capacity to Pay as disposable income. To supplement these indicators, both frameworks also measure Willingness to Pay. UNICEF Sustainability Checks measures

Willingness to Pay based on the percent of households who would spend part of a hypothetical monthly salary raise on sanitation. SEI Sustainability Criteria measured Willingness to Pay using the contingent valuation method (Fujita et al. 2005), which is the most widely-accepted measurement of willingness to pay for sanitation, but the method is highly time-intensive since it requires large sample sizes, piloting, and complicated statistical analysis (Akeju et al. 2018).

Overall, there is a lack of agreement on how to best measure important sustainability factors.

In addition to different measurements, frameworks proposed few economic indicators compared to other topics. Traditionally, the economic pillar of sustainability focused solely on capital and O&M costs (Laramee and Davis 2013; Rebitzer et al. 2003; Santoyo-Castelazo and

Azapagic 2014), without considering the sources of funds. Infrastructure, especially sanitation, in resource-limited communities is often heavily subsidized by governments, implementing organizations, and/or donors (Andersson et al. 2016; Whittington et al. 2012). Sanitation economic

120 sustainability efforts are now focused on reducing reliance on subsidies due to their uncertainty

(WaterAid 2011a; Whittington et al. 2012) and because subsidies limit the development of local capacity (Seppälä 2002). Despite this shift in discourse, most existing frameworks’ economic indicators still do not consider the influence of subsidies. The SEI Sustainability Checks and TAF frameworks stand out as exceptions, with the former considering the percent contributions from individuals (i.e., community) and society (i.e., government or external organizations) and the latter evaluating affordability based on the need for permanent external subsidies.

There was wide variation in the social indicators proposed and their measurements. Some frameworks encouraged the selection of context-specific indicators and metrics by providing optional indicators (SEI Sustainability Criteria) or multiple metrics for one indicator (UNICEF

Sustainability Checks). Others proposed social indicators that are posited to be universal (e.g.,

Odors, Visual Impacts, Noise, Convenience), but the importance of these has been shown to vary by context (Davis et al. 2019a; Nawab et al. 2006). Research demonstrates that priorities are context-specific (Kapiriri and Norheim 2002; Nawab et al. 2006), and especially that some communities value some aspects (e.g., Privacy, Low Cost) far more than topics specified in some frameworks like Odors or Noise (Davis et al. 2019a). Therefore, measurements of social sustainability may also need to be community-specific. One possible metric that could improve social sustainability measurement is the Addressed Sanitation Priorities score described in Davis et al. (2019a), which accounts for context-specific priorities and demonstrated strong differentiation between successful and failed systems. In this approach, community-specific priorities are identified and ranked based on local importance, and sanitation technologies are evaluated to determine how well they address those priorities.

121 Another difference between frameworks is that some measure potential values and others measure actual values. Evaluating the theoretical, or potential, values of indicators is useful for comparing between technology alternatives in the planning phase but does not accurately characterize the sustainability of built infrastructure. For example, the Composite Indicator

Approach measures Odors based on theoretical odor emissions factors estimated for each treatment module (Stellacci et al. 2010) but does not account for how odor emissions may be greater when systems are not performing to their intended designs. In a different example,

CDC/ARC WASH’s Occurrence of Natural Disasters does not account for if and how the event impacted the system or if damage was repaired. The ability for sustainability frameworks to differentiate between sustainable and unsustainable systems could be improved by proposing indicators that reflect the statuses and effects of implementations.

When applying the frameworks, some adaptation of indicators was encouraged by frameworks to help users select context-appropriate indicators and metrics (e.g., UNICEF

Sustainability Checks’ use of multiple metrics for indicators such as Affordability). This approach can be helpful to include in a framework since it is important to engage local stakeholders in the sustainability evaluation itself (Lennartsson et al. 2009; Olschewski and Casey 2015), However, these context-specific adaptations were for a minority of indicators. In contrast, one-third of indicators needed adaptations due to lack of clear definitions and metrics (Table H4). Three frameworks (TechSelect, Composite Indicator Approach, and SEI Sustainability Criteria) required adaptations for close to 50 percent of their indicators. This high level of adaptation was mainly due to the frameworks’ use of scales without clear differentiation between scale steps (e.g., a generic three-level scale of low, medium, and high). Only one framework, TAF, required no adaptation because the framework included lengthy appendices with complete indicator

122 definitions, measurement questions, and evaluation methods (Skat Foundation and Rural Water

Supply Network 2018). Overall, frameworks would benefit from increased specificity, such as providing more comprehensive and clear definitions for each indicator and either a specific metric to use or a range of metrics to select from based on local context.

Within the frameworks’ indicator categories, technical indicators required the most adaptation (76%), followed by social (55%), institutional (44%), health (43%), economic (33%), and environmental (26%) (Table H5). Economic and environmental indicators have long been the most studied aspects of sustainability (Balkema et al. 2002; Basiago 1998; Dillard et al. 2008) and have well-established and widely accepted measurement methods (Hunkeler and Rebitzer 2003;

ISO 1997). Social, institutional, and health indicators have been previously considered important for sustainability (Azar et al. 1996; Bell and Morse 2008), but have been historically under- emphasized and typically not quantified or measured (Boström 2012; Hutchins and Sutherland

2008). The high amount of adaptation required for technical indicators could be due to the large variety of proposed technical indicators. The specificity of technical indicators could be improved by drawing from extensive monitoring and evaluation processes (Jacimovic and Bostoen 2017;

Mercy Corps 2006; World Health Organization 2012). For example, Jacimovic and Bostoen

(2017) propose five well-defined technical indicators, (e.g., Sanitation Coverage, Average

Unavailability of WASH Services) to monitor humanitarian WASH interventions and align these measurements with WASH humanitarian sector (i.e., Sphere) standards.

Overall, sustainability frameworks should also be practical, replicable, and easy for implementing organizations to use and understand. Since the resource requirements of sustainability assessments are a main barrier to framework use (UNICEF 2017), frameworks should optimize the numbers and types of indicators to minimize data collection efforts without

123 compromising the amount and quality of data. A framework needs to include as many diverse, non-redundant indicators as needed to meet its objective. Frameworks should also select indicators that are less resource-intensive, where possible. For example, the SEI Sustainability Criteria proposed measuring Risk of Infection using either a health risk assessment or a qualitative scale.

Due to resource-intensity, health risk assessments are beyond most implementing organizations’ regular capabilities (Dourson et al. 2013). Frameworks, like the SEI Sustainability Criteria, that provide well-defined alternatives (e.g., qualitative scales) to complex measurement methods, make frameworks more accessible. The extensive amount of adaptation required by most frameworks also highlights the need for frameworks to contain clear definitions and metrics associated with each indicator. The ease of framework use could be improved by clearly defining all indicators and identifying specific metrics, data collection, and analysis methods for each indicator.

Framework Application Monitoring Frameworks. The UNICEF Sustainability Checks and CDC/ARC WASH were both frameworks with the purpose of monitoring a single WASH project. While a comparison of the cases by individual indicator highlighted differences between systems, these monitoring frameworks did not definitively identify which systems were successful or failed. This is expected since the goal of both frameworks was to monitor a single sanitation system and not compare between systems. For the 12 cases evaluated, though, there were indicators that most clearly differentiated between successful and failed systems. For the UNICEF Sustainability Checks frameworks, those indicators were Adequate O&M, Existence of Community-Based Body,

Existence of Post-Triggering Follow-up Support, and Functional Monitoring System. The presence of these factors have been shown to be essential for sanitation success (Davis et al. 2019b; Starkl et al. 2013b). For the CDC/ARC WASH framework, those indicators were Sanitation Coverage,

Water Infrastructure, Presence of Active Committees, Follow-up from Outside Organizations, and

124 System Performance. These indicators reflect system performance and essential maintenance support. For all of these indicators, the failed cases had lower performance than the other successful systems. For some indicators though, a failed system had a high performance. For example, failed case 17 had the third highest rate of appropriate Handwashing Practice in UNICEF

Sustainability Checks.

The main benefit of both frameworks was their indicator summary table (Tables H6 and

H7). Each table can be used to generate discussions with local stakeholders to identify improvements to an existing system. For example, CDC/ARC WASH identifies the presence of a highly active committee to maintain WASH infrastructure as critically important for the long-term provision of sustainable WASH service; systems with no committee (as in Cases 12 and 17) and committees that had some maintenance or organizational challenges (as in Cases 1, 2, 3, and 14) could use this monitoring information to strengthen committees’ maintenance preparedness.

Further, these frameworks can be used to set quantitative improvement targets. The CDC/ARC

WASH framework had three indicators that identified a quantitative target: 100% coverage for

Water Infrastructure and Sanitation Coverage; 75% coverage for Hygiene Behavior. The UNICEF

Sustainability Checks framework recognized 100% as the ideal target for all indicators, but encouraged that framework users set realistic, context-specific targets based on national standards or implementing organization program goals. Qualitative or quantitative targets for each indicator would help provide a clear definition of a successful (i.e., sustainable) system. Alternatively, the frameworks could have a relative definition of a successful system such that they allowed comparisons and benchmarking between systems. Either approach, a clear absolute or relative definition for sustainability overall, would strengthen the information generated from monitoring.

125 Comparison Frameworks. The TAF and SEI Sustainability Checks frameworks were both semi-quantitative frameworks that compare sanitation alternatives. They provide indicator data in summary tables (Table 5.3 and Table H8), similar to the monitoring frameworks. However, for these comparison frameworks, each indicator was scored on a relative basis; TAF evaluated impacts using +, 0, or – based on the whether an impact was positive, potentially at risk, or critical;

SEI evaluated impacts using ++, +, 0, -, or -- relative to a baseline. When evaluating a case based on its total number of positive or neutral indicator scores versus its total number of negative indicator scores, both frameworks had clear differentiation between successful and failed systems.

When the 12 cases were evaluated using TAF, all indicator scores were positive or neutral for the successful cases, and all but two indicators were negative for the failed cases, providing a very clear distinction between successful and failed cases (Table 5.3). For example, successful case 18 had 16 positive impacts (green, +), two potential impacts (yellow, 0) that should receive greater attention or resources, and negative impacts (red, -); failed case 17 had zero positive impacts, two potential impacts, and 16 negative impacts that seriously hinder the system’s sustainability (Table 5.3). For TAF, all indicators clearly differentiated between successful and failed cases except for Legal Regulation and Requirements for Registration of Producers and

Potential for Local Production of Product or Spares since their indicator scores were the same between all cases because regulations and market opportunities were similar for all communities.

TAF’s clear differentiation between cases could be in part because TAF was revised after being piloted in real-world applications (UNICEF Sustainability Checks also did this). For the SEI

Sustainability Criteria framework, the successful cases, compared to the failed cases, had at least five more indicators scores that were positive or neutral (Table H8). The indicators that differentiated most clearly between successful and failed systems were Appropriateness to Local

126 Context, Responsibility Distribution, Ability to Address Awareness Information Needs, Potential for Reuse of Water, and Potential for Reuse of Nutrients. These indicators reflect system performance and key community buy-in and maintenance resources.

Neither of these frameworks incorporated indicator weighting or aggregation, and instead emphasized that sustainability is context-specific, and therefore the interpretation of the framework results must be done in partnership with local stakeholders. Since both frameworks provide a relative evaluation, the inclusion of a baseline begins to highlight which systems are sustainable or unsustainable. Although TAF did not quantify or define overall sustainability directly, its use of color coding the indicator scores in the summary table makes the results table easy to interpret when considering equal weights between indicators. The most sustainable technologies should have mostly green indicators, which signifies positive or neutral impacts; systems with one or more red indicators have negative or critical impacts that need immediate attention. Since the SEI Sustainability Checks framework had the most indicators among the selected frameworks, color coding the summary table (as done in Table H8) could help with results interpretation, if the user wants to assume equal weights across indicators. This framework had significantly more indicators in the technical and environmental categories than other categories, so assuming equal weights could shift the focus of the sustainability evaluations. In addition to technology comparison and selection, these frameworks could also be useful for monitoring: the red or negative impacts identify critical issues and areas for improvement. Overall, the results of these comparison frameworks highlight the importance of balancing focus across sustainability pillars and of engaging stakeholders to understand the importance of the relative impacts.

127 Table 5.3. Summary of the results from TAF.

Social Economic Environmental Institutional Capacity Technical

on (user) on

(regulator/investor)

Case NumberCase

impacts(user)

technology (user)

satisfacti

needed (producer) needed

Affordability (user)

(regulator/investor) (regulator/investor)

research (producer) research

Profitability (producer)

and validation procedures andprocedures validation

product or (producer) productspares

followup (regulator/investor)

Sector capacity for validation, validation, capacitySector for

marketing (regulator/investor)

technology (regulator/investor)

spares, and services (producer) spares, and services

larger scale

Need for promotion and market market for promotionNeed and of production Potential for local

benefits for natural resources on on benefits for natural resources

Potential for benefits or negative benefits orPotential negative for or negativePotential impacts for

introduction of technologies, and ofand introduction technologies,

Supportive financial mechanisms Supportive financial

Viable supply chains for product, Viable product, supply for chains

Demand for the technology (user) Demand the technology for anduser Reliability technology of

Alignment with national strategies national Alignmentstrategies with

Support mechanisms for upscalingSupport mechanisms for

technology & accountabilitytechnology &(users)

Legal structures for management offor Legal management structures

Need for behavior change and social and forsocial behaviorNeed change

registration of producers (producer) registrationof producers

Level of technical and business skills Level of and business technical

Skillset of user or operator to manage manage orto Skillset operator user of Legal regulation and requirements for and Legal requirements regulation

18 + + + + + + + 0 + + 0 + + + + + + + 11 + + + + + + + 0 + + 0 + + + + + + + 13 0 0 + + + + + 0 + + 0 + + + + + + + 7 + 0 + + 0 + + 0 0 + 0 + + + + + + 0 15 + 0 + + 0 + 0 0 0 + 0 + + + + + + 0 14 + 0 + 0 0 0 + 0 + + 0 0 + + 0 + + 0 8 0 0 0 0 0 + + 0 0 + 0 + + + + + + 0 2 + 0 0 0 0 0 0 0 + + 0 0 + 0 + + + 0 3 + 0 + 0 0 0 0 0 0 + 0 0 + 0 + + + 0 1 - 0 0 0 0 + 0 0 0 + 0 0 + 0 + + + 0 12* ------0 - - 0 ------17* ------0 - - 0 ------Note: + = high value, neutral or positive, supportive characteristics; 0 = potential impact, could become critical, needs follow up; - = low value, negative, critical, hindering characteristics. Systems are in descending order from greatest number of positive indicators (and least negative). * denotes failed systems, all other systems were successful.

128 Single-score Frameworks. The Composite Indicator Approach and TechSelect frameworks used a framework-specified weighting scheme to aggregate the quantitative indicator data into a single-score. When applying both to the 12 cases, neither clearly differentiated between the successful and failed cases and both lacked a clear quantitative sustainability definition. The

Composite Indicator Approach employed an indicator weighting scheme using expert opinions

(Molinos-Senante et al. 2014). While the ten successful cases had higher scores than the two failed cases, there was less than a 1% difference between the “worst” successful case score and the “best” failed case score (Figure 5.1a), making it difficult to determine which systems were unsustainable.

The expert weights prioritized environmental indicators, giving it a 47% weight compared to 31% and 22% weights for the economic and social indicators, respectively. When the expert weights were replaced with equal weights across indicators (5.8% per indicator), the score difference between successful and failed case increased by 33%, which presents a clearer distinction (Figure

5.1b). This demonstrates that weights may need to be context-specific but gathering that data will require more resources. There were some individual indicators in this framework that were most helpful for differentiating between successful and failed systems; they included all of the environmental indicators: Nitrogen Efficiency Removal, Phosphorous Efficiency Removal,

Organic Matter Efficiency Removal, Suspended Solids Efficiency Removal, Potential to Recover

Products, and Potential for Water Reuse. Since these indicators were all based on measures of system performance, failed systems that were not functional scored below successful systems.

129

Figure 5.1. Comparison of weighted (a) and unweighted (b) results from the single-score aggregation framework, the Composite Indicator Approach. The clear gap between successful and failed systems in (b) highlights the need for frameworks to also explore unweighted or context- specific weighted results.

The TechSelect framework also did not have a clear differentiation between successful and failed systems. Users must select one of six scenarios options based on context (i.e., urban, suburban, or rural) and whether the system included water recycling (Kalbar et al. 2016). The closest scenario to the systems evaluated was Scenario 1 (Urban, No Recycling). The weighted, aggregated scores showed failed case 12 seventh with a score 0.3% higher than the next closest successful system (Figure H1). When the weights were replaced with equal weights across all indicators (8.3% per indicator), the two failed systems were ranked last, as would be expected; but there was only a small 3% difference between the “worst” successful case score and the “best” failed case score (Figure H1). Like the Composite Indicator Approach, these results demonstrate that the framework-specified indicator weights may not always be appropriate. Also, there were some individual indicators in this framework that were most helpful for differentiating between successful and failed systems; they included: Reliability, Flexibility, Participation, Promotion of

Sustainable Behavior, and Acceptance. Although these all required adaptation (Table H2), this combination of technical and social indicators reflected both system performance and key drivers

130 of sanitation success. These single-score frameworks may be most relevant for comparing the overall potential or actual impacts of different sanitation systems in planning or post- implementation, respectively.

Comparison of Framework Results To help compare across the diverse frameworks, each framework’s results were used to assign each case a rank from one (best, most sustainable system) to 12 (worst, least sustainable system) (Appendix H, Section H5). Each framework defined sanitation sustainability differently

(Table H10), and therefore, the results of which systems were the most and least sustainable were different (Figure 5.2). For example, successful case 13 rank ranged from first (CDC/ARC WASH) to eleventh (TechSelect). Four successful cases (Cases 7, 11, 13, and 18) were ranked highly across most frameworks, receiving the top rank from one or more frameworks. However, due to the variation in ranks between frameworks, there was no clear “most sustainable” system. The other six successful cases (Cases 1, 2, 3, 8, 14, and 15) were generally ranked in the middle by all frameworks. The two failed cases (Cases 12 and 17) were ranked last for nearly all frameworks; the exception was TechSelect, which gave failed case 12 a rank of seventh. While most of the frameworks ranked failed systems last, indicating that those systems were the “least sustainable”, the wide range of rankings between frameworks for the successful cases demonstrates that there is not yet clear agreement on the definition of a sustainable sanitation system.

131

Figure 5.2. Ranking from 1 (best) to 12 (worst) for sanitation systems from the selected frameworks. Systems are ordered on the x-axis from the best to worst average rank across all six frameworks. All systems have six symbols, but symbols overlap when multiple frameworks ranked a system the same. Results from the TechSelect and Composite Indicator Approach used the expert weightings, as presented in the frameworks. The numerical rankings are summarized in Table H9. Most frameworks have common elements, including long-term functionality and provision of services, resource management, user satisfaction, and minimization of negative impacts, but have different definitions for sustainability (Table H10). Despite the differences, each definition aligns with the foundational definitions in sustainability literature, where there is consensus that sustainability is multi-dimensional and that holistic sustainability evaluations should include social, economic, and environmental pillars (Balkema et al. 2002; Dillard et al. 2008; Vleuten-

Balkema 2003; WCED 1987). Applying each framework to the 12 illustrative cases showed the importance of a comprehensive set of indicators. For example, TAF’s Viable Supply Chains, a technical indicator, and Legal Structures for Management of Technology & Accountability, an institutional indicator, capture important factors that are often bottlenecks to sanitation service provision (Davis et al. 2019b; Devas and Grant 2003; USAID 2014). Similarly, the SEI

132 Sustainability Criteria and UNICEF Sustainability Checks’ focus on health impacts is valuable to understand the local impacts of sanitation.

While the use of broadly-accepted sustainability definitions can serve as a foundation, frameworks should include sanitation-specific definitions that provide clear guidance for results interpretation and differentiation between systems. For example, the Composite Indicator

Approach and TechSelect could add a quantitative threshold for their aggregated sustainability scores below which systems are considered entirely unsustainable. Relative comparison frameworks could follow the example of TAF, which declares that any sanitation systems that have a negative indicator score for any indicator require serious consideration and re-evaluation of the viability and resources required to ensure system sustainability. Indicator-specific targets, that identify an ideal goal (e.g., 100% sanitation coverage) and a minimum score for systems to be considered sustainable (e.g., 75% sanitation coverage), could also be a helpful way to measure sustainability in monitoring. Overall, frameworks should identify the characteristics of sustainable and unsustainable sanitation systems which will improve the sanitation sector’s understanding and measurement of sanitation sustainability.

Conclusion This evaluation of sanitation sustainability assessment frameworks identified how sustainability is currently measured. Despite the inclusion of common sustainability pillars, there is great variation between the proposed indicators, their definitions and metrics, and data processing methods. Even within common indicators, metrics vary, which makes it difficult to directly compare framework results. Most frameworks lacked complete indicator definitions or measurement guidance, so framework effectiveness could be improved with more detailed methods for indicator evaluation. A database of indicators with common definitions and metrics

133 would assist in unifying disparate sustainability evaluations, which could improve understanding of sanitation impacts globally.

The application of frameworks to successful and failed systems illuminated further insight into sustainability assessments. While most frameworks rated failed systems as the least sustainable, there was neither consensus on what constitutes a sustainable system nor which systems were the most sustainable. Framework results were influenced by the selection, categorization, and redundancy between indicators, especially within the technical and environmental pillars, and by data processing methods. While monitoring and relative comparison frameworks are useful to understand the impacts of a single system over time or the highest risk factors, respectively, both types of frameworks could benefit from providing guidance to aggregate or compare indicators and overall sustainability. The results from single-score aggregation frameworks highlight the importance of incorporating local context. If weighting schemes are used, weights should be context-specific and determined by local stakeholders instead of or in addition to expert weightings; this may align framework results better with the success or failure of implemented systems.

Finally, frameworks would also benefit from greater differentiation between sustainable and unsustainable systems and identification of thresholds for sustainability or targets for indicators. Across all frameworks, the indicators that consistently differentiated between successful and failed systems were drivers of success (external support, clear O&M responsibilities, maintenance committees, community buy-in) and system performance indicators.

Therefore, frameworks should comprehensively measure sustainability and identify minimum criteria or thresholds for all indicators. These improvements could increase use of frameworks by reducing the resource-intensity of applying the frameworks, which could lead to better sanitation

134 decision-making and in turn, improved public and environmental health, economic viability, and sanitation use and acceptance.

Chapter 5 References Akeju, T., Oladehinde, G., Abubakar, K., 2018. An Analysis of Willingness to Pay (WTP) for Improved Water Supply in Owo Local Government, Ondo State, Nigeria. Asian Res. J. Arts Soc. Sci. 5, 1–15. https://doi.org/10.9734/ARJASS/2018/39282 Andersson, K., Dickin, S., Rosemarin, A., 2016. Towards “Sustainable” Sanitation: Challenges and Opportunities in Urban Areas. Sustainability 8, 1289. https://doi.org/10.3390/su8121289 Ashley, R., Blackwood, D., Butler, D., Jowitt, P., Davies, J., Smith, H., Gilmour, D., Oltean- Dumbrava, C., 2008. Making asset investment decisions for wastewater systems that include sustainability. J. Environ. Eng. 134, 200–209. Azar, C., Holmberg, J., Lindgren, K., 1996. Socio-ecological indicators for sustainability. Ecol. Econ. 18, 89–112. https://doi.org/10.1016/0921-8009(96)00028-6 Balkema, A.J., Preisig, H.A., Otterpohl, R., Lambert, F., 2002. Indicators for the sustainability assessment of wastewater treatment systems. Urban Water 4, 153–161. https://doi.org/10.1016/S1462-0758(02)00014-6 Barnes, R., Roser, D., Brown, P., 2011. Critical evaluation of planning frameworks for rural water and sanitation development projects. Dev. Pract. 21, 168–189. https://doi.org/10.1080/09614524.2011.543269 Basiago, A.D., 1998. Economic, social, and environmental sustainability in development theory and urban planning practice. The Environmentalist 19, 145–161. Bell, S., Morse, S., 2008. Sustainability indicators: measuring the immeasurable?, 2nd ed. ed. Earthscan, London ; Sterling, VA. Boström, M., 2012. A missing pillar? Challenges in theorizing and practicing social sustainability: introduction to the special issue. Sustain. Sci. Pract. Policy 8. Brunner, N., Starkl, M., Kazmi, A., Real, A., Jain, N., Mishra, V., 2018. Affordability of Decentralized Wastewater Systems: A Case Study in Integrated Planning from India. Water 10, 1644. https://doi.org/10.3390/w10111644 CDC, 2008. Evaluation of the Sustainability of Water and Sanitation Interventions in Central America after Hurricane Mitch: February 12 – 27, 2006. U.S. Centers for Disease Control and Prevention (CDC), Atlanta. Chevalier, C., Meunier, F., 2005. Environmental assessment of biogas co- or tri-generation units by life cycle analysis methodology. Appl. Therm. Eng. 25, 3025–3041. https://doi.org/10.1016/j.applthermaleng.2005.03.011 Davis, A., Javernick-Will, A., Cook, S., 2019a. The use of qualitative comparative analysis to identify pathways to successful and failed sanitation systems. Sci. Total Environ. 663, 507–517. https://doi.org/10.1016/j.scitotenv.2019.01.291 Davis, A., Javernick-Will, A., Cook, S., 2019b. Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities. Environ. Sci. Technol. 53, 401–411. https://doi.org/10.1021/acs.est.8b04761 Devas, N., Grant, U., 2003. Local government decision-making? Citizen participation and local accountability: some evidence from Kenya and Uganda. Public Adm. Dev. 23, 307–316. https://doi.org/10.1002/pad.281

135 Dillard, J., Dujon, V., King, M.C., 2008. Understanding the Social Dimension of Sustainability. Routledge. Dourson, M., Becker, R.A., Haber, L.T., Pottenger, L.H., Bredfeldt, T., Fenner-Crisp, P.A., 2013. Advancing human health risk assessment: Integrating recent advisory committee recommendations. Crit. Rev. Toxicol. 43, 467–492. https://doi.org/10.3109/10408444.2013.807223 Fang, L.L., Valverde-Pérez, B., Damgaard, A., Plósz, B.G., Rygaard, M., 2016. Life cycle assessment as development and decision support tool for wastewater resource recovery technology. Water Res. 88, 538–549. https://doi.org/10.1016/j.watres.2015.10.016 Flores, A., Buckley, C., Fenner, R., 2009. Selecting sanitation systems for sustainability in developing countries. Water Sci. Technol. 60, 2973. https://doi.org/10.2166/wst.2009.375 Fujita, Y., Fujii, A., Furukawa, S., Ogawa, T., 2005. Estimation of Willingness-to-Pay (WTP) for Water and Sanitation Services through Contingent Valuation Method (CVM): A Case Study in Iquitos City, The Republic of Peru (Review No. 11). Japan Bank for International Cooperation. Gallego, A., Hospido, A., Moreira, M.T., Feijoo, G., 2008. Environmental performance of wastewater treatment plants for small populations. Resour. Conserv. Recycl. 52, 931– 940. https://doi.org/10.1016/j.resconrec.2008.02.001 Gaulke, L.S., Weiyang, X., Scanlon, A., Henck, A., Hinckley, T., 2010. Evaluation Criteria for Implementation of a Sustainable Sanitation and Wastewater Treatment System at Jiuzhaigou National Park, Sichuan Province, China. Environ. Manage. 45, 93–104. https://doi.org/10.1007/s00267-009-9398-1 Guest, J.S., Cook, S.M., Skerlos, S.J., Love, N.G., 2009. A Methodology to Assess the Environmental Impacts of Upgrading Wastewater Infrastructure: A Case Study to Evaluate Energy Recovery from Black Water. Proc. Water Environ. Fed. 2009, 6584– 6602. Hunkeler, D., Rebitzer, G., 2003. Life Cycle costing—paving the road to sustainable development? Int. J. Life Cycle Assess. 8, 109–110. Hutchins, M.J., Sutherland, J.W., 2008. An exploration of measures of social sustainability and their application to supply chain decisions. J. Clean. Prod. 16, 1688–1698. https://doi.org/10.1016/j.jclepro.2008.06.001 Iribarnegaray, M.A., Copa, F.R., Gatto D’Andrea, M.L., Arredondo, M.F., Cabral, J.D., Correa, J.J., Liberal, V.I., Seghezzo, L., 2012. A comprehensive index to assess the sustainability of water and sanitation management systems. J. Water Sanit. Hyg. Dev. 2, 205. https://doi.org/10.2166/washdev.2012.005 ISO, 1997. Environmental Management: Life Cycle Assessment: Principles and Framework. International Organization for Standardization. Jacimovic, R., Bostoen, K., 2017. Monitoring and Evaluation Framework for WASH Market- Based Humanitarian Programming (Guidance Document). Oxfam. Kalbar, P.P., Karmakar, S., Asolekar, S.R., 2016. Life cycle-based decision support tool for selection of wastewater treatment alternatives. J. Clean. Prod. 117, 64–72. https://doi.org/10.1016/j.jclepro.2016.01.036 Kalbar, P.P., Karmakar, S., Asolekar, S.R., 2012a. Technology assessment for wastewater treatment using multiple-attribute decision-making. Technol. Soc. 34, 295–302. https://doi.org/10.1016/j.techsoc.2012.10.001

136 Kalbar, P.P., Karmakar, S., Asolekar, S.R., 2012b. Selection of an appropriate wastewater treatment technology: A scenario-based multiple-attribute decision-making approach. J. Environ. Manage. 113, 158–169. https://doi.org/10.1016/j.jenvman.2012.08.025 Kapiriri, L., Norheim, O.F., 2002. Whose priorities count? Comparison of community-identified health problems and Burden-of-Disease-assessed health priorities in a district in Uganda. Health Expect. 5, 55–62. Katukiza, A.Y., Ronteltap, M., Niwagaba, C.B., Foppen, J.W.A., Kansiime, F., Lens, P.N.L., 2012. Sustainable sanitation technology options for urban slums. Biotechnol. Adv. 30, 964–978. https://doi.org/10.1016/j.biotechadv.2012.02.007 Laramee, J., Davis, J., 2013. Economic and environmental impacts of domestic bio-digesters: Evidence from Arusha, Tanzania. Energy Sustain. Dev. 17, 296–304. https://doi.org/10.1016/j.esd.2013.02.001 Lennartsson, M., Kvarnström, E., Lundberg, T., Buenfil, J., Sawyer, R., 2009. Comparing sanitation systems using sustainability criteria. Stockholm Environment Institute, SEI, Stockholm. Malekpour, S., Langeveld, J., Letema, S., Clemens, F., van Lier, J.B., 2013. Judgment under uncertainty; a probabilistic evaluation framework for decision-making about sanitation systems in low-income countries. J. Environ. Manage. 118, 106–114. https://doi.org/10.1016/j.jenvman.2013.01.007 Mercy Corps, 2006. Design, Monitoring & Evaluation (DM&E) Guidebook. Mercy Corps. Molinos-Senante, M., Gómez, T., Caballero, R., Hernández-Sancho, F., Sala-Garrido, R., 2015. Assessment of wastewater treatment alternatives for small communities: An analytic network process approach. Sci. Total Environ. 532, 676–687. https://doi.org/10.1016/j.scitotenv.2015.06.059 Molinos-Senante, M., Gómez, T., Garrido-Baserba, M., Caballero, R., Sala-Garrido, R., 2014. Assessing the sustainability of small wastewater treatment systems: A composite indicator approach. Sci. Total Environ. 497–498, 607–617. https://doi.org/10.1016/j.scitotenv.2014.08.026 Molinos-Senante, M., Hernández-Sancho, F., Sala-Garrido, R., 2010. Economic feasibility study for wastewater treatment: A cost–benefit analysis. Sci. Total Environ. 408, 4396–4402. https://doi.org/10.1016/j.scitotenv.2010.07.014 Muga, H.E., Mihelcic, J.R., 2008. Sustainability of wastewater treatment technologies. J. Environ. Manage. 88, 437–447. https://doi.org/10.1016/j.jenvman.2007.03.008 Murphy, H.M., McBean, E.A., Farahbakhsh, K., 2009. Appropriate technology – A comprehensive approach for water and sanitation in the developing world. Technol. Soc. 31, 158–167. https://doi.org/10.1016/j.techsoc.2009.03.010 Murray, A., Ray, I., Nelson, K.L., 2009. An innovative sustainability assessment for urban wastewater infrastructure and its application in Chengdu, China. J. Environ. Manage. 90, 3553–3560. https://doi.org/10.1016/j.jenvman.2009.06.009 Nawab, B., Nyborg, I.L.P., Esser, K.B., Jenssen, P.D., 2006. Cultural preferences in designing ecological sanitation systems in North West Frontier Province, Pakistan. J. Environ. Psychol. 26, 236–246. https://doi.org/10.1016/j.jenvp.2006.07.005 Noga, J., Wolbring, G., 2012. The Economic and Social Benefits and the Barriers of Providing People with Disabilities Accessible Clean Water and Sanitation. Sustainability 4, 3023– 3041. https://doi.org/10.3390/su4113023

137 Olschewski, A., Casey, V., 2015. The Technology Applicability Framework. A Participatory Tool to Validate Water, Sanitation, and Hygiene Technologies for Low-Income Urban Areas, in: Technologies for Development. Springer International Publishing, Cham, pp. 185–197. https://doi.org/10.1007/978-3-319-16247-8_18 Palme, U., Lundin, M., Tillman, A.-M., Molander, S., 2005. Sustainable development indicators for wastewater systems – researchers and indicator users in a co-operative case study. Resour. Conserv. Recycl. 43, 293–311. https://doi.org/10.1016/j.resconrec.2004.06.006 Rebitzer, G., Hunkeler, D., Jolliet, O., 2003. LCC—the economic pillar of sustainability: methodology and application to wastewater treatment. Environ. Prog. Sustain. Energy 22, 241–249. Rodriguez-Garcia, G., Molinos-Senante, M., Hospido, A., Hernández-Sancho, F., Moreira, M.T., Feijoo, G., 2011. Environmental and economic profile of six typologies of wastewater treatment plants. Water Res. 45, 5997–6010. https://doi.org/10.1016/j.watres.2011.08.053 Sabogal, R.I., Medlin, E., Aquino, G., Gelting, R.J., 2014. Sustainability of water, sanitation and hygiene interventions in Central America. J. Water Sanit. Hyg. Dev. 4, 89–99. https://doi.org/10.2166/washdev.2013.130 Santoyo-Castelazo, E., Azapagic, A., 2014. Sustainability assessment of energy systems: integrating environmental, economic and social aspects. J. Clean. Prod. 80, 119–138. https://doi.org/10.1016/j.jclepro.2014.05.061 Schoen, M.E., Xue, X., Wood, A., Hawkins, T.R., Garland, J., Ashbolt, N.J., 2017. Cost, energy, global warming, eutrophication and local human health impacts of community water and sanitation service options. Water Res. 109, 186–195. https://doi.org/10.1016/j.watres.2016.11.044 Schweitzer, R., Grayson, C., Lockwood, H., 2014. Mapping of Water, Sanitation, and Hygiene Sustainability Tools. Working Paper 10, IRC/Aguaconsult/Triple-S. 2014. Available online: http://www. ircwash. org/sites/default/files/triple- s_wp10mappingofwashsustainabilitytools. pdf (accessed on 7 March 2017). Schwemlein, S., Cronk, R., Bartram, J., 2016. Indicators for Monitoring Water, Sanitation, and Hygiene: A Systematic Review of Indicator Selection Methods. Int. J. Environ. Res. Public. Health 13, 333. https://doi.org/10.3390/ijerph13030333 Seppälä, O.T., 2002. Effective water and sanitation policy reform implementation: need for systemic approach and stakeholder participation. Water Policy 4, 367–388. Skat Foundation, 2013. Technology Applicability Framework (TAF) Manual. Skat Foundation, Rural Water Supply Network, 2018. [TAF] Technology Applicability Framework [WWW Document]. TAF Technol. Appl. Framew. URL https://technologyapplicability.wordpress.com/ (accessed 2.12.19). Sperling, J., Romero-Lankao, P., Beig, G., 2016. Exploring citizen infrastructure and environmental priorities in Mumbai, India. Environ. Sci. Policy 60, 19–27. https://doi.org/10.1016/j.envsci.2016.02.006 Starkl, M., Brunner, N., López, E., Martínez-Ruiz, J.L., 2013a. A planning-oriented sustainability assessment framework for peri-urban water management in developing countries. Water Res. 47, 7175–7183. https://doi.org/10.1016/j.watres.2013.10.037 Starkl, M., Brunner, N., Stenström, T.-A., 2013b. Why Do Water and Sanitation Systems for the Poor Still Fail? Policy Analysis in Economically Advanced Developing Countries. Environ. Sci. Technol. 47, 6102–6110. https://doi.org/10.1021/es3048416

138 Stellacci, P., Liberti, L., Notarnicola, M., Haas, C.N., 2010. Hygienic sustainability of site location of wastewater treatment plants. Desalination 253, 51–56. https://doi.org/10.1016/j.desal.2009.11.034 Tilley, E., Strande, L., Lüthi, C., Mosler, H.-J., Udert, K.M., Gebauer, H., Hering, J.G., 2014. Looking beyond Technology: An Integrated Approach to Water, Sanitation and Hygiene in Low Income Countries. Environ. Sci. Technol. 48, 9965–9970. https://doi.org/10.1021/es501645d UNICEF, 2017. Sustainability Checks: Guidance to Design and Implement Sustainability Monitoring in WASH. UNICEF. UNICEF, 2014. Madagascar WASH Sector Sustainability Check. UNICEF, Madagascar. UNICEF, 2012. Sustainability Checks: the UNICEF Experience in Eastern and Southern Africa. UNICEF. USAID, 2014. Local Systems: A Framework for Supporting Sustainable Development. USAID, Washington, D.C. Van Minh, H., Hung, N.V., 2011. Economic Aspects of Sanitation in Developing Countries. Environ. Health Insights 5, 63–70. https://doi.org/10.4137/EHI.S8199 Vleuten-Balkema, A.J. van der, 2003. Sustainable wastewater treatment: developing a methodology and selecting promising systems (Dissertation). Eindhoven University, Eindhoven. WaterAid, 2011. Sustainability framework (Framework). WaterAid, London, UK. WCED, 1987. Report of the World Commission on Environment and Development: Our Common Future. World Commission on Environment and Development, Oxford. Whittington, D., Jeuland, M., Barker, K., Yuen, Y., 2012. Setting Priorities, Targeting Subsidies among Water, Sanitation, and Preventive Health Interventions in Developing Countries. World Dev. 40, 1546–1568. https://doi.org/10.1016/j.worlddev.2012.03.004 WHO, UNICEF, 2017. Progress on Drinking Water, Sanitation and Hygiene: Update and SDG Baselines. Joint Monitoring Program, Geneva. Wongburi, P., Park, J.K., 2018. Decision making tools for selecting sustainable wastewater treatment technologies in Thailand. IOP Conf. Ser. Earth Environ. Sci. 150, 012013. https://doi.org/10.1088/1755-1315/150/1/012013 World Health Organization, 2012. A toolkit for monitoring and evaluating household water treatment and safe storage programmes. World Health Organization, Geneva.

139 Chapter 6: Research Contributions

In this dissertation, I employed a multi-method approach, conducting in-depth case studies in 20 resource-limited communities in southern India with small-scale sanitation systems, to comprehensively examine how to improve the success and sustainability of sanitation systems in resource-limited contexts by identifying: the most effective way to identify priorities (Chapter 2), a protocol that evaluates how well sanitation systems address priorities (Chapter 3), causes of sanitation system success and failure (Chapter 4), and how sanitation sustainability is—and should be—measured (Chapter 5). This final chapter summarizes the key findings and contributions of this dissertation (Table 6.1) and provides recommendations for future research (further reflections and suggestions are included in Appendix I).

140 Table 6.11. Summary of the results and research contributions from each chapter. Chapter/Citation Results Contributions Chapter 2: Davis, A., Javernick-Will, A., Cook, S., 2018. A comparison of interviews, Interviews were the most effective method to identify Theoretical: Relative effectiveness of interviews, photovoice, and focus groups, and photovoice to identify comprehensive lists of sanitation and community priorities and focus groups for identifying unique priorities; Knowledge of most sanitation priorities and increase success of the top most important priorities. effective method, role, and sex to identify priorities. community-based sanitation systems. Community members identified more priorities than community Practical: Recommendations of how to collect priorities and from Environmental Science: Water Research & leaders and operators; females identified more priorities than whom, which are contradictory with current practice; Evidence for Technology 4, 1451–1463. males. efficient use of limited resources for data collection. https://doi.org/10.1039/C8EW00391B Theoretical: Comparison of priorities for sanitation to understand if Most priorities and their relative importance were case-specific; and how communities value resource recovery benefits; thus, priority assessments should be conducted in each Understanding the ability of existing technologies to address Chapter 3: Davis, A., Javernick-Will, A., community. Priorities were also project-specific, highlighting priorities; New social sustainability assessment developed to Cook, S., 2019. Priority Addressment that it is much more effective to use sanitation versus general evaluate how technologies address priorities and how addressment Protocol: Understanding the Ability and community priorities to select sanitation technologies. Potential of Sanitation Systems to Address can be improved. Existing systems do not address priorities well. Priority Priorities. Environmental Science & Practical: Dataset of newly identified community and sanitation addressment could be improved by ensuring systems function Technology 53, 401–411. priorities; Determination of technologies/approaches that best according to their intended designs, increasing knowledge of https://doi.org/10.1021/acs.est.8b04761 address priorities; Identification of the limited benefits of resource priorities in planning, or adding resource recovery when recovery based on priorities; New social sustainability assessment communities value those benefits. that can be used to evaluate any community and technology. Theoretical: Enhanced understanding of causal relationships for Chapter 4: Davis, A., Javernick-Will, A., sanitation with identified pathways that lead to success and failure Cook, S., 2019. The use of qualitative Successful sanitation systems were ones where: (1) communities in India; Calibration of causal conditions and outcomes that can be comparative analysis to identify pathways to had adequate technical and financial resources for maintenance; used in future fsQCAs of sanitation. successful and failed sanitation systems. (2) community and municipality stakeholders were engaged in Practical: Pathways to sanitation success and failure; Identification Science of The Total Environment 663, planning and knew their maintenance responsibilities; and (3) of alternatives to enable sanitation system success: when 507–517. appropriate technologies that met community needs and achieve municipalities cannot join in planning, engaging communities and https://doi.org/10.1016/j.scitotenv.2019.01.2 community buy-in were selected. addressing their priorities can achieve success. When communities 91 are not involved in planning, implementers should use behavior change education to achieve community buy-in. Most sanitation sustainability frameworks evaluate economic impacts, benefits, performance impacts, resource use & impacts, Chapter 5: Analyzing sanitation use & acceptance, and technical & maintenance aspects. Theoretical: Understanding of how sustainability is measured in sustainability assessment frameworks to frameworks used in research and practice; Recommendations for Sanitation sustainability frameworks could be improved by identify ways to improve how sanitation how to improve sustainability frameworks. providing specific definitions and metrics for indicators, sustainability is measured in resource- reducing redundancy, and using context-specific information. Practical: Adaptations and recommendations for how to measure limited communities sustainability and improve the utility of frameworks; Identification Most frameworks rate failed systems as the least sustainable, but of the best indicators of sustainability that provide the most amount (publication forthcoming) all lacked clear differentiation between sustainable and of information for the effort needed to collect the data. unsustainable systems, which could be improved with clear definitions and thresholds or targets for sustainability.

141 Addressing Priorities In Chapter 2, I compared the effectiveness of interviews, focus groups, and photovoice, as well as the influence of respondent attributes, on priority identification effectiveness. Interviews identified the most complete lists of priorities and the most important priorities. Community members identified more priorities than community leaders or sanitation system managers, and females identified slightly more priorities than males. In Chapter 3, sanitation and community priorities and their relative importance (i.e., rankings) were identified. Most priorities and their rankings were community-specific, highlighting the need to conduct priority assessments in each community. I then created and applied a new priority addressment protocol that identifies a community’s priorities and then quantitatively measures how well a sanitation system does or could address their priorities, based on importance. Existing sanitation systems do not currently address priorities well. Priority addressment improves when systems function according to their intended designs and when systems have expanded benefits (e.g., recovery of water or biogas) that align with local priorities.

Theoretical Contributions. To my knowledge, this is the first study to systematically compare and evaluate how three commonly used qualitative priority identification methods, as well as respondent role and sex, influence priority identification effectiveness. The priority addressment protocol presents a new social sustainability assessment to evaluate the current ability of sanitation systems to address local priorities and to evaluate how sanitation technology alternatives could improve addressment.

Recommendations for Implementing Organizations. Knowledge of the most effective method and sources to identify sanitation priorities can help implementing organizations select appropriate ways to identify priorities. Instead of relying heavily on focus groups and community leader and male operator perspectives, implementing organizations should use interviews and

142 solicit a diversity of community member perspectives. Implementing organizations can also use the priority addressment protocol to strengthen sanitation technology selection in planning and evaluate the effectiveness of their interventions on addressing priorities post-implementation.

Since not all types of resource recovery are equally valued and beneficial, implementing organizations should instead focus on recovering resources that best match local priorities, which most often are water and biogas. The use of this new tool could help to elucidate which technologies and strategies minimize tradeoffs and meet the most priorities long-term.

Recommendations for Municipalities. Municipalities can use the priority addressment protocol to identify technologies that are compatible with municipal priorities and resources.

Municipalities can also use the results of poorly addressed or unaddressed priorities to revise policies and resource allocation for sanitation. For example, within the 20 case studies, some priorities were unaddressable because municipalities commissioned community-wide treatment systems and did not sponsor individual household septic tanks.

Recommendations for Communities. Communities could use the priority addressment protocol to better understand their collective priorities and relative importance. This could facilitate consensus building and allow communities to better communicate their needs, increase buy-in, and generate evidence for how different sanitation technologies align with or are incompatible with their priorities.

Limitations and Recommendations for Future Research. The comparison of priority identification methods was not exhaustive, so this work could be expanded by comparing these results with other methods such as community mapping or surveys. It may also be useful to examine how to improve photovoice’s ability to capture abstract priorities. Additionally, despite a consensus for the need to incorporate local priorities in sanitation system design, few organizations

143 can regularly conduct priority assessments and priorities remain unaddressed. Research should continue to explore how to increase the use of priority assessments and how to better incorporate priorities into decision-making. Future research should also apply the priority addressment protocol to a wider variety of technologies (e.g., household latrines) and approaches (e.g., market- based solutions) to understand their ability to address priorities. Together, these efforts may lead to an even greater understanding of appropriate sanitation technologies, acceptance, and use.

Pathways to Sanitation Success and Failure In Chapter 4, I used fuzzy-set qualitative comparative analysis (fsQCA) to uncover the combinations of social, technical, economic, and institutional factors that led to sanitation success or failure. Successful systems required adequate operation & maintenance (O&M) funds, clear

O&M plans, technical support, local stakeholder engagement, and community buy-in; failed systems lacked many of these important factors.

Theoretical Contributions. This study furthers understanding of complex causal relationships by illuminating how social, technical, economic, and institutional factors interact in combination to influence sanitation success and failure. These results demonstrate the importance of holistic approaches to studying sanitation success and failure. The calibrations—or common measurements of differences between cases for a given causal condition—can aid future cross- case comparative studies of sanitation success or failure.

Recommendations for Implementing Organizations. Organizations who initiate sanitation projects in resource-limited communities can use these findings to ensure that (1) communities have adequate technical and financial resources for maintenance; (2) community and municipality stakeholders are engaged in planning and know their maintenance responsibilities; and (3) appropriate technologies are selected that meet community needs and achieve community buy-in.

There was more than one pathway to success, which can enable implementing organizations to

144 focus on the pathway that aligns best with local context and available resources. For instance, if a community’s priorities are not addressable due to technology or project limitations, behavior change education can help incentivize use and maintenance. If a municipality lacks the resources to participate in planning, communities should be involved in more decisions.

Recommendations for Municipalities. Municipalities can use these findings to (1) highlight the importance of their attendance in planning meetings and in co-creating O&M plans; (2) ensure their actions and policy decisions do not inhibit sanitation system function; and (3) increase the amount of human and financial resources dedicated to sanitation maintenance support.

Recommendations for Communities. Resource-limited communities can use these findings to (1) advocate for sanitation technology selection that aligns with their priorities, especially when communities are expected to shoulder maintenance responsibilities; (2) communicate with implementing organizations and municipalities to understand maintenance roles and system ownership; and (3) identify how to effectively manage system income to have funds available for maintenance emergencies and/or for local development opportunities, such as micro-loan programs.

Limitations and Recommendations for Future Research. The fsQCA study identified important pathways to sanitation success and failure for small-scale sanitation systems in resource- limited communities in southern India. Future research should explore how these findings may be similar or different when there is variability in the important domain conditions, such as socio- economic status, geographical location, and culture. Replication of this study in other contexts can further illuminate the influence of local and national policies and cultures on the important causal conditions for sanitation success or failure. Additionally, it would be valuable to identify the institutional supports and barriers for implementing sanitation success pathways. For example,

145 there is a need to understand the drivers behind municipalities lacking the financial or human resources to participate in planning and support maintenance.

Measuring Sanitation Sustainability In Chapter 5, I evaluated and applied six sanitation sustainability frameworks. Common indicator types included economic impacts, benefits, performance impacts, resource use and impacts, use and acceptance, and technical and maintenance aspects. Most frameworks lacked indicator definitions or measurement guidance; framework effectiveness could be improved with more detailed methods for indicator evaluation. Additionally, while most frameworks rated failed systems as the least sustainable, there was neither consensus on what constitutes a sustainable system nor which systems were the most sustainable. Framework results were influenced by the selection, categorization, and redundancy between indicators, especially within the technical and environmental pillars. Results were also influenced by the system evaluation and comparison methods, especially when expert opinions determined indicator weights. For frameworks that used expert weightings of indicators, differences between successful and failed system sustainability was even smaller; in contrast, unweighted results and results using context-specific weightings improved differentiation between systems. Overall, frameworks would benefit from clear differentiation between sustainable and unsustainable systems and identification of thresholds for sustainability or targets for indicators.

Theoretical Contributions. The comparison of sustainability frameworks illuminated how sanitation sustainability is commonly measured and identified drivers behind the limited use of these frameworks. Since the results of using sustainability frameworks are rarely shared publicly and frameworks are often developed without pilot testing in real-world applications, this study provides the first comparison of the results of applying multiple frameworks within the same context. This study takes first steps towards elucidating how to adjust the selection and definitions

146 of sustainability indicators to achieve greater consensus and to increase the effectiveness of sustainability frameworks.

Recommendations for Implementing Organizations. Implementing organizations can use these findings to choose the best framework based on their needs and project phase (e.g., alternatives comparison vs. monitoring). As with priorities, implementing organizations should elucidate community-specific weightings of sustainability indicators to measure sustainability in a context-specific manner as well as compare weighted and unweighted results. When implementing organizations use frameworks, they should share the results of their sustainability evaluations to increase understanding of how sustainability is measured and the overall impact of sanitation interventions.

Recommendations for Municipalities. Municipalities should incorporate sustainability frameworks into sanitation planning and evaluation processes. Municipalities can improve framework adaptation by making indicators specific to municipal, regional, and national goals, which can also help define “sustainability” and differentiate between sustainable and unsustainable systems. Municipalities can also use framework results to support the implementation of more sustainable sanitation systems.

Limitations and Recommendations for Future Research. The comparison of sanitation sustainability frameworks identified the benefits and drawbacks of six approaches to measure sustainability. This comparison was not exhaustive and comparing more frameworks could expand understanding on how to best measure sustainability. Similarly, since all frameworks were applied to small-scale sanitation systems, it would be beneficial to also apply frameworks to other types of sanitation interventions, such as large-scale municipal systems and individual household toilets and treatment systems. Additionally, while many implementing organizations and governments

147 have strong monitoring and evaluation programs, widespread application of sustainability frameworks would enable a better understanding of the impacts of implemented technologies.

Particularly as the Sustainable Development Goals aim to greatly increase sanitation coverage, it is imperative to measure the positive and negative impacts of the implemented sanitation infrastructure. Therefore, future research should focus on (1) how to increase the use of sanitation sustainability frameworks; (2) how to make frameworks easier to use by optimizing indicator measurement, information gained, and resources used; and (3) how to incorporate framework insights into improved decision-making for future implementations. A database of indicators with common definitions and metrics would assist in unifying disparate sustainability evaluations, which could improve our understanding of sanitation impacts globally.

Final Dissemination Plans Thus far, I have disseminated results through (a) journal articles (Davis et al. 2018b, 2019a; b), (b) conference papers (Davis et al. 2017, 2018a), (c) conference presentations (Davis 2016c,

2017a; b, 2018a; c), and (d) seminars (Davis 2016a; b, 2018b). I have also created relationships with USAID’s India Mission in Delhi, with TERI University in Delhi, IIT Madras in Chennai, and with several additional sanitation consulting firms and non-governmental organizations in India.

With my research results and relationships, I will be working with local governments, implementing organizations, and international aid agencies to extend the practical impacts of the work. Specifically, during a trip to India in May 2019, in partnership with the Centre for Advanced

Sanitation Solutions, a knowledge management and training center that seeks to build capacity in the sanitation sector in India, I will work with local engineers, project planners, and government officials to improve their understanding of important social, technical, economic, and institutional factors that support sustainable sanitation outcomes. I will host workshops with participating communities that seek to build local capacity for sustainable sanitation systems. Using my results,

148 I will focus training on empowering community members to identify their priorities and advocate for their needs with implementing organizations, and I will focus on the conditions that lead to sustainable systems, pointing to key takeaways from this study that can be shared. I plan to publish both a peer-reviewed journal article for Chapter 5 and a policy brief to ensure that the findings on how to improve sanitation sustainability frameworks are shared widely within research and practice communities. Finally, I hope to partner with implementing organizations to identify how failed sanitation systems can be renovated and how the supporting environment can be strengthened so that these systems do not have prolonged negative impacts on communities.

Reflections

Summary Overall, this dissertation aims to help implementing organizations, communities, and municipalities improve their project planning, decision-making, and stakeholder engagement strategies to facilitate sustainable sanitation outcomes. The findings encourage a renewed emphasis on appropriate technology selection that incorporates local priorities and engages local community and government stakeholders, which has the potential to increase system use and acceptance. The findings also demonstrate ways to effectively use limited resources to improve the long-term functionality of sanitation and understand system impacts. Together, these findings and tools chart a path for improving sanitation success and sustainability, especially in resource- limited communities.

Chapter 6 References Davis, A. (2016a). “Sustainable Community-Based Sanitation: Improving Resource Recovery Acceptance and Life Cycle Performance.” Guest Seminar, Bangalore, India. Davis, A. (2016b). “Qualitative Comparative Analysis: A Logical Method to Simplify Complexity.” Guest Seminar, Bangalore, India.

149 Davis, A. (2016c). “Priorities for Sanitation and Energy in Resource-Limited Communities.” Conference Presentation, Chapel Hill, North Carolina. Davis, A. (2017a). “Multi-Method Approach to Identify Community Priorities for Sanitation Systems.” Conference Presentation, Lake Tahoe, CA. Davis, A. (2017b). “Avoiding Failure: The Use of Qualitative Comparative Analysis to Identify Pathways to Successful Sanitation Interventions.” Conference Presentation, Chapel Hill, North Carolina. Davis, A. (2018a). “Avoiding Failure: The Use of Qualitative Comparative Analysis to Identify Pathways to Successful Sanitation Interventions.” Conference Presentation, Brijuni, Croatia. Davis, A. (2018b). “Sustainable Sanitation Systems: Understanding Priorities, Processes, and Pathways to Success.” Guest Seminar, Zurich, Switzerland. Davis, A. (2018c). “Why do sanitation systems still not address user priorities?” Conference Presentation, Chapel Hill, North Carolina. Davis, A., Javernick-Will, A., and Cook, S. (2017). “Multi-Method Approach to identify Community Priorities for Sanitation Systems.” Stakeholder Perspectives and Engineering Projects, EPOS Society, Lake Tahoe, CA. Davis, A., Javernick-Will, A., and Cook, S. (2018a). “Avoiding Failure: The Use of Qualitative Comparative Analysis to Identify Pathways to Successful Sanitation Interventions.” Brijuni, Croatia, 25. Davis, A., Javernick-Will, A., and Cook, S. (2018b). “A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems.” Environmental Science: Water Research & Technology, 4, 1451– 1463. Davis, A., Javernick-Will, A., and Cook, S. (2019a). “Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities.” Environmental Science & Technology, 53(1), 401–411. Davis, A., Javernick-Will, A., and Cook, S. (2019b). “The use of qualitative comparative analysis to identify pathways to successful and failed sanitation systems.” Science of The Total Environment, 663(1), 507–517.

150 Complete Dissertation References

Aboud, F. E., and Yousafzai, A. K. (2018). “Health and nutrition interventions for infant development.” The Lancet Child & Adolescent Health, 2(4), 231–233.

Ajayi, R. (2006). “Women’s participation in self-help community development projects in Ndokwa agricultural zone of Delta State, Nigeria.” Community Development Journal, 41(2), 189–209.

Akeju, T., Oladehinde, G., and Abubakar, K. (2018). “An Analysis of Willingness to Pay (WTP) for Improved Water Supply in Owo Local Government, Ondo State, Nigeria.” Asian Research Journal of Arts & Social Sciences, 5(3), 1–15.

Alkire, S. (2005). Valuing Freedoms: Sen’s Capability Approach and Poverty Reduction. Oxford University Press, Oxford.

Allotey, P., Reidpath, D., Kouamé, A., and Cummins, R. (2003). “The DALY, context and the determinants of the severity of disease: an exploratory comparison of paraplegia in Australia and Cameroon.” Social Science & Medicine, 57(5), 949–958.

Alsaawi, A. (2014). “A critical review of qualitative interviews.” European Journal of Business and Social Sciences, 3(4), 149–156.

Andersson, K., Dickin, S., and Rosemarin, A. (2016). “Towards ‘Sustainable’ Sanitation: Challenges and Opportunities in Urban Areas.” Sustainability, 8(12), 1289.

Andres, L. A., Briceno, B., Chase, C., and Echenique, J. A. (2014). Sanitation and Externalities: Evidence from Early Childhood Health in Rural India. Policy Research Working Paper, The World Bank, South Asia Region.

Arnstein, S. R. (1969). “A ladder of citizen participation.” Journal of the American Planning Association, 35(4), 216–224.

Ashley, R., Blackwood, D., Butler, D., Jowitt, P., Davies, J., Smith, H., Gilmour, D., and Oltean- Dumbrava, C. (2008). “Making asset investment decisions for wastewater systems that include sustainability.” Journal of Environmental Engineering, 134(3), 200–209.

Azar, C., Holmberg, J., and Lindgren, K. (1996). “Socio-ecological indicators for sustainability.” Ecological Economics, 18(2), 89–112.

Badowski, N., Castro, C. M., Montgomery, M., Pickering, A. J., Mamuya, S., and Davis, J. (2011). “Understanding Household Behavioral Risk Factors for Diarrheal Disease in Dar es

151 Salaam: A Photovoice Community Assessment.” Journal of Environmental and Public Health, 2011, 1–10.

Balkema, A. J., Preisig, H. A., Otterpohl, R., and Lambert, F. (2002). “Indicators for the sustainability assessment of wastewater treatment systems.” Urban Water, 4(2), 153–161.

Banerjee, A. (2013). Legal and policy framework for wastewater treatment and reuse in India: A background review. GIZ-India.

Bao, P. N., Aramaki, T., and Hanaki, K. (2013). “Assessment of stakeholders’ preferences towards sustainable sanitation scenarios: Assessment on sanitation scenarios.” Water and Environment Journal, 27(1), 58–70.

Bapat, M., and Agarwal, I. (2003). “Our needs, our priorities; women and men from the slums in Mumbai and Pune talk about their needs for water and sanitation.” Environment and Urbanization, 15(2), 71–86.

Barnes, R., Roser, D., and Brown, P. (2011). “Critical evaluation of planning frameworks for rural water and sanitation development projects.” Development in Practice, 21(2), 168–189.

Basiago, A. D. (1998). “Economic, social, and environmental sustainability in development theory and urban planning practice.” The Environmentalist, 19(2), 145–161.

Basurto, X., and Speer, J. (2012). “Structuring the Calibration of Qualitative Data as Sets for Qualitative Comparative Analysis (QCA).” Field Methods, 24(2), 155–174.

Battacharyya, U. (2015). “Community participation in water and sanitation services: role of Panchayati Raj in India.” Water, sanitation and hygiene services beyond 2015: Improving access and sustainability, Loughborough, England.

Bazeley, P., and Jackson, K. (2013). Qualitative Data Analysis with NVivo. Sage Publications.

Bell, S., and Morse, S. (2008). Sustainability indicators: measuring the immeasurable? Earthscan, London ; Sterling, VA.

Birt, L., Scott, S., Cavers, D., Campbell, C., and Walter, F. (2016). “Member checking: A tool to enhance trustworthiness or merely a nod to validation?” Qualitative Health Research, 26(13), 1802–1811.

Bisung, E., Elliot, S. J., Abudho, B., Karanja, D. M., and Schuster-Wallace, C. J. (2015a). “Using Photovoice as a Community Based Participatory Research Tool for Changing Water, Sanitation, and Hygiene Behaviors in Usoma, Kenya.” BioMed Research International, 8(2), 1–10.

152 Bisung, E., Elliott, S. J., Abudho, B., Schuster-Wallace, C. J., and Karanja, D. M. (2015b). “Dreaming of toilets: Using photovoice to explore knowledge, attitudes and practices around water–health linkages in rural Kenya.” Health & Place, 31, 208–215.

Black, M. (1998). Learning What Works: A 20 Year Retrospective View on International Water and Sanitation Cooperation. UNDP - World Bank Water and Sanitation Program, Washington, D.C.

Boström, M. (2012). “A missing pillar? Challenges in theorizing and practicing social sustainability: introduction to the special issue.” Sustainability: Science, Practice, & Policy, 8(1).

Bouabid, A., and Louis, G. E. (2015). “Capacity factor analysis for evaluating water and sanitation infrastructure choices for developing communities.” Journal of Environmental Management, 161, 335–343.

Breslin, E. D. (2003). Demand response approach in practice: why sustainability remains elusive. WaterAid, Mozambique.

Brikké, F. (2000). “Operation and Maintenance of rural water supply and sanitation systems.” A training package for managers and planners. IRC International Water and Sanitation Center, WHO.

Brikké, F., and Bredero, M. (2003). Linking technology choice with operation and maintenance in the context of community water supply and sanitation. WHO and IRC WASH, Geneva, Switzerland.

Brown, A. (2010). “Empowerment and Gender Equality for Water and Sanitation in Rural India: Two Case Studies.” Master of Science, Oregon State University, Corvallis, Oregon.

Brunner, N., Starkl, M., Kazmi, A., Real, A., Jain, N., and Mishra, V. (2018). “Affordability of Decentralized Wastewater Systems: A Case Study in Integrated Planning from India.” Water, 10(11), 1644.

Burra, S., Patel, S., and Kerr, T. (2003). “Community-designed, built and managed toilet blocks in Indian cities.” Environment and Urbanization, 15(2), 11–32.

Cairncross, S., Shordt, K., Zacharia, S., and Govindan, B. K. (2005). “What causes sustainable changes in hygiene behaviour? A cross-sectional study from Kerala, India.” Social Science & Medicine, 61(10), 2212–2220.

153 Carey, J. M. M. A. (1994). “The Group Effect in Focus Groups: Planning, Implementing, and Interpreting Focus Group Research.” Critical Issues in Qualitative Research Methods, Morse, Janice M., ed., SAGE, Thousand Oaks, California.

Catalani, C., and Minkler, M. (2010). “Photovoice: A Review of the Literature in Health and Public Health.” Health Education & Behavior, 37(3), 424–451.

CDC. (2008). Evaluation of the Sustainability of Water and Sanitation Interventions in Central America after Hurricane Mitch: February 12 – 27, 2006. U.S. Centers for Disease Control and Prevention (CDC), Atlanta.

Central Pollution Control Board. (2017). “Environmental Standards: Water Quality Criteria.” Central Pollution Control Board, (May 17, 2016).

Chambers, R. (1981). “Rapid rural appraisal: rationale and repertoire.” Public Administration and Development, 1, 95–106.

Chaplin, S. E. (2011). “Indian cities, sanitation and the state: the politics of the failure to provide.” Environment and Urbanization, 23(1), 57–70.

Chatterley, C., Javernick-Will, A., Linden, K. G., Alam, K., and Bottinelli, L. (2014). “A qualitative comparative analysis of well-managed school sanitation in Bangladesh.” BMC Public Health, 14(6), 1–14.

Chatterley, C., Linden, K. G., and Javernick-Will, A. (2013). “Identifying pathways to continued maintenance of school sanitation in Belize.” Journal of Water, Sanitation and Hygiene for Development, 3(3), 411–422.

Chevalier, C., and Meunier, F. (2005). “Environmental assessment of biogas co- or tri-generation units by life cycle analysis methodology.” Applied Thermal Engineering, 25(17–18), 3025–3041.

Chowns, E. (2015). “Is Community Management an Efficient and Effective Model of Public Service Delivery? Lessons from the Rural Water Supply Sector in Malawi: Is Community Management Efficient and Effective?” Public Administration and Development, 35(4), 263–276.

Cornejo, P. K., Zhang, Q., and Mihelcic, J. R. (2013). “Quantifying benefits of resource recovery from sanitation provision in a developing world setting.” Journal of Environmental Management, 131, 7–15.

154 Crawford, P., and Bryce, P. (2003). “Project monitoring and evaluation: a method for enhancing the efficiency and effectiveness of aid project implementation.” International Journal of Project Management, 21(5), 363–373.

Cronin, A. A., Ohikata, M., and Kumar, M. (2014). “Social and economic cost-benefit analysis of sanitation in Odisha State, India.” Journal of Water, Sanitation and Hygiene for Development, 4(3), 521.

Davis, A. (2016a). “Sustainable Community-Based Sanitation: Improving Resource Recovery Acceptance and Life Cycle Performance.” Guest Seminar, Bangalore, India.

Davis, A. (2016b). “Qualitative Comparative Analysis: A Logical Method to Simplify Complexity.” Guest Seminar, Bangalore, India.

Davis, A. (2016c). “Priorities for Sanitation and Energy in Resource-Limited Communities.” Conference Presentation, Chapel Hill, North Carolina.

Davis, A. (2017a). “Multi-Method Approach to Identify Community Priorities for Sanitation Systems.” Conference Presentation, Lake Tahoe, CA.

Davis, A. (2017b). “Avoiding Failure: The Use of Qualitative Comparative Analysis to Identify Pathways to Successful Sanitation Interventions.” Conference Presentation, Chapel Hill, North Carolina.

Davis, A. (2018a). “Avoiding Failure: The Use of Qualitative Comparative Analysis to Identify Pathways to Successful Sanitation Interventions.” Conference Presentation, Brijuni, Croatia.

Davis, A. (2018b). “Sustainable Sanitation Systems: Understanding Priorities, Processes, and Pathways to Success.” Guest Seminar, Zurich, Switzerland.

Davis, A. (2018c). “Why do sanitation systems still not address user priorities?” Conference Presentation, Chapel Hill, North Carolina.

Davis, A., Javernick-Will, A., and Cook, S. (2017). “Multi-Method Approach to identify Community Priorities for Sanitation Systems.” Stakeholder Perspectives and Engineering Projects, EPOS Society, Lake Tahoe, CA.

Davis, A., Javernick-Will, A., and Cook, S. (2018a). “Avoiding Failure: The Use of Qualitative Comparative Analysis to Identify Pathways to Successful Sanitation Interventions.” Brijuni, Croatia, 25.

155 Davis, A., Javernick-Will, A., and Cook, S. (2018b). “A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems.” Environmental Science: Water Research & Technology, 4, 1451– 1463.

Davis, A., Javernick-Will, A., and Cook, S. (2019a). “Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities.” Environmental Science & Technology, 53(1), 401–411.

Davis, A., Javernick-Will, A., and Cook, S. (2019b). “The use of qualitative comparative analysis to identify pathways to successful and failed sanitation systems.” Science of The Total Environment, 663(1), 507–517.

Davis, S. (2015). “Statistics on Sanitation Failures.” Improve International.

Devas, N., and Grant, U. (2003). “Local government decision-making? Citizen participation and local accountability: some evidence from Kenya and Uganda.” Public Administration and Development, 23(4), 307–316.

Diener, S., Semiyaga, S., Niwagaba, C. B., Muspratt, A. M., Gning, J. B., Mbgur, M., Ennin, J. E., Zurbrugg, C., and Strande, L. (2014). “A value proposition: Resource recovery from faecal sludge? Can it be the driver for improved sanitation?” Resources, Conservation and Recycling, 88, 32–38.

Dillard, J., Dujon, V., and King, M. C. (2008). Understanding the Social Dimension of Sustainability. Routledge.

Dourson, M., Becker, R. A., Haber, L. T., Pottenger, L. H., Bredfeldt, T., and Fenner-Crisp, P. A. (2013). “Advancing human health risk assessment: Integrating recent advisory committee recommendations.” Critical Reviews in Toxicology, 43(6), 467–492.

Eales, K., Blackett, I., Siregar, R., and Febriani, E. (2013). Review of Community-Managed Decentralized Wastewater Treatment Systems in Indonesia. WSP.

Eisenberg, D., Soller, J., Sakaji, R., and Olivieri, A. (2001). “A methodology to evaluate water and wastewater treatment plant reliability.” Water Science and Technology, 43(10), 91–99.

Fang, L. L., Valverde-Pérez, B., Damgaard, A., Plósz, B. G., and Rygaard, M. (2016). “Life cycle assessment as development and decision support tool for wastewater resource recovery technology.” Water Research, 88, 538–549.

156 Fischer, M., and Maggetti, M. (2017). “Qualitative Comparative Analysis and the Study of Policy Processes.” Journal of Comparative Policy Analysis: Research and Practice, 19(4), 345– 361.

Flores, A., Buckley, C., and Fenner, R. (2009). “Selecting sanitation systems for sustainability in developing countries.” Water Science & Technology, 60(11), 2973.

Freudenberger, K. S. (2011). Rapid Rural Appraisal and Participatory Rural Appraisal: A Manual for CRS field Workers and Partners. Manual, Catholic Relief Services, Baltimore, Maryland.

Fujita, Y., Fujii, A., Furukawa, S., and Ogawa, T. (2005). Estimation of Willingness-to-Pay (WTP) for Water and Sanitation Services through Contingent Valuation Method (CVM): A Case Study in Iquitos City, The Republic of Peru. Review, Japan Bank for International Cooperation, 59–87.

Gabe, J., Trowsdale, S., and Vale, R. (2009). “Achieving integrated urban water management: planning top-down or bottom-up?” Water Science and Technology, 59(10), 1999–2008.

Gallego, A., Hospido, A., Moreira, M. T., and Feijoo, G. (2008). “Environmental performance of wastewater treatment plants for small populations.” Resources, Conservation and Recycling, 52(6), 931–940.

Garrido, J. M., Fdz-Polanco, M., and Fdz-Polanco, F. (2013). “Working with energy and mass balances: a conceptual framework to understand the limits of municipal wastewater treatment.” Water Science and Technology, 67(10), 2294–2301.

Gaulke, L. S., Weiyang, X., Scanlon, A., Henck, A., and Hinckley, T. (2010). “Evaluation Criteria for Implementation of a Sustainable Sanitation and Wastewater Treatment System at Jiuzhaigou National Park, Sichuan Province, China.” Environmental Management, 45(1), 93–104.

Government of India. (2016). “Swachh Bharat.” Swachh Bharat Mission “My Clean India,” (Sep. 9, 2016).

Guest, J. S., Cook, S. M., Skerlos, S. J., and Love, N. G. (2009a). “A Methodology to Assess the Environmental Impacts of Upgrading Wastewater Infrastructure: A Case Study to Evaluate Energy Recovery from Black Water.” Proceedings of the Water Environment Federation, 2009(9), 6584–6602.

Guest, J. S., Skerlos, S. J., Barnard, J. L., Beck, M. B., Daigger, G. T., Hilger, H., Jackson, S. J., Karvazy, K., Kelly, L., Macpherson, L., Mihelcic, J. R., Pramanik, A., Raskin, L., Van Loosdrecht, M. C. M., Yeh, D., and Love, N. G. (2009b). “A New Planning and Design

157 Paradigm to Achieve Sustainable Resource Recovery from Wastewater.” Environmental Science & Technology, 43(16), 6126–6130.

Gupta, A. (2015). “EcoSan: Sustainable solution to sanitation problem in rural India.” Water, sanitation and hygiene services beyond 2015: Improving access and sustainability, Loughborough, England.

Hacker, M. E., and Kaminsky, J. A. (2017). “Cultural preferences for the methods and motivation of sanitation infrastructure development.” Journal of Water Sanitation and Hygiene for Development, 7(3), 407–415.

Haller, L., Hutton, G., and Bartram, J. (2007). “Estimating the costs and health benefits of water and sanitation improvements at global level.” Journal of Water and Health, 5(4), 467–480.

Harrell, M. C., and Bradley, L. M. A. (2009). Data Collection Methods: Semi-Structured Interviews and Focus Groups. Rand Corporation for the National Defense Research Institute, Arlington, VA.

Harris, D., Kooy, M., and Jones, L. (2011). Analysing the governance and political economy of water and sanitation service delivery. Overseas Development Institute, London, UK.

Harris, M., Alzua, M. L., Osbert, N., and Pickering, A. (2017). “Community-Level Sanitation Coverage More Strongly Associated with Child Growth and Household Drinking Water Quality than Access to a Private Toilet in Rural Mali.” Environmental Science & Technology, 51(12), 7219–7227.

Harris-Lovett, S. R., Binz, C., Sedlak, D. L., Kiparsky, M., and Truffer, B. (2015). “Beyond User Acceptance: A Legitimacy Framework for Potable Water Reuse in California.” Environmental Science & Technology, 49(13), 7552–7561.

Hawkins, P., Blackett, I., and Heymans, C. (2013). Poor-inclusive urban sanitation: An overview. Targeting the Urban Poor and Improving Services in Small Towns, WSP, The World Bank, Washington, D.C.

Hino, A., and Imai, R. (2018). “Ranking and Rating: Neglected Biases in Factor Analysis of Postmaterialist Values.” International Journal of Public Opinion Research, 1–14.

Hofstede, G. (1998). “Attitudes, values and organizational culture: Disentangling the concepts.” Organization studies, 19(3), 477–493.

Hofstede, G. (2012). “India.” Cultural Dimensions, (Aug. 7, 2017).

158 Hubbard, B., Sarisky, J., Gelting, R., Baffigo, V., Seminario, R., and Centurion, C. (2011). “A community demand-driven approach toward sustainable water and sanitation infrastructure development.” International Journal of Hygiene and Environmental Health, 214(4), 326– 334.

Hunkeler, D., and Rebitzer, G. (2003). “Life Cycle costing—paving the road to sustainable development?” The international journal of life cycle assessment, 8(2), 109–110.

Hutchins, M. J., and Sutherland, J. W. (2008). “An exploration of measures of social sustainability and their application to supply chain decisions.” Journal of Cleaner Production, 16(15), 1688–1698.

Hutton, G. (2013). “Global costs and benefits of reaching universal coverage of sanitation and drinking-water supply.” Journal of Water and Health, 11(1), 1–12.

Hutton, G., Haller, L., and Water, S. (2004). Evaluation of the costs and benefits of water and sanitation improvements at the global level. World Health Organization, Geneva, Switzerland.

IDECK. (2015). Strengthening the operation & maintenance sector for servicing decentralised urban sanitation infrastructure in Bangalore: Landscape synthesis report. CDD Society and BORDA, Bangalore, India.

Iribarnegaray, M. A., Copa, F. R., Gatto D’Andrea, M. L., Arredondo, M. F., Cabral, J. D., Correa, J. J., Liberal, V. I., and Seghezzo, L. (2012). “A comprehensive index to assess the sustainability of water and sanitation management systems.” Journal of Water, Sanitation and Hygiene for Development, 2(3), 205.

ISO. (1997). Environmental Management: Life Cycle Assessment: Principles and Framework. International Organization for Standardization.

Jacimovic, R., and Bostoen, K. (2017). Monitoring and Evaluation Framework for WASH Market- Based Humanitarian Programming. Guidance Document, Oxfam.

Jenkins, M. W., and Cairncross, S. (2010). “Modelling latrine diffusion in Benin: towards a community typology of demand for improved sanitation in developing countries.” Journal of Water and Health, 08(1), 166.

Jenkins, M. W., and Curtis, V. (2005). “Achieving the ‘good life’: Why some people want latrines in rural Benin.” Social Science & Medicine, 61(11), 2446–2459.

159 Jenkins, M. W., and Scott, B. (2007). “Behavioral indicators of household decision-making and demand for sanitation and potential gains from social marketing in Ghana.” Social Science & Medicine, 64(12), 2427–2442.

Jordan, E., Javernick-Will, A., and Tierney, K. (2016). “Post-tsunami recovery in Tamil Nadu, India: combined social and infrastructural outcomes.” Natural Hazards, 84(2), 1327–1347.

Jordan, S., and Wagner, F. (1993). “Meeting women’s needs and priorities for water and sanitation in cities.” Environment and Urbanization, 5(2), 135–1445.

Judd, C. M., McClelland, G. H., and Ryan, C. S. (2009). Data Analysis: A Model Comparison Approach. Routledge, New York, NY.

Kalbar, P. P., Karmakar, S., and Asolekar, S. R. (2012a). “Technology assessment for wastewater treatment using multiple-attribute decision-making.” Technology in Society, 34(4), 295– 302.

Kalbar, P. P., Karmakar, S., and Asolekar, S. R. (2012b). “Selection of an appropriate wastewater treatment technology: A scenario-based multiple-attribute decision-making approach.” Journal of Environmental Management, 113, 158–169.

Kalbar, P. P., Karmakar, S., and Asolekar, S. R. (2016). “Life cycle-based decision support tool for selection of wastewater treatment alternatives.” Journal of Cleaner Production, 117, 64–72.

Kamble, S. M. (2014). “Participatory rural appraisal: a tool for inclusive growth and participatory development a case study of Village Marale, MS, India.” Int Res J of Soc Sci, 3(3), 48–50.

Kaminsky, J. (2016). “Institutionalizing infrastructure: photo-elicitation of cultural-cognitive knowledge of development.” Construction Management and Economics, 33(11–12), 1–15.

Kaminsky, J. A. (2015). “Cultured Construction: Global Evidence of the Impact of National Values on Sanitation Infrastructure Choice.” Environmental Science & Technology, 49(12), 7134–7141.

Kaminsky, J., and Javernick-Will, A. (2012). “Causes for Sustainable Maintenance and Operation of On-Site Sanitation Systems.” Proceedings of the 2012 Construction Research Congress, 21–23.

Kaminsky, J., and Jordan, E. (2017). “Qualitative comparative analysis for WASH research and practice.” Journal of Water Sanitation and Hygiene for Development, 7(2), 196–208.

160 Kamyotra, J. S., and Bhardwaj, R. M. (2011). “Municipal wastewater management in India.” India Infrastructure Report, 299.

Kapiriri, L., and Norheim, O. F. (2002). “Whose priorities count? Comparison of community- identified health problems and Burden-of-Disease-assessed health priorities in a district in Uganda.” Health Expectations, 5(1), 55–62.

Katukiza, A. Y., Ronteltap, M., Niwagaba, C. B., Foppen, J. W. A., Kansiime, F., and Lens, P. N. L. (2012). “Sustainable sanitation technology options for urban slums.” Biotechnology Advances, 30(5), 964–978.

Katukiza, A. Y., Ronteltap, M., Oleja, A., Niwagaba, C. B., Kansiime, F., and Lens, P. N. L. (2010). “Selection of sustainable sanitation technologies for urban slums — A case of Bwaise III in Kampala, Uganda.” Science of The Total Environment, 409(1), 52–62.

Kooy, M., and Harris, D. (2012). Political economy analysis for water, sanitation and hygiene (WASH) service delivery. Project Briefing, Overseas Development Institute, London, UK.

Krueger, R. A., and Casey, M. A. (2014). Focus Groups: A Practical Guide for Applied Research. SAGE Publications.

Kunz, N. C., Fischer, M., Ingold, K., and Hering, J. G. (2015). “Why Do Some Water Utilities Recycle More than Others? A Qualitative Comparative Analysis in New South Wales, Australia.” Environmental Science & Technology, 49(14), 8287–8296.

Laramee, J., and Davis, J. (2013). “Economic and environmental impacts of domestic bio- digesters: Evidence from Arusha, Tanzania.” Energy for Sustainable Development, 17(3), 296–304.

Leder, S. (2007). “Gender issues in water and sanitation programmes: Lessons from India.” South Asian Water Studies, (A. A. Cronin, P. K. Mehta, and A. Prakash, eds.), 5(4), 59–62.

Lennartsson, M., Kvarnström, E., Lundberg, T., Buenfil, J., and Sawyer, R. (2009). Comparing sanitation systems using sustainability criteria. Stockholm Environment Institute, SEI, Stockholm.

Lloyd-Jones, T., and Rakodi, C. (2014). Urban Livelihoods: A People-Centred Approach to Reducing Poverty. Routledge, London, UK.

López, E. D. S., Eng, E., Randall-David, E., and Robinson, N. (2005). “Quality-of-Life Concerns of African American Breast Cancer Survivors Within Rural North Carolina: Blending the Techniques of Photovoice and Grounded Theory.” Qualitative Health Research, 15(1), 99– 115.

161 Luthi, C., Morel, A., Tilley, E., and Ulrich, L. (2011). Community-Led Urban Environmental Sanitation Planning (CLUES). UN Habitat and Eawag-Sandec.

Malekpour, S., Langeveld, J., Letema, S., Clemens, F., and van Lier, J. B. (2013). “Judgment under uncertainty; a probabilistic evaluation framework for decision-making about sanitation systems in low-income countries.” Journal of Environmental Management, 118, 106–114.

Mansuri, G. (2004). “Community-Based and -Driven Development: A Critical Review.” The World Bank Research Observer, 19(1), 1–39.

Mara, D., Lane, J., Scott, B., and Trouba, D. (2010). “Sanitation and Health.” PLoS Medicine, 7(11), 1–7.

Mariwah, S., and Drangert, J.-O. (2011). “Community perceptions of human excreta as fertilizer in peri-urban agriculture in Ghana.” Waste Management & Research, 29(8), 815–822.

Marks, S. J., and Davis, J. (2012). “Does User Participation Lead to Sense of Ownership for Rural Water Systems? Evidence from Kenya.” World Development, 40(8), 1569–1576.

Marks, S. J., Komives, K., and Davis, J. (2014). “Community participation and water supply sustainability evidence from handpump projects in rural Ghana.” Journal of Planning Education and Research, 34(3), 276–286.

Marks, S. J., Kumpel, E., Guo, J., Bartram, J., and Davis, J. (2018). “Pathways to sustainability: A fuzzy-set qualitative comparative analysis of rural water supply programs.” Journal of Cleaner Production, 205, 789–798.

McConville, J. R., and Mihelcic, J. R. (2007). “Adapting Life-Cycle Thinking Tools to Evaluate Project Sustainability in International Water and Sanitation Development Work.” Environmental Engineering Science, 24(7), 937–948.

McNall, M., and Foster-Fishman, P. G. (2007). “Methods of Rapid Evaluation, Assessment, and Appraisal.” American Journal of Evaluation, 28(2), 151–168.

Mendoza, G. A., and Martins, H. (2006). “Multi-criteria decision analysis in natural resource management: A critical review of methods and new modelling paradigms.” Forest Ecology and Management, 230(1–3), 1–22.

Mercy Corps. (2006). Design, Monitoring & Evaluation (DM&E) Guidebook. Mercy Corps.

Metcalf & Eddy. (2004). Wastewater Engineering: Treatment and Resource Recovery. McGraw- Hill, New York, NY.

162 Milgram, S., and Gudehus, C. (1978). Obedience to authority. Ziff-Davis Publishing Company.

Ministry of Urban Development. (2008). National Urban Sanitation Policy. Government of India, New Delhi, India.

Molinos-Senante, M., Gómez, T., Caballero, R., Hernández-Sancho, F., and Sala-Garrido, R. (2015). “Assessment of wastewater treatment alternatives for small communities: An analytic network process approach.” Science of The Total Environment, 532, 676–687.

Molinos-Senante, M., Gómez, T., Garrido-Baserba, M., Caballero, R., and Sala-Garrido, R. (2014). “Assessing the sustainability of small wastewater treatment systems: A composite indicator approach.” Science of The Total Environment, 497–498, 607–617.

Molinos-Senante, M., Hernández-Sancho, F., and Sala-Garrido, R. (2010). “Economic feasibility study for wastewater treatment: A cost–benefit analysis.” Science of The Total Environment, 408(20), 4396–4402.

Mosler, H.-J. (2012). “A systematic approach to behavior change interventions for the water and sanitation sector in developing countries: a conceptual model, a review, and a guideline.” International Journal of Environmental Health Research, 22(5), 431–449.

Muga, H. E., and Mihelcic, J. R. (2008). “Sustainability of wastewater treatment technologies.” Journal of Environmental Management, 88(3), 437–447.

Murphy, H. M., McBean, E. A., and Farahbakhsh, K. (2009). “Appropriate technology – A comprehensive approach for water and sanitation in the developing world.” Technology in Society, 31(2), 158–167.

Murray, A., Ray, I., and Nelson, K. L. (2009). “An innovative sustainability assessment for urban wastewater infrastructure and its application in Chengdu, China.” Journal of Environmental Management, 90(11), 3553–3560.

Mwirigi, J., Balana, B. B., Mugisha, J., Walekhwa, P., Melamu, R., Nakami, S., and Makenzi, P. (2014). “Socio-economic hurdles to widespread adoption of small-scale biogas digesters in Sub-Saharan Africa: A review.” Biomass and Bioenergy, 70, 17–25.

Mwirigi, J. W., Makenzi, P. M., and Ochola, W. O. (2009). “Socio-economic constraints to adoption and sustainability of biogas technology by farmers in Nakuru Districts, Kenya.” Energy for Sustainable Development, 13(2), 106–115.

Nair, S., and Howlett, M. (2015). “Scaling up of Policy Experiments and Pilots: A Qualitative Comparative Analysis and Lessons for the Water Sector.” Water Resources Management, 29(14), 4945–4961.

163 Nawab, B., Nyborg, I. L. P., Esser, K. B., and Jenssen, P. D. (2006). “Cultural preferences in designing ecological sanitation systems in North West Frontier Province, Pakistan.” Journal of Environmental Psychology, 26(3), 236–246.

Nickerson, R. S. (1998). “Confirmation bias: A ubiquitous phenomenon in many guises.” Review of general psychology, 2(2), 175.

Noga, J., and Wolbring, G. (2012). “The Economic and Social Benefits and the Barriers of Providing People with Disabilities Accessible Clean Water and Sanitation.” Sustainability, 4(12), 3023–3041.

Novotný, J., Hasman, J., and Lepič, M. (2018). “Contextual factors and motivations affecting rural community sanitation in low- and middle-income countries: A systematic review.” International Journal of Hygiene and Environmental Health, 221(2), 121–133.

Oliveira, S. C., and Von Sperling, M. (2008). “Reliability analysis of wastewater treatment plants.” Water Research, 42(4–5), 1182–1194.

Olschewski, A., and Casey, V. (2015). “The Technology Applicability Framework. A Participatory Tool to Validate Water, Sanitation, and Hygiene Technologies for Low- Income Urban Areas.” Technologies for Development, Springer International Publishing, Cham, 185–197.

Opdyke, A. (2017). “Resilient and Sustainable Infrastructure Systems: A Comparative Analysis of Post-Disaster Shelter Coordination, Stakeholder Participation, and Training.” University of Colorado at Boulder.

Opdyke, A., Javernick-Will, A., and Koschmann, M. (2018). “A Comparative Analysis of Coordination, Participation, and Training in Post-Disaster Shelter Projects.” Sustainability, 10(11), 4241.

O’Reilly, K. (2004). “Developing Contradictions: Women’s Participation as a Site of Struggle Within an Indian NGO∗.” The Professional Geographer, 56(2), 174–184.

Orner, K. D., and Mihelcic, J. R. (2018). “A review of sanitation technologies to achieve multiple sustainable development goals that promote resource recovery.” Environmental Science: Water Research & Technology, 4, 16–32.

Palaniappan, M., Lang, M., and Gleick, P. H. (2008). A Review of Decsion-Making Support Tools in the Water, Sanitation, and Hygiene Sector. The Pacific Institute, Oakland, California.

164 Palme, U., Lundin, M., Tillman, A.-M., and Molander, S. (2005). “Sustainable development indicators for wastewater systems – researchers and indicator users in a co-operative case study.” Resources, Conservation and Recycling, 43(3), 293–311.

Peletz, R., Kisiangani, J., Bonham, M., Ronoh, P., Delaire, C., Kumpel, E., Marks, S., and Khush, R. (2018). “Why do water quality monitoring programs succeed or fail? A qualitative comparative analysis of regulated testing systems in sub-Saharan Africa.” International Journal of Hygiene and Environmental Health, 221(6), 907–920.

Plakas, K. V., Georgiadis, A. A., and Karabelas, A. J. (2016). “Sustainability assessment of tertiary wastewater treatment technologies: a multi-criteria analysis.” Water Science and Technology, 73(7), 1532–1540.

Platteau, J.-P., and Abraham, A. (2002). “Participatory Development in the Presence of Endogenous Community Imperfections.” Journal of Development Studies, 39(2), 104–136.

Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., and Podsakoff, N. P. (2003). “Common method biases in behavioral research: A critical review of the literature and recommended remedies.” Journal of Applied Psychology, 88(5), 879–903.

Prokopy, L. S. (2005). “The relationship between participation and project outcomes: Evidence from rural water supply projects in India.” World Development, 33(11), 1801–1819.

QSR International. (2015). NVivo qualitative data analysis software. QSR International Pty Ltd.

Ragin, C. C. (1987). The comparative method: moving beyond qualitative and quantitative strategies. University of California Press, Berkeley, Calif.

Ragin, C. C. (2006). “Set Relations in Social Research: Evaluating Their Consistency and Coverage.” Political Analysis, 14(3), 291–310.

Ragin, C. C. (2008). Redesigning social inquiry: fuzzy sets and beyond. University of Chicago Press, Chicago.

Ragin, C. C. (2013). Fuzzy-Set/Qualitative Comparative Analysis. Department of Sociology, University of Arizona, Tuscon Arizona.

Ragin, C. C., Patros, T., Strand, S. I., and Rubinson, C. (2017). “User’s Guide to Fuzzy-Set / Qualitative Comparative Analysis.”

Ramani, S. V., SadreGhazi, S., and Duysters, G. (2012). “On the diffusion of toilets as bottom of the pyramid innovation: Lessons from sanitation entrepreneurs.” Technological Forecasting and Social Change, 79(4), 676–687.

165 Rao, V., and Ibanez, A. M. (2003). The social impact of social funds in Jamaica: A mixed-methods analysis of participation, targeting, and collective action in community-driven development. World Bank Publications.

Raymond, C. M., Bryan, B. A., MacDonald, D. H., Cast, A., Strathearn, S., Grandgirard, A., and Kalivas, T. (2009). “Mapping community values for natural capital and ecosystem services.” Ecological Economics, 68(5), 1301–1315.

Rebitzer, G., Hunkeler, D., and Jolliet, O. (2003). “LCC—the economic pillar of sustainability: methodology and application to wastewater treatment.” Environmental Progress & Sustainable Energy, 22(4), 241–249.

Rihoux, B., and Ragin, C. C. (2009). Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques. Sage Publications, Inc.

Rodriguez-Garcia, G., Molinos-Senante, M., Hospido, A., Hernández-Sancho, F., Moreira, M. T., and Feijoo, G. (2011). “Environmental and economic profile of six typologies of wastewater treatment plants.” Water Research, 45(18), 5997–6010.

Roma, E., and Jeffrey, P. (2010). “Evaluation of community participation in the implementation of community-based sanitation systems: a case study from Indonesia.” Water Science & Technology, 62(5), 1028–1036.

Rosenquist, L. E. D. (2005). “A psychosocial analysis of the human-sanitation nexus.” Journal of Environmental Psychology, 25(3), 335–346.

Ryan, M, Da, Scott, Reeves, C., Bate, A., van Teijlingen, Er, Russell, E., Napper, M., and Robb, C. (2001). “Eliciting public preferences for healthcare: a systematic review of techniques.” Health technology assessment (Winchester, England), 5(5), 1–186.

Saaty, R. W. (1987). “The analytic heirarchy process - what it is and how it is used.” Mathematical Modeling, 9(3–5), 161–176.

Saaty, T. L. (1977). “A Scaling Method for Priorities in Hierarchical Structures.” Journal of Mathematical Psychology, 15, 234–281.

Saaty, T. L. (2008). “Decision making with the analytic hierarchy process.” International Journal of Services Sciences, 1(1), 83–98.

Sabogal, R. I., Medlin, E., Aquino, G., and Gelting, R. J. (2014). “Sustainability of water, sanitation and hygiene interventions in Central America.” Journal of Water, Sanitation and Hygiene for Development, 4(1), 89–99.

166 Sakthivel, S. R., Seshadri, A., Rahman, A., and Chariar, V. M. (2014). “Standardisation of Design and Maintenance of DEWATS Plants in India.” India Institute of Technology Technical Paper Series.

Saldana, J. (2009). The Coding Manual for Qualitative Researchers. Sage, London.

Sansom, K. (2011). “Complementary roles? NGO-Government relations for community-based sanitation in South Asia.” Public Administration and Development, 31(4), 282–293.

Santoyo-Castelazo, E., and Azapagic, A. (2014). “Sustainability assessment of energy systems: integrating environmental, economic and social aspects.” Journal of Cleaner Production, 80, 119–138.

Sara, J., and Katz, T. (1998). Making Rural Water Supply Sustainable: Report on the Impact of Project Rules. Water and Sanitation Program, World Bank, Washington, D.C.

Scheyvens, R., and Storey, D. (Eds.). (2003). Development Fieldwork: A Practical Guide. Sage Publications, London.

Schmitt, R. J. P., Morgenroth, E., and Larsen, T. A. (2017). “Robust planning of sanitation services in urban informal settlements: An analytical framework.” Water Research, 110, 297–312.

Schoen, M. E., Xue, X., Wood, A., Hawkins, T. R., Garland, J., and Ashbolt, N. J. (2017). “Cost, energy, global warming, eutrophication and local human health impacts of community water and sanitation service options.” Water Research, 109, 186–195.

Schwartz, S., and Rubel, T. (2005). “Sex differences in value priorities: Cross-cultural and multimethod studies.” Journal of Personality and Social Psychology, 89(6), 1010–1028.

Schweitzer, R., Grayson, C., and Lockwood, H. (2014). Mapping of Water, Sanitation, and Hygiene Sustainability Tools. Working Paper 10, IRC/Aguaconsult/Triple-S. 2014. Available online: http://www. ircwash. org/sites/default/files/triple- s_wp10mappingofwashsustainabilitytools. pdf (accessed on 7 March 2017).

Schwemlein, S., Cronk, R., and Bartram, J. (2016). “Indicators for Monitoring Water, Sanitation, and Hygiene: A Systematic Review of Indicator Selection Methods.” International Journal of Environmental Research and Public Health, 13(3), 333.

Scope. (2012). Ecological Sanitation in Trichy, India. Scope Trichy, Tamil Nadu, India.

Seppälä, O. T. (2002). “Effective water and sanitation policy reform implementation: need for systemic approach and stakeholder participation.” Water Policy, 4(4), 367–388.

167 Seymour, Z. (2014). “Sanitation in developing countries: a systematic review of user preferences and motivations.” Journal of Water, Sanitation and Hygiene for Development, 4(4), 681– 691.

Singhirunnusorn, W., and Stenstrom, M. K. (2009). “Appropriate wastewater treatment systems for developing countries: criteria and indictor assessment in Thailand.” Water Science and Technology, 59(9), 1873–1884.

Skat Foundation. (2013). “Technology Applicability Framework (TAF) Manual.” WASHTech, European Commission’s 7th Framework Programme in Africa.

Skat Foundation, and Rural Water Supply Network. (2018). “[TAF] Technology Applicability Framework.” [TAF] Technology Applicability Framework, (Feb. 12, 2019).

Sperling, J., Romero-Lankao, P., and Beig, G. (2016). “Exploring citizen infrastructure and environmental priorities in Mumbai, India.” Environmental Science & Policy, 60, 19–27.

Starkl, M., Brunner, N., Feil, M., and Hauser, A. (2015). “Ensuring Sustainability of Non- Networked Sanitation Technologies: An Approach to Standardization.” Environmental Science & Technology, 49(11), 6411–6418.

Starkl, M., Brunner, N., López, E., and Martínez-Ruiz, J. L. (2013a). “A planning-oriented sustainability assessment framework for peri-urban water management in developing countries.” Water Research, 47(20), 7175–7183.

Starkl, M., Brunner, N., and Stenström, T.-A. (2013b). “Why Do Water and Sanitation Systems for the Poor Still Fail? Policy Analysis in Economically Advanced Developing Countries.” Environmental Science & Technology, 47, 6102–6110.

Stellacci, P., Liberti, L., Notarnicola, M., and Haas, C. N. (2010). “Hygienic sustainability of site location of wastewater treatment plants.” Desalination, 253(1–3), 51–56.

Stewart, D. W., and Shamdasani, P. N. (2014). Focus Groups: Theory and Practice. SAGE Publications, Thousand Oaks, California.

Stokes, D., and Bergin, R. (2006). “Methodology or ‘methodolatry’? An evaluation of focus groups and depth interviews.” Qualitative Market Research: An International Journal, 9(1), 26–37.

Sujaritpong, S., and Nitivattananon, V. (2009). “Factors influencing wastewater management performance: Case study of housing estates in suburban Bangkok, Thailand.” Journal of Environmental Management, 90(1), 455–465.

168 Tilley, E., Strande, L., Lüthi, C., Mosler, H.-J., Udert, K. M., Gebauer, H., and Hering, J. G. (2014a). “Looking beyond Technology: An Integrated Approach to Water, Sanitation and Hygiene in Low Income Countries.” Environmental Science & Technology, 48(17), 9965– 9970.

Tilley, E., Ulrich, L., Lüthi, C., Reymond, P., and Zurbrügg, C. (2014b). Compendium of Sanitation Systems and Technologies. Eawag.

Tofail, F., Fernald, L. C., Das, K. K., Rahman, M., Ahmed, T., Jannat, K. K., Unicomb, L., Arnold, B. F., Ashraf, S., Winch, P. J., Kariger, P., Stewart, C. P., Colford, J. M., and Luby, S. P. (2018). “Effect of water quality, sanitation, hand washing, and nutritional interventions on child development in rural Bangladesh (WASH Benefits Bangladesh): a cluster- randomised controlled trial.” The Lancet Child & Adolescent Health, 2(4), 255–268.

Törnqvist, R., Norström, A., Kärrman, E., and Malmqvist, P.-A. (2008). “A framework for planning of sustainable water and sanitation systems in peri-urban areas.” Water Science & Technology, 58(3), 563–570.

Trimmer, J. T., Cusick, R. D., and Guest, J. S. (2017). “Amplifying Progress toward Multiple Development Goals through Resource Recovery from Sanitation.” Environmental Science & Technology, 51(18), 10765–10776.

Uddin, S. M. N., Muhandiki, V. S., Sakai, A., Al Mamun, A., and Hridi, S. M. (2014). “Socio- cultural acceptance of appropriate technology: Identifying and prioritizing barriers for widespread use of the urine diversion toilets in rural Muslim communities of Bangladesh.” Technology in Society, 38, 32–39.

UN-HABITAT. (2010). Chapter 1: Development Context and the Millennium Agenda. The Challenge of Slums: Global Report on Human Settlements 2003, United Nations, Geneva, Switzerland.

UNICEF. (2012). Sustainability Checks: the UNICEF Experience in Eastern and Southern Africa. UNICEF.

UNICEF. (2014). Madagascar WASH Sector Sustainability Check. UNICEF, Madagascar.

UNICEF. (2017). Sustainability Checks: Guidance to Design and Implement Sustainability Monitoring in WASH. UNICEF, 44.

United Nations. (2015a). “Transforming our world: the 2030 Agenda for Sustainable Development.” Sustainable Development Knowledge Platform, (Nov. 19, 2015).

169 United Nations. (2015b). Sustainable Development Goals. United Nations.

USAID. (2014). Local Systems: A Framework for Supporting Sustainable Development. USAID, Washington, D.C.

Van Minh, H., and Hung, N. V. (2011). “Economic Aspects of Sanitation in Developing Countries.” Environmental Health Insights, 5, 63–70.

Venkataramanan, V., Crocker, J., Karon, A., and Bartram, J. (2018). “Community-Led Total Sanitation: A Mixed-Methods Systematic Review of Evidence and Its Quality.” Environmental Health Perspectives, 126(2), 026001.

Vleuten-Balkema, A. J. van der. (2003). “Sustainable wastewater treatment: developing a methodology and selecting promising systems.” Dissertation, Eindhoven University, Eindhoven.

Walekhwa, P. N., Lars, D., and Mugisha, J. (2014). “Economic viability of biogas energy production from family-sized digesters in Uganda.” Biomass and Bioenergy, 70, 26–39.

Walekhwa, P. N., Mugisha, J., and Drake, L. (2009). “Biogas energy from family-sized digesters in Uganda: Critical factors and policy implications.” Energy Policy, 37(7), 2754–2762.

Wang, C., and Burris, M. A. (1997). “Photovoice: Concept, Methodology, and Use for Participatory Needs Assessment.” Health, Education & Behavior, 24(3), 369–387.

Wang, C. C., Morrel-Samuels, S., Hutchison, P. M., Bell, L., and Pestronk, R. M. (2004). “Flint photovoice: Community building among youths, adults, and policymakers.” American journal of public health, 94(6), 911–913.

Wang, C. C., and Redwood-Jones, Y. A. (2001). “Photovoice ethics: Perspectives from Flint photovoice.” Health education & behavior, 28(5), 560–572.

Wang, J.-J., Jing, Y.-Y., Zhang, C.-F., and Zhao, J.-H. (2009). “Review on multi-criteria decision analysis aid in sustainable energy decision-making.” Renewable and Sustainable Energy Reviews, 13(9), 2263–2278.

Wankhade, K. (2015). “Urban sanitation in India: key shifts in the national policy frame.” Environment and Urbanization, 27(2), 555–572.

WaterAid. (2011a). Sustainability framework. Framework, WaterAid, London, UK.

WaterAid. (2011b). Sanitation Framework. WaterAid, London, UK.

170 Watkins, R., West Meiers, M., and Visser, Y. (2012). A Guide to Assessing Needs: Essential Tools for Collecting Information, Making Decisions, and Achieving Development Results. The World Bank, Washington, D.C.

WCED. (1987). Report of the World Commission on Environment and Development: Our Common Future. World Commission on Environment and Development, Oxford.

Wegelin-Schuringa, M. (2000). “Public awareness and mobilisation for ecosanitation.” IRC International Water and Sanitation Center, Delft, Netherlands, IRC International Water and Sanitation Center, Delft, Netherlands, Bonn, Germany.

Whittington, D., Jeuland, M., Barker, K., and Yuen, Y. (2012). “Setting Priorities, Targeting Subsidies among Water, Sanitation, and Preventive Health Interventions in Developing Countries.” World Development, 40(8), 1546–1568.

WHO. (2004). “Rapid Needs Assessment for Water, Sanitation and Hygiene.”

WHO, and UNICEF. (2015). Progress on sanitation and drinking water: 2015 Update and MDG Report. Joint Monitoring Programme, World Health Organization, Geneva.

WHO, and UNICEF. (2017). Progress on Drinking Water, Sanitation and Hygiene: Update and SDG Baselines. Joint Monitoring Program, Geneva.

Wongburi, P., and Park, J. K. (2018). “Decision making tools for selecting sustainable wastewater treatment technologies in Thailand.” IOP Conference Series: Earth and Environmental Science, 150, 012013.

Wood, B. T., Dougill, A. J., Quinn, C. H., and Stringer, L. C. (2016). “Exploring Power and Procedural Justice Within Climate Compatible Development Project Design: Whose Priorities Are Being Considered?” The Journal of Environment & Development, 25(4), 363–395.

World Health Organization. (2012). A toolkit for monitoring and evaluating household water treatment and safe storage programmes. World Health Organization, Geneva.

WSP. (2012). “Economic Assessment of Sanitation Interventions in Cambodia.” The World Bank.

WSP. (2017). “Economics of Sanitation Initiative.” The World Bank, (Nov. 9, 2017).

Yin, R. K. (2009). Case study research: design and methods. Applied social research methods, Sage Publications, Los Angeles, Calif.

171 Appendix A: Glossary of Terms

Table A1. Glossary of terms used throughout the dissertation.

Term Definition Abstract Priorities Abstract priorities are not visual and usually require a social or organizational intervention, such as women’s empowerment or government support Added Resource In the Added Resource Recovery Scenario, all systems are Recovery Scenario functional (as in the Intended Design Scenario), and technologies are changed such that they have full resource recovery capabilities, i.e., the ability to produce and recover biogas, water, and compost AHP Group A focus group of community members that used the Analytical Hierarchy Process (AHP) to rank community and sanitation priorities. This AHP group is different from focus groups, which were conducted to identify, but not rank community and sanitation priorities Case Used interchangeably with “community”, a case refers to one of the 20 communities in which an in-depth case study was conducted Causal Conditions Factors that are hypothesized to influence outcomes of interest, in QCA Community Priorities Community needs and values for general life such as education, jobs, and drinking water Conventional Systems Conventional sanitation systems are designed solely to contain and treat waste and do not produce or recover energy, nutrients, or water Current Scenario The current scenario reflects the currently implemented technologies and present status of the systems. No changes are made to system designs and functionality DEWATS Decentralized wastewater treatment system that consists of either a biogas digester or a conventional settling tank, a baffled reactor, and a planted gravel filter EcoSan Ecological sanitation is a type of urine diverting dry toilet. In this research, EcoSan toilets are implemented as shared community toilets followed by a gravel filter (for anal cleansing water) and have the ability to produce compost, recover urine, and reuse water for irrigation Failure A failed sanitation system does not meet one or more of the criteria for success (use, maintenance, or performance) Improved Septic Tank An improved septic tank is a settling tank with two or more baffles to increase retention time and settling of solids Indicator A measurable factor that may influence or reflect sustainability proposed in a sustainability framework

172 Intended Design Scenario The Intended Design Scenario reflects the currently implemented technologies, but all systems are changed to function according to their intended designs Maintenance Adequate maintenance occurs when more than 90 percent of required maintenance tasks are performed correctly and on time Metric The specific, measurable quantity used to measure an indicator in a sustainability framework Pathways In QCA, a pathway is a combination of multiple factors (causal conditions) that together produce the outcome of interest Performance System performance occurs when the treated effluent meets or exceeds local regulations. Regulations in India are for pH, BOD, and COD and depend on the end use of water, such as discharge to a municipal sewer, irrigation, surface water, or coastal waters Physical Priorities Physical priorities can be seen, captured visually, and addressed through a specific infrastructure solution, such as biogas production, which could be addressed by the implementation of a well-maintained biogas digester Priority A priority is a problem, need, or value of significant importance to a community Priority Category Priority category refers to the broad group of priorities: community or sanitation Priority Type Priority type refers to the specific manifestation of the priority: physical or abstract Resource Recovery The production and recovery of energy, nutrients, and/or water in wastewater treatment processes for beneficial use Resource Recovery Resource recovery systems are sanitation systems that are Systems designed to produce and recover energy (biogas), nutrients (compost or urine), or water for beneficial use Sanitation Priorities Community needs, values, and preferences specific to sanitation systems such as cleanliness, safety, and low cost SPISF Single pass intermittent sand filter that consists of a common settling tank and a sand and gravel filter Success A successful sanitation system is one that is continually used and maintained, and performs to local regulations Sustainability Impacts The social, economic, and environmental impacts of sanitation systems Sustainable Sanitation A sustainable sanitation system is a successful sanitation system that minimizes negative social, economic, and environmental impacts Use Continual use occurs when more than 75 percent of community members (in the system’s target population) use the system correctly, daily, and exclusively (i.e., no open defecation)

173 Appendix B: Case Summaries

Case 1: Successful Conventional System in Tamil Nadu Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 800 Implemented By: Out-of-State Indian NGO with prior sanitation experience and no prior community experience O&M Manager: Male Operator from the community Outcome: Successful (100% Used, 91% Maintained, 3 of 3 Regulations Met) Case Summary:  NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality;  NGO conducted a limited priority assessment that did not include sanitation-specific priorities;  Community was informed but not involved in planning or decision-making;  Behavior change education activities occurred before and after implementation and included sanitation seminars and mapping to identify open defecation sites and effects;  Municipality was involved in planning, decided the site location, and approved the treatment technology;  A male operator from community performs daily O&M;  Sanitary engineers from municipality visit system bi-weekly and provide technical assistance; municipality also pays for all O&M costs, which is included in their annual budget.

Case 2: Successful Resource Recovery System in Karnataka Technology: Resource Recovery (DEWATS with Biogas Digester; Intended to recover biogas and water) User Connection: Individual Household Toilets Number of Intended Users: 850 Implemented By: Local Indian NGO with prior sanitation experience and no prior experience in community O&M Manager: Implementing Organization Engineers Outcome: Successful (100% Used, 90% Maintained, 3 of 3 Regulations Met) Case Summary:  NGO partnered with community leader to engage community and plan project to stop open defecation and improve environmental health;  NGO did not conduct any priority assessment;  In planning, community members attended meetings and contributed to some design decisions;  Limited behavior change education occurred, focused only on the benefits of sanitation;  Municipality gave permission for the project but was not otherwise involved. NGO did not attempt to engage municipality and instead engaged community;  O&M is performed by NGO because system is right across the street from the organization's office and they use it as a showpiece for prospective customers;  NGO pays for O&M;  Water is reused to irrigate a small vegetable farm, and the vegetables are distributed to community; biogas is used as cooking fuel for community’s kindergarten.

174

Case 3: Successful Conventional System in Tamil Nadu Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 800 Implemented By: Out-of-State Indian NGO with prior sanitation experience and no prior community experience O&M Manager: Male Operator Outcome: Successful (84% Used, 90% Maintained, 2 of 2 Regulations Met) Case Summary:  NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality;  NGO did not conduct any priority assessment;  Community was informed but not involved in planning or decision-making;  Behavior change education activities occurred before and after implementation and included sanitation seminars, sanitation games, and discussions of healthy behaviors;  Municipality was involved in the planning of system, decided the site location, and approved the treatment technology;  A male operator maintains system, primarily turning the pumps on/off daily and removing sewer blockages;  Municipality provides O&M technical and financial assistance;  Community generates income for O&M costs from annual fees collected by the panchayat (community leadership organization).

Case 4: Failed Conventional System in Karnataka Technology: Conventional (Baffled Settling Tank + Gravel Filter) User Connection: Shared Community Toilets Number of Intended Users: 1,010 Implemented By: International NGO with sanitation experience and no prior community experience O&M Manager: None Outcome: Failed (0% Used, 0% Maintained, 0 of 3 Regulations Met) Case Summary:  NGO without prior experience in India implemented system to end open defecation and protect surface water;  NGO conducted a limited priority assessment that did not include sanitation-specific priorities;  Community was not involved in planning and NGO efforts in planning to create a community sanitation committee were unsuccessful;  Limited behavior change education occurred focused only on the benefits of sanitation;  Municipality was uninvolved in planning, and formal system handover was never given to municipality;  A caretaker was appointed to clean the toilets but not to perform O&M. After one year, community struggled with the toilet water supply and cleaning stopped;  Neither municipality nor NGO provided ongoing O&M or financial assistance for O&M;  Toilets were closed because of O&M problems, and people in community returned to open defecation.

175 Case 5: Failed Resource Recovery System in Tamil Nadu Technology: Resource Recovery (Baffled Settling Tank + Gravel Filter; Intended to recover water) User Connection: Shared Community Toilets Number of Intended Users: 800 Implemented By: Local Indian NGO without sanitation experience and no prior community experience O&M Manager: None Outcome: Failed (22% Used, 0% Maintained, 1 of 2 Regulations Met) Case Summary:  NGO received international donor funding to implement a prefabricated system and end open defecation;  NGO did not conduct any priority assessment;  Community was not involved in planning. NGO spoke with community leader about a possible sanitation project, but unclear communication meant the leader was unaware the project was going to happen. NGO made no other attempts to engage community in planning;  Behavior change education did not occur;  Municipality lacked resources to be involved in planning;  No operator was appointed;  Community receives no technical or financial assistance and is unable to repair system;  The toilets are extremely unclean, and most of community goes for open defecation;  Treated water was supposed to be used for irrigation, but instead discharges to a drainage canal.

Case 6: Failed Resource Recovery System in Karnataka Technology: Resource Recovery (DEWATS with Biogas Digester; Intended to recover biogas and water) User Connection: Individual Household Toilets Number of Intended Users: 1,015 Implemented By: Local Indian NGO without prior sanitation experience and with prior experience in community O&M Manager: Women’s Self-Help Group (WSHG) Outcome: Failed (48% Used, 7% Maintained, 0 of 3 Regulations Met) Case Summary:  NGO implemented system to upgrade a failed system;  NGO conducted a limited priority assessment that did not include sanitation-specific priorities;  In planning, community members were only informed of the project and did not contribute to decision- making, and NGO made no other attempts to engage community in planning;  Behavior change education occurred pre-implementation and included sanitation seminars, theatrical plays to promote positive sanitation behaviors, demonstrations of healthy behaviors, and educational pamphlets;  Municipality lacked resources to be involved in planning;  A WSHG was primarily responsible for O&M, but WSHG experienced internal conflict and lacked the technical and financial capacity to continue O&M;  Income was generated from monthly household fees, but many community members could not pay or refused to pay; community received no municipal support and limited technical assistance from NGO;  When community began experiencing water scarcity, O&M problems increased, and most community members returned to open defecation;  Biogas was used for lights in the toilets, and water irrigated plants at the nearby park.

176 Case 7: Successful Conventional System in Tamil Nadu Technology: Conventional (Settling Tank + Single-Pass Intermittent Sand Filter) User Connection: Individual Household Toilets Number of Users: 820 Implemented By: Out-of-State Indian NGO with prior sanitation experience and no prior community experience O&M Manager: Male Operator Outcome: Successful (95% Used,100% Maintained, 2 of 2 Regulations Met) Case Summary:  NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality;  NGO did not conduct any priority assessment;  Community was informed but not involved in planning, and NGO made no other attempts to engage community in planning;  Behavior change education occurred before and after implementation and included sanitation seminars, games to identify unsanitary behaviors, and community mapping to identify open defecation sites and effects;  Municipality was involved in the planning of system, decided the site location, and approved the treatment technology;  A dedicated male operator performs daily O&M. Community attributes their system’s success to the operator’s dedication;  Municipality pays the operator’s salary and provides technical assistance when larger problems occur.

Case 8: Successful Conventional System in Tamil Nadu Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 875 Implemented By: Out-of-State Indian NGO with prior sanitation experience and no prior community experience O&M Manager: Municipality Engineers Outcome: Successful (90% Used, 100% Maintained, 2 of 2 Regulations Met) Case Summary:  NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality;  NGO conducted a limited priority assessment that did not include sanitation-specific priorities;  Community was informed but not involved in planning. NGO met with community leader to discuss the project, but he did not contribute further to decision-making;  Behavior change education occurred before and after implementation and included sanitation seminars, educational pamphlets, and discussions of healthy behaviors;  Municipality was involved in the planning of system, decided the site location, and approved the treatment technology; received formal handover and have collected monitoring data;  Municipality has a highly motivated sanitary inspector, who oversees engineers who perform O&M;  Municipality pays the operators and large O&M costs;  Community generates income for O&M costs from annual fees collected by the community leader.

177 Case 9: Failed Resource Recovery System in Karnataka Technology: Resource Recovery (DEWATS; Intended to recover water and compost) User Connection: Individual Household Toilets Number of Intended Users: 800 Implemented By: Local Indian NGO without prior sanitation experience and with prior experience in community O&M Manager: WSHG Outcome: Failed (85% Used, 20% Maintained, 0 of 3 Regulations Met) Case Summary:  NGO implemented system to end open defecation;  NGO conducted a limited priority assessment that did not include sanitation-specific priorities;  In planning, community members attended meetings and selected the site;  Behavior change education occurred pre-implementation and included sanitation seminars, discussions of social norms, demonstrations of healthy behaviors, and theatrical plays to promote positive sanitation behaviors;  Municipality was uninvolved in planning because NGO did not engage them;  Income was generated from monthly household fees; community received limited technical assistance from NGO and no municipal support;  A WSHG oversaw O&M, but lacked financial and technical capacity to overcome system damage and municipal opposition;  Water was used for groundwater recharge and irrigation of a small garden, and vegetables were sold; sludge was land applied on nearby farms;  Municipality constructed 120 new households nearby and connected them to system without making modifications. System is over-capacity and completely flooded; municipality has no plans to repair system.

Case 10: Failed Conventional System in Tamil Nadu Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 810 Implemented By: Out-of-State Indian NGO without prior sanitation experience and no prior community experience O&M Manager: Male Operator Outcome: Failed (6% Used, 0% Maintained, 1 of 2 Regulations Met) Case Summary:  NGO implemented system to replace damaged septic tanks;  NGO did not conduct any priority assessment;  Community learned of the project during construction, and NGO made no attempts to engage community in planning;  Behavior change education did not occur;  Municipality lacked the resources to be involved in planning;  A male community member oversaw O&M, but once the pumps failed, he was unable to perform O&M;  Community receives no O&M financial or technical assistance;  Community has a very strong, negative view of system, indicating that it has “destroyed” their community because of persistent problems system has caused (blockages, smells).

178 Case 11: Successful Resource Recovery System in Tamil Nadu Technology: Resource Recovery (DEWATS with Biogas Digester; Intended to recover biogas and water) User Connection: Shared Community Toilets Number of Intended Users: 1,006 Implemented By: Local Indian NGO with prior sanitation experience and prior experience in community O&M Manager: WSHG Outcome: Successful (100% Used, 92% Maintained, 3 of 3 Regulations Met) Case Summary:  NGO implemented system to replace damaged community toilets;  NGO conducted an extensive priority assessment that included sanitation-specific priorities;  In planning, community members attended meetings, selected site, determined number of toilets, and designed farm;  Behavior change education occurred pre-implementation and included sanitation seminars, educational pamphlets, and community mapping to identify open defecation sites;  Municipality lacked resources to be involved in planning but have since contributed to O&M;  A WSHG receives technical and financial assistance from municipality;  Income is generated from pay-per-use fees and sale of vegetables from the farm;  Treated effluent is used for irrigation of small farm nearby; system has a biogas digester which was formerly functional and used for cooking fuel for a tea shop run by WSHG. Biogas is no longer in use due to broken pipes that transported the gas from the digester to the tea shop.

Case 12: Failed Resource Recovery System in Karnataka Technology: Resource Recovery (DEWATS with Biogas Digester; Intended to recover biogas) User Connection: Shared Community Toilets Number of Intended Users: 802 Implemented By: Local Indian NGO without prior sanitation experience and with prior experience in community O&M Manager: Male Operator Outcome: Failed (32% Used, 7% Maintained, 0 of 3 Regulations Met) Case Summary:  NGO implemented system to end open defecation;  NGO did not conduct any priority assessment;  Community was only informed of the project but was not actively involved in planning;  Limited behavior change education occurred focused only on the benefits of sanitation;  Municipality lacked resources to be involved in planning;  A male community member was appointed for cleaning and O&M but only performs occasional cleaning and no O&M;  Community receives no O&M technical or financial assistance;  Income is generated from pay-per-use fees and is used for micro-loans for community, so there are insufficient O&M funds;  Community never used the biogas produced from system because of bad smells;  System has been damaged by vandalism and community lacks the finances to pay for repairs;  Municipality has threatened to revoke the land lease because of the poor O&M but has not yet acted.

179 Case 13: Successful Resource Recovery System in Tamil Nadu Technology: Resource Recovery (DEWATS with Biogas Digester; Intended to recover biogas, water, and compost) User Connection: Shared Community Toilets Number of Intended Users: 860 Implemented By: Local Indian NGO with prior sanitation experience and prior community experience O&M Manager: Male Operator Outcome: Successful (100% Used, 100% Maintained, 3 of 3 Regulations Met) Case Summary:  NGO implemented system as part of an agreement with municipality to add composting to the city’s solid waste management center;  NGO did not conduct any priority assessment;  Community members attended planning meetings;  Behavior change education initiatives were led by community members before and after implementation and included sanitation seminars, discussions of social norms, sanitation games, and community mapping to identify open defecation sites;  Municipality commissioned the project and attended all meetings;  A male community member performs O&M;  Municipality provides technical and financial assistance;  Biogas is used by two households nearby system for cooking fuel. Water is used for irrigation of banana plants. Sludge is mixed with organic waste from municipality and used to produce compost, which is sold for agriculture, generating more than 100% of O&M costs.

Case 14: Successful Resource Recovery System in Tamil Nadu Technology: Resource Recovery (Community Ecological Sanitation; Intended to recover water and compost) User Connection: Shared Community Toilets Number of Intended Users: 845 Implemented By: Local Indian NGO with prior sanitation experience and no prior community experience O&M Manager: Male Operator Outcome: Successful (100% Used, 100% Maintained, 3 of 3 Regulations Met) Case Summary:  NGO implemented system as a pilot project for ecological sanitation technology with international donor funding;  NGO did not conduct any priority assessment;  In planning, community members attended meetings, selected the site, and approved the final design;  Behavior change education occurred pre-implementation and included sanitation seminars and community mapping to identify open defecation sites;  Municipality lacked the resources to be involved in planning but gave permission for the project;  A male community member with a sanitary engineering degree performs O&M;  Community receives ongoing technical and financial assistance from NGO;  System generates income from resource recovery sales;  Fecal waste is mixed with ash to produce compost; anal cleansing water is treated in a gravel filter and used for irrigation for a small nearby farm; urine is collected, mixed with water, and used as fertilizer.

180

Case 15: Successful Conventional System in Tamil Nadu Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 850 Implemented By: Out-of-State Indian NGO with prior sanitation experience and no prior community experience O&M Manager: Male Operator Outcome: Successful (91% Used, 90% Maintained, 2 of 2 Regulations Met) Case Summary:  NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality;  NGO conducted a limited priority assessment that did not include sanitation-specific priorities;  Community was informed but not involved in planning because many community members lacked time for planning;  Behavior change education activities were conducted before and after implementation and included sanitation seminars, and demonstrations of healthy behaviors;  Municipality was involved in the planning of system, decided the site location, and approved the treatment technology;  A male operator performs O&M;  Municipality provides technical and financial assistance;  Community generates income for O&M costs from annual fees collected by the panchayat (community leadership organization).

Case 16: Failed Conventional System in Tamil Nadu Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 805 Implemented By: Out-of-State Indian NGO without prior sanitation experience and no prior community experience O&M Manager: None Outcome: Failed (31% Used, 3% Maintained, 1 of 2 Regulations Met) Case Summary:  NGO implemented system to replace damaged septic tanks;  NGO did not conduct any priority assessment;  Community learned of the project during construction, and NGO did not make any attempts to engage community in planning;  Behavior change education did not occur;  NGO did not handover system to municipality, so municipality refuses to assist with O&M;  An operator was not appointed, so no O&M has been performed;  Municipality also lacks the human and financial resources to provide support for system;  Community members state that the construction quality is poor because the contractors are only interested in profit and rushed to complete the job, and community was not able to make meaningful changes during construction;  Most of community practices open defecation.

181

Case 17: Failed Conventional System in Karnataka Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 1,010 Implemented By: International NGO with sanitation experience and no prior community experience O&M Manager: None Outcome: Failed (88% Used, 11% Maintained, 0 of 3 Regulations Met) Case Summary:  NGO implemented system to end open defecation and protect surface water;  NGO conducted a limited priority assessment that did not include sanitation-specific priorities;  Community was not involved in planning, and efforts to create a community sanitation committee were unsuccessful;  Limited behavior change education occurred focused only on the benefits of sanitation;  Formal handover never was given to municipality, who was also uninvolved in planning, so they refused to provide O&M technical and financial assistance;  No operator was appointed so no O&M was performed;  NGO provided no technical or financial assistance;  System was never desludged, and the pumps broke very early on in the few short months of operation;  Toilet wastewater now discharges directly to the nearby lake.

Case 18: Successful Resource Recovery System in Tamil Nadu Technology: Resource Recovery (DEWATS with Biogas Digester; Intended to recover biogas and water) User Connection: Shared Community Toilets Number of Intended Users: 1,000 Implemented By: Local Indian NGO with prior sanitation experience and prior community experience O&M Manager: WSHG Outcome: Successful (100% Used, 100% Maintained, 3 of 3 Regulations Met) Case Summary:  NGO implemented system to replace damaged community toilets and improve environmental health;  NGO conducted an extensive priority assessment that included sanitation-specific priorities;  In planning, community members attended meetings, selected the site, requested child-friendly toilets, and helped design community kitchen;  Behavior change education was led by community members before and after implementation and included sanitation seminars, discussions of healthy behaviors, discussions of social norms, theatrical plays to promote positive sanitation behaviors, and community mapping to identify open defecation sites;  Municipality lacked resources to be involved in planning but is now supportive;  A WSHG oversaw O&M of system;  Municipality pays for major O&M costs and performs O&M for the biogas digester;  WSHG generates income for O&M from pay-per-use fees and employs over 30 women monthly;  Biogas is used for a free community kitchen with 20 cookstoves;  WSHG generated enough income from toilets to construct water storage tank to improve community’s water supply and to construct community hall.

182

Case 19: Failed Conventional System in Tamil Nadu Technology: Conventional (DEWATS) User Connection: Individual Household Toilets Number of Intended Users: 1,015 Implemented By: International NGO without sanitation experience and no prior community experience O&M Manager: Male Operator Outcome: Failed (82% Used, 60% Maintained, 1 of 3 Regulations Met) Case Summary:  NGO implemented system to end open defecation;  NGO did not conduct any priority assessment;  Community learned of the project during construction, and NGO did not make any attempts to engage community in planning;  Behavior change education did not occur;  Municipality was not involved in planning. Case 18 technically falls under village panchayat authority, although neighboring communities are under municipal authority. There is a lack of clarity of which government agency is responsible for community, so community receives no assistance;  An unskilled male operator runs the pumps, and some households desludge the tanks on an emergency basis, when tanks start to overflow.

Case 20: Failed Resource Recovery System in Karnataka Technology: Resource Recovery (DEWATS with Biogas Digester; Intended to recover biogas and water) User Connection: Shared Community Toilets Number of Intended Users: 850 Implemented By: Local Indian NGO with prior sanitation experience and prior community experience O&M Manager: WSHG Outcome: Failed (0% Used, 0% Maintained, 0 of 3 Regulations Met) Case Summary:  NGO implemented system to end open defecation;  NGO conducted an extensive priority assessment that included sanitation-specific priorities;  In planning, community members attended meetings, went on exposure visits, and the WSHG was actively involved in system design and site selection;  Behavior change education occurred post-implementation and included sanitation seminars, sanitation games, demonstrations of healthy behaviors, and theatrical plays to promote positive sanitation behaviors;  Municipality was not involved in planning but agreed to lease the land for system;  A WSHG who received training and had technical knowledge oversaw O&M;  WSHG generated income from pay-per-use fees and resource recovery. Biogas was used to heat water which was sold for bathing. Treated water was sold for construction and sludge was sold for agriculture application;  Municipality revoked the land lease and appointed their own employees to manage system and its revenue. Municipality failed to perform O&M and eventually locked toilets.

183

Appendix C: Interview and Survey Questionnaire Scripts

Overview: Part 1: Semi-structured Interview, Photovoice, and Focus Group Scripts

Table C1. Summary of interview, photovoice, and focus group scripts, durations, and topics. Interview Participants Interview Expected Estimated Number of Topic Duration Participants (per interview type per community system) CM1 Community Community 30 10 Members (users of Priorities minutes the system) CM2 Community Photovoice 24 hours 8 Members (Community Priorities) CM3 Community Focus Group 75 8-12 Members (Community minutes Priorities) CM4 Community Existing 1 hour 10 (can be redundant Members conditions, with CM1) outcome conditions O1 Operators Existing 1 hour All – ideally 2-3 (people maintaining conditions, the system— outcome different from user) conditions IO1 Implementing Existing 1 hour 2-3 Organizations conditions, (persons who work outcome for org. who were conditions involved in project) M1 Managers Existing 1 hour 1-2 Conditions G1 Local Government Existing 1 hour 1-2 Conditions

Some interview questions were adapted from the Small-Scale Sanitation Scaling-Up (4S) Project managed by Eawag-Sandec and funded by the Bill & Melinda Gates Foundation.

Citation: Eawag-Sandec. (2017). “4S: Small-Scale Sanitation Scaling-Up.” Eawag, (Nov. 4, 2017).

Part 2: Sustainability Survey Scripts 1.0 Questions for Community Members 2.0 Questions for Municipalities

184

C1: Semi-structured Interview, Photovoice, and Focus Group Scripts Interview CM1: Identifying Community & Sanitation Priorities Participants: Community Members

1. Please complete the following demographic information: a. Gender: Male Female b. Age: 18-24 25-34 35-44 45-54 55-64 Over 65 2. How long have you lived in this community? 3. What is your occupation? 4. Do you hold a leadership position or participate in any community committees such as a water or sanitation committee? a. If yes, can you describe your role and responsibilities in this position? 5. What is your home’s average monthly income? 6. What are your home’s primary expenditures? 7. What water sources do you use (tanker, city water supply, borewell)? a. How much water do you use each day? b. How often are there water cuts? c. Are you able to get water whenever you need it? 8. How many hours do you or someone in your home spend cooking each day? 9. What does your home use for cooking fuel? a. How much time do you or someone in your home spend obtaining that cooking fuel? b. How much money does your home spend on cooking fuel each week? 10. Do you have electricity in your home? a. Where does the electricity come from? (Grid, generator, solar panels, etc.) b. What do you use electricity for? (Lights, charging, etc.) 11. Does your family have a garden or farm? a. If yes, can you estimate the size of the garden or farm and list what is grown there? b. Does your family use any kind of fertilizer? c. What does your family use for fertilizer?

Overall Community Priorities/Problems 12. What is most important to your community? 13. What are some of the largest problems in your community? 14. If you had more resources, what problem would you address in your community first? Second? Third? 15. What are the top three things that you would not change (keep the same) in your community? a. What works well in your community? 16. What are some of the problems you have with infrastructure in your community? (Infrastructure can mean things like water systems, sanitation systems, roads, communication lines, buildings, electrical lines, etc.) 17. If you had more resources, what infrastructure problem would you address in your community first? Second? Third? 18. What social issues are most important? 19. If you had more resources, what social problem would you address in your community first? Second? Third? 20. What economic issues are most important? 21. If you had more resources, what economic problem would you address in your community first? Second? Third? 22. What are your community’s values?

185

23. Does your community ever have community meetings? If so, what is usually discussed in these meetings?

Community Sanitation Priorities/Problems 24. What are your community’s goals for sanitation? a. Has the implementing organization, [insert implementing organization name] achieved these goals? b. If yes, how? 25. What are your community’s goals for energy? a. Has the implementing organization, [insert implementing organization name] achieved these goals? b. If yes, how? 26. What are your community’s goals for water? (Reuse, conservation, supply) a. Has the implementing organization, [insert implementing organization name] achieved these goals? b. If yes, how? 27. What was sanitation in your community like before this sanitation system? 28. What is sanitation in your community like now? 29. What was energy (cooking fuel and electricity sources) like in your community before this sanitation (biogas) system? 30. What is energy (cooking fuel and electricity sources) like now? 31. How satisfied or unsatisfied are you with your community’s sanitation system? Very Unsatisfied Unsatisfied Neither satisfied Satisfied Very or unsatisfied Satisfied a. Why are you [insert previous question’s answer] with your community’s sanitation system? 32. What do you like about your current sanitation system? 33. What do you dislike about your community’s sanitation system? 34. What do you think is important to consider when choosing a sanitation system? 35. What benefits are important from your sanitation system? 36. Has the sanitation system met the needs of the community? a. What needs have been met by the sanitation system? b. What needs have not been met by the sanitation system? 37. How do you think the system could be improved? (Alternate question: what would you change about the sanitation system?) 38. How do you think the way the organization implemented the system could be improved? (Focus on the processes, how the organization implemented or managed things) 39. What would you keep the same about the sanitation system? 40. Can you please describe an ideal sanitation system for your community? 41. How is your system similar or different from that ideal? 42. What design features are important for toilets? 43. What design features are important for a sanitation treatment system? 44. What factors are important for using the sanitation system? 45. What factors are important for operating the sanitation system? 46. What factors are important for maintaining the sanitation system? 47. How well or poorly did the implementing organization consider and address your community’s priorities when planning the sanitation system? Very Poorly Neither Well or Well Very Poorly Poorly Well a. Why do you say the organization considered and addressed your community’s priorities [insert previous question’s answer] in planning?

186

48. You have told me that [insert general priorities here] are the most important factors in your community. Are there any priorities not on this list? 49. You have told me that [insert sanitation priorities here] are the most important factors for toilets and a treatment system in your community. Are there any priorities not on this list? 50. Did the implementing organization conduct a survey or ask questions about your community’s priorities before they did the sanitation project? Can you describe what the organization did? 51. Which priorities, needs, or problems from this list were addressed by the implementing organization in the sanitation system and planning processes? 52. Which priorities, needs, or problems from this list were not addressed by the implementing organization in the sanitation system and planning processes?

Conclusion 53. Is there anything else that is important for us to know about the sanitation system or your community’s priorities?

Photovoice CM2: Photovoice Introduction Participants: Community members

Hello and welcome. Thank you all for agreeing to join us today. This is an information session on photovoice, a method where each of you will be given a camera and asked to take pictures in response to a prompt. Photovoice is used to learn about community needs and values that are sometimes difficult to articulate or talk about in an interview.

We will check out a digital camera to each of you for the next 24 hours. Then, we will print the photographs that you took and ask you to come for an interview to discuss the photographs and your experience with photovoice. The interview will be between 30 minutes and 1 hour. Following this brief presentation, we will demonstrate how to use the cameras and answer any questions that you may have. As a thank you for participating, we would like to provide you photographs of your family. Please take 5 photos of your family and we will print them.

We are conducting a study on community priorities and how these relate to success of sanitation and energy systems. We are interested in both overall community priorities as well as priorities specific to sanitation and energy. As such, we have five questions that we would like you to consider when taking photographs over the next 24 hours. Please take as many photographs as you think are necessary to document these parts of your community, and then please write down what you would caption the pictures.

From here, the researchers will open up the information session to questions from the participants. The researchers will also demonstrate how to use the cameras and work with participants to ensure they understand how to use the cameras and do not have any further questions. The session will end with checking cameras out to each participant and handing them a sheet with the above prompts as a reminder.

Photovoice CM2: Photovoice Follow-Up Interview 1. May we show your photographs to other people?

187

2. Where was this photograph taken? 3. Can you tell me about this photograph? 4. Why did you choose to take this photograph? 5. What does this photograph have to do with sanitation (or energy)? 6. What does this photograph tell us about sanitation (or energy) in your community? 7. What else should we know about this photograph?

Focus Group CM3: Focus Groups for Identifying Community & Sanitation Priorities

Thank you for joining us today. We’re going to discuss important priorities, needs, and problems in your community. At the end of our time, we hope to have a complete list of community priorities, which are needs and issues and important aspects of the community overall. We also hope to have a complete list of sanitation priorities, which are needs, issues, and important aspects specifically related to a sanitation system.

The following questions can be used to facilitate discussions in the group, but not all questions must be asked. The focus group ends once the group agrees that all community and sanitation priorities have been mentioned.

1. What do you like about sanitation in your community? 2. What problems do you have with sanitation in your community? 3. What do you like about your energy (cooking fuel and electricity) sources in your community? 4. What problems do you have with your energy (cooking fuel and electricity) sources in your community? 5. What is important to think about when deciding to use a sanitation (biogas) system in your community? 6. What is most important to your community? 7. What are some of the largest problems in your community? 8. If you had more resources, what problem would you address in your community first? Second? Third? 9. What are the top three things that you would not change (keep the same) in your community? a. What works well in your community? 10. What are some of the problems you have with infrastructure in your community? (Infrastructure can mean things like water systems, sanitation systems, roads, communication lines, buildings, electrical lines, etc.) 11. If you had more resources, what infrastructure problem would you address in your community first? Second? Third? 12. What social issues are most important? 13. If you had more resources, what social problem would you address in your community first? Second? Third? 14. What economic issues are most important? 15. If you had more resources, what economic problem would you address in your community first? Second? Third? 16. What are your community’s values? 54. What do you like about your current sanitation system? 55. What do you dislike about your community’s sanitation system? 56. What do you think is important to consider when choosing a sanitation system? 57. What benefits are important from your sanitation system? 58. Has the sanitation system met the needs of the community? c. What needs have been met by the sanitation system? d. What needs have not been met by the sanitation system?

188

59. How do you think the system could be improved? (Alternate question: what would you change about the sanitation system?) 60. How do you think the way the organization implemented the system could be improved? (Focus on the processes, how the organization implemented or managed things) 61. What would you keep the same about the sanitation system? 62. Can you please describe an ideal sanitation system for your community? 63. How is your system similar or different from that ideal? 64. What design features are important for toilets? 65. What design features are important for a sanitation treatment system? 66. What factors are important for using the sanitation system? 67. What factors are important for operating the sanitation system? 68. What factors are important for maintaining the sanitation system?

Interview CM4: Causal Conditions & Outcomes Participants: Community members

1. Please complete the following demographic information: a. Gender: Male Female b. Age: 18-24 25-34 35-44 45-54 55-64 Over 65 2. How long have you lived in this community? 3. What is your occupation? 4. Do you hold a leadership position or participate in any community committees such as a water or sanitation committee? a. If yes, can you describe your role and responsibilities in this position? 5. What is your home’s average monthly income? 6. What are your home’s primary expenditures? 7. What water sources do you use (tanker, city water supply, borewell)? 8. What does your home use for cooking fuel? a. How much time do you or someone in your home spend obtaining that cooking fuel? b. How much money does your home spend on cooking fuel each week? 9. How many hours do you or someone in your home spend cooking each day? 10. Do you have electricity in your home? a. Where does the electricity come from? (Grid, generator, solar panels, etc.) 11. Does your family have a garden or farm? a. If yes, can you estimate the size of the garden or farm and list what is grown there? b. Does your family use any kind of fertilizer? c. What does your family use for fertilizer?

Planning and Decision-Making 12. Were you here when the toilet was built? 13. Can you please start at the very beginning of the sanitation project and walk me through everything that happened for the project and how it was accomplished? 14. Who initiated the sanitation project? 15. What were the goals for the sanitation project? Why was it initiated? What problems did the project aim to solve? 16. Who was involved with the sanitation project? a. Besides community members, implementing organization, and government, was anyone else involved in the project? 17. Can you please describe how [insert implementing organization name] completed the sanitation project in your community? 18. Can you describe how that organization contacted you and told you about the sanitation system?

189

19. In what ways (if any) did the organization include community members in the planning processes of the sanitation system? 20. What do you know about the technology used for the sanitation system? Were you familiar with this technology before it was built in your community? 21. In what ways (if any) did community members help with the design of the sanitation system? 22. For which decisions or aspects of the sanitation system was your input asked? 23. How was the cost of the system considered in decisions? 24. How were the benefits of the system considered in decisions? a. How was biogas production considered in the planning process? b. How was compost considered in the planning process? c. How was water reuse considered in the planning process? 25. How were the challenges of the system considered in decisions? 26. How satisfied or unsatisfied are you with how the implementing organization included your community in the planning of the sanitation system? Very Unsatisfied Unsatisfied Neither satisfied Satisfied Very or unsatisfied Satisfied a. Why are you [insert previous question’s answer] with how the organization included your community in planning? 27. Did the organization conduct a priority or needs assessment in your community? a. If yes: Can you please describe how the organization identified priorities/needs? 28. How well or poorly did the implementing organization consider and address your community’s priorities when planning the sanitation system? Very Poorly Neither Well or Well Very Poorly Poorly Well a. Why do you say the organization considered and addressed your community’s priorities [insert previous question’s answer] in planning? 29. Which priorities, needs, or problems were addressed by the implementing organization in the sanitation system and planning processes? 30. Which priorities, needs, or problems were not addressed by the implementing organization in the sanitation system and planning processes? 31. If you were in charge of the project, what would you do differently than the organization? 32. What would you do the same as the organization? 33. What is the most important thing for us to know about the planning process for this project?

Construction 34. Can you describe for me how the sanitation system in your community was built? 35. In what ways (if any) did the organization include community members in the construction of the sanitation system? 36. Did you pay for the construction of the system? a. If yes: How much? 37. Do you pay ongoing fees for operation and maintenance of the system? a. If yes: How much? 38. Do you pay ongoing fees for operation and maintenance of the system? a. If yes: What did you provide or how did you help? 39. How satisfied or unsatisfied are you with how the implementing organization included your community in the construction of the sanitation system? Very Unsatisfied Unsatisfied Neither satisfied Satisfied Very or unsatisfied Satisfied a. Why are you [insert previous question’s answer] with how the organization included your community in the construction?

190

40. How well or poorly did the implementing organization consider and address your community’s priorities when constructing the sanitation system? Very Poorly Neither Well or Well Very Poorly Poorly Well a. Why do you say the organization considered and addressed your community’s priorities [insert previous question’s answer] in construction? 41. What did you like about how the organization planned and built the system? 42. What do you dislike about how the organization planned and built the system?

Training 43. Did the implementing organization provide training for community members about the sanitation system? a. Can you please describe the trainings? b. What was the subject of the training? c. What kind of training? (e.g., lecture, demonstration, active participation, etc.) d. What were the training objectives or goals? e. How did the organization teach the training objectives? f. Were tests performed to assess learning objectives or check trainee understanding? g. What type of tests or assessments were performed? h. Can you please describe this the materials provided by the organization (e.g., manuals, flyers, etc.)? Do you still have them? 44. How would you judge the effectiveness of the training? Very Ineffective Neither Effective Effective Very Ineffective or Ineffective Effective a. Why did you say the training was [insert previous question’s response]?

Use & Maintenance 45. Do you use the system? a. Why or why not? b. How often do you use the sanitation system? c. How are you affected by [water scarcity at the toilets?]? d. Are there ever times when the community toilet is closed? e. How often do you use the community toilet? f. Were you here when the toilet was built? g. How are you affected by [water scarcity at the toilets?]? h. Are there ever times when the community toilet is closed? i. How often do you use the community toilet? j. How many days per month do you not use the community toilet? 46. Does your community use biogas produced by the system? a. If yes: How does your home or your community use the biogas generated from the system? i. How much biogas does your home use? ii. How many hours does it last? b. If no: if you had biogas, what would you use it for? c. What do you think about using biogas produced from the system? 47. Does your community use compost produced by the system? How? a. If yes: How does your home or your community use the compost generated from the system? i. How much compost does your home use? b. If no: if you had compost, what would you use it for? c. What do you think about using compost made with sludge from the system?

191

48. What happens to the treated water from the system? Is any of the water reused (toilet flushing, irrigation, etc.)? a. How much water is reused? What is it used for? b. If no: if your community did reuse water, what would you use it for? c. What do you think about using treated water from the system? 49. Do you help operate or maintain the system? a. If yes: Can you describe what you do? 50. How well or poorly did the implementing organization consider and address your community’s priorities in the operation of the sanitation system? Very Poorly Neither Well or Well Very Poorly Poorly Well a. Why do you say the organization considered and addressed your community’s priorities [insert previous question’s answer] in operation? 51. What do they think is important for us to know about the women's self help groups or the community-based organization that manages the toilets? 52. How was the management structure for the community-based organization or women’s self help group developed? 53. What is the motivation to spend so much time and resources managing the system?

System Performance 54. How would you rate the overall function (performance, reliability) of your community’s sanitation system? Very Poor Neither Good or Good Very Poor Poor Good a. Why is the performance of your community’s sanitation system [insert previous question’s answer]? 55. Is the system working now? 56. When is the system working? Not working? 57. If the system is not working, what are the reasons why it is not working? 58. Has the system ever had a problem? When/why did that happen? 59. Were you affected by any of these problems? How? 60. What do you do when there is a problem with the sanitation system? 61. Does the organization help with problems? a. If yes: What do they do? 62. How long does it usually take for a repair to happen?

Communication with Implementing Organization 63. How often does the implementing organization talk with you or someone from your community? a. Who typically initiates the communication? 64. What do you discuss with the organization? 65. How do you provide feedback or complaints to the organization? 66. What support or services does the organization provide in your community? 67. What support or services does the organization provide for the sanitation system? 68. When was the last time the organization came to your community?

Government Role 69. What support or services does the government provide in your community? 70. What support or services does the government provide for the sanitation system? 71. When was the last time the government came to your community? 72. How do you provide feedback or complaints to the government?

192

Conclusion 73. Is there anything else that is important for me to know about your community? 74. From the factsheet: do your community goals match these? Did you achieve these goals? 75. What were some of the other lessons learned through the sanitation project? What did you learn? 76. What are some of the things that she has learned in 15 years of working here? 77. Is there anything else that is important for us to know about the sanitation system, your role, or your community’s participation?

Interview O1: Causal Conditions & Outcomes Participants: Operators, WSHG Members

1. Please complete the following demographic information: a. Gender: Male Female b. Age: 18-24 25-34 35-44 45-54 55-64 Over 65 2. Do you live in the community? a. If yes: How long have you lived here? 3. How long have you been an operator of the sanitation system? 4. Is the operation and maintenance of the sanitation system your primary job? a. If no: what other jobs do you hold? 5. Are you paid for your operation and maintenance work? a. How much are you paid per month? 6. How were you selected to be the system operator? a. What experience did you have with sanitation system operation and maintenance prior to this position? 7. How often do you perform operation or maintenance on the system? 8. Can you please describe what you do to operate or maintain the system? 9. How much time each day do you spend operating or maintaining the sanitation system? 10. What supplies and tools do you own or have access to for operation, repairs, or maintenance of the sanitation system? 11. How much energy does the sanitation system use each month (either kWh or costs)? 12. How much biogas is produced each day by the sanitation (biogas) system? 13. How much compost is produced each day by the sanitation (biogas) system? 14. How often do you feed (input feedstocks) the sanitation (biogas) system? a. How much feedstocks do you put into the system at each feeding? What are the feedstocks used? 15. Does the system use other resources (i.e. water)? a. If so, how much of each of these resources does it use each month (or each day)? 16. How would you rate the overall function (performance, reliability) of your community’s sanitation system? Very Poor Neither Good or Good Very Poor Poor Good a. Why is the performance of the sanitation/biogas system [insert previous question’s answer]?

System Performance and Use & Maintenance 17. Is the system working now? 18. When is the system working? Not working? 19. If the system is not working, what happens when it is not working? a. What are the reasons why it is not working? 20. Is the system producing biogas now? Compost? a. Are there times when the system is not producing biogas or compost?

193

b. Are there times when the system produces more biogas or compost than other times? Why do you think this happens? 21. Has the system ever had a problem? When/why did that happen? 22. What do you do when there is a problem with the sanitation system? 23. Does the organization help with problems? a. If yes: What do they do? 24. How often does the system need repairs? 25. How often are system repairs needed because of how the system was constructed originally? 26. If maintenance or repairs are needed for the sanitation system, what is it typically for? 27. When maintenance or repairs are needed, what do you do to address these? a. Can you give examples? b. Do these actions fix the problem? 28. How do you think the system could be improved? 29. What are some of the things that you have learned in XX years of working here? 30. What do you think is important for us to know about the women's self help groups? 31. What is the most important thing for us to know about the planning process for this project? 32. What is the motivation to spend so much time and resources managing the system?

Training 33. Did the implementing organization provide training for community members about the sanitation system? a. How often did the trainings occur and who was at the trainings? b. How long did each training last? c. Can you please describe the trainings? d. What was the subject of the training? e. What kind of training? (e.g., lecture, demonstration, active participation, etc.) f. What were the training objectives or goals? g. How did the organization teach the training objectives? h. Were tests performed to check trainee understanding/assess objectives? i. What type of tests or assessments were performed? 34. How would you judge the effectiveness of the training? Very Ineffective Neither Effective Effective Very Ineffective or Ineffective Effective a. Why did you say the training was [insert previous question’s response]? 35. Did the implementing organization provide a manual or reading material? a. If yes: Can you describe these materials? Do you still have them? 36. In what ways were the trainings useful? Not useful? 37. What other support for the system is provided by the organization? 38. What education or training would be useful for you?

Conclusion 39. Is there anything else that is important for us to know about the sanitation system, your role, or your community’s participation?

Interview IO1: Causal Conditions & Outcomes Participants: Implementing Organizations

1. Please complete the following demographic information: a. Gender: Male Female b. Age: 18-24 25-34 35-44 45-54 55-64 Over 65 2. What is your role with [Insert name of implementing organization]?

194

3. How long have your worked for this organization? 4. Can you describe for me the sanitation project in [insert name of community]? a. When did the organization begin the sanitation project? b. Why was the sanitation project started? c. Who was involved in the project?

Community Priorities 5. What were your organization’s goals for the sanitation system? a. Can you describe how your organization developed these goals? b. Were these goals achieved? c. How? 6. Did your organization conduct a priority or needs assessment for the community? a. Can you please describe how you determined and integrated community needs/priorities into the planning and design processes? Into the construction and operation processes? b. What were the community’s priorities for the sanitation system? c. How were these priorities incorporated? d. Which priorities are addressed by the sanitation system and/or your organization’s planning process? e. Were there any priorities that your organization was unable to address in the project? Why? f. What methods were used to conduct needs assessments? g. What were the results of those needs assessments? 7. Can you please describe the benefits the community experiences from the sanitation system? 8. Has the sanitation system met the needs of the community? a. What needs have been met by the sanitation system? b. What needs have not been met by the sanitation system? c. How were community priorities identified and incorporated in the implementation process for existing successful and failed sanitation systems? d. What methods were used to conduct needs assessments? e. What were the results of those needs assessments? f. How was community input elicited during the planning process? g. What elements of the toilets and the treatment system did community members provide input? h. What were community priorities in general and for the sanitation project? i. What was identified?

Decision-Making and Planning 9. How much time did your organization spend planning the sanitation project? 10. In what ways (if any) did the organization include community members in the planning of the sanitation system? 11. In what ways (if any) did community members help with the design of the sanitation system? 12. For which decisions or aspects of the sanitation system did the organization ask for community member input? 13. How was community input elicited during the planning process? 14. What elements of the toilets and the treatment system did community members provide input? 15. What was considered when selecting and designing the sanitation/biogas technology? a. Why was biogas the technology selected (over other non-energy producing technologies)? 16. How was biogas production considered in system planning? 17. How was compost considered in system planning? 18. How was water reuse considered in decisions?

195

a. If water is reused, what is the water reused for? 19. How was sludge considered in decisions? a. If sludge is reused, what is it reused for? 20. How was the cost of the system considered in decisions? 21. How were the benefits of the system considered in decisions? 22. How were the challenges of the system considered in decisions? 23. What were some of the activities or things you did in the community to prepare them? 24. How was the management structure for the CBO developed?

Implementation 25. If not answered previously: Can you please describe how the sanitation system was built? 26. How much time did your organization spend building the system? 27. Were there any problems that happened during construction of the system? 28. In what ways (if any) did the organization include community members in the construction of the sanitation system? 29. Did community members provide materials or help build the system? a. What did they provide and/or how did they help? 30. Who paid for the construction of the sanitation system? a. How much did each individual/group pay? 31. Who pays for ongoing operations and maintenance costs for the sanitation system? a. How much does each individual/group pay?

Training 32. What training was provided to community members and/or operators about the project? a. Were there different trainings for operators and community members? b. How often did training sessions occur? c. How long did each training last? d. How many community members (or operators) attended each session? e. Can you please describe the trainings? f. What was the subject of the training? g. What kind of training? (e.g., lecture, demonstration, active participation, etc.) h. What were the training objectives or goals? i. How did the organization teach the training objectives? j. Were tests performed to check trainee understanding/assess objectives? k. What type of tests or assessments were performed? 33. How would you judge the effectiveness of the training? Very Ineffective Neither Effective Effective Very Ineffective or Ineffective Effective a. Why did you say the training was [insert previous question’s response]? 34. Were the training objectives or goals achieved? 35. Did you provide a manual or other reading material to operators and/or community members? a. Can you describe these for me? Do you have a copy?

System Performance and Use & Maintenance 36. How would you rate the overall function (performance, reliability) of the community’s sanitation system? Very Poor Neither Good or Good Very Poor Poor Good a. Why is the performance of the community’s sanitation system [insert previous question’s answer]? 37. Is the system working now?

196

38. When is the system working? Not working? 39. Is the system producing biogas now? Compost? a. How much biogas is produced each day by the sanitation (biogas) system? b. How does the community use the biogas? c. How much compost is produced each day by the sanitation (biogas) system? d. How does the community use the compost? e. Are there times when the system is not producing biogas or compost? 7.1 f. Are there times when the system produces more biogas or compost than other times? Why do you think this happens? 40. Has the system ever had a problem? When/why did this happen? a. Did the community report these problems? 41. When a problem occurs, how does your organization respond? 42. How much energy does the community sanitation system use each month (either kWh or costs)? 43. Does the system use other resources (i.e. water)? a. If so, how much of each of these resources does it use each month (or each day)?

Communication with the Community 44. What support does your organization provide for the community regarding the sanitation system? 45. How often do you (or did you in the past) communicate with community members? 46. What monitoring and evaluation has your organization performed since implementation? 47. What are the results of this monitoring and evaluation?

Other Stakeholders 48. Who else has been involved in the project/Who are the other stakeholders? 49. How did your organization include those stakeholders in the project? 50. How were each of these stakeholders informed and consulted for the project? 51. What contributions did these stakeholder make to the project? Can you elaborate? (e.g., financial, time, information, decisions, system management, etc.)

Conclusion 52. What are the biggest lessons learned from the planning and implementation processes of the sanitation system? 53. If you could do the project again, what would you change about the system or process? 54. What would you keep the same about the system or process? 55. Is there anything else that is important for us to know about the sanitation project, your organization’s role, or the community’s participation?

Interview M1: Causal Conditions & Outcomes Participants: Managers

General Information 1. IRAP Community Case Study Code 2. Community Name 3. Please complete the following demographic information: a. Gender: Male Female b. Age: 18-24 25-34 35-44 45-54 55-64 Over 65 4. Manager/community leader contact information 5. GPS Location of Interview 6. Do you live in this community? a. Yes/No 7. If yes, How long have you lived in this community? If no, where do you live?

197

8. What is your occupation? 9. Are you a manager of the sanitation system? a. Yes/No 10. Do you hold a leadership position or participate in any community committees such as a water or sanitation committee? a. If yes, can you describe your role and responsibilities in this position? b. How long have you been in this role?

Sanitation History and Planning 11. When did the sanitation system start its operation? 12. What was the most common sanitation option before implementation of this system? 13. Can you describe how the sanitation project was initiated? 14. Why was this sanitation project initiated? 15. Did you have to obtain a consent of establishment from the pollution control board to implement this system? 16. What were the primary goals or objectives of this sanitation system? 17. What is the reason why this particular technology combination was chosen for this system? 18. How familiar were you with the sanitation technology before it was built? 19. Can you discuss how decisions were made regarding the sanitation project? 20. Can you discuss how the community or users were involved in the planning process? 21. Can you discuss how the community or users were involved in the implementation? 22. What challenges occurred during the planning of the system? 23. What challenges occurred during the implementation of the system?

Management 24. Who has the overall management responsibility for the treatment system? 25. What are the duties of this managing body? What is it responsible for? 26. What challenges, if any, have occurred with the operator?

Government Role 27. What support or services does the government provide in your community? 28. What support or services does the government provide for the sanitation system? 29. Other notes

Interview G1: Causal Conditions & Outcomes Participants: Local Government

General Information 1. IRAP Community Case Study Codes (communities in their purview) 2. Unit of Government a. Town Panchayat b. Village Panchayat c. Gram Panchayat d. Corporation e. District f. State g. National 3. Please complete the following demographic information: a. Gender: Male Female b. Age: 18-24 25-34 35-44 45-54 55-64 Over 65 4. Government employee/representative contact information:

198

5. GPS Location of Interview 6. Do you live in this community? a. Yes/No 7. If yes, How long have you lived in this community? If no, where do you live? 8. What is your occupation? 9. Are you a manager of the sanitation system? a. Yes/No 10. Can you please describe your role at this [unit of government]?

Strategic Plans and Government Services 11. Does your [unit of government] have any strategic plans? a. If so, what are the plans for? b. If there is a strategic plan for water, sanitation, or solid waste management, what are the plans? Can I see a copy? c. How were those plans developed? d. How are the plans implemented? 12. What support or services does the government provide to communities? 13. What support or services related to sanitation does the government provide to communities?

Sanitation System Planning 14. There is a sanitation system in [insert community name]. What was the role of the government in the planning of the project? 15. What decisions did the government make for the project? 16. Who else made decisions? 17. Is there a formal agreement between the community and the government (or the implementing organization and the government)?

Sanitation System Construction and Land Ownership/Lease 18. What was the role of the government in construction? 19. Did the government contribute any funds for the sanitation system? 20. What government subsidies exist for sanitation (or energy) projects? 21. Does the government require construction permits or construction inspection? 22. Who is the legal owner of the land where the sanitation system was built? 23. Does the community lease the land? 24. Can you describe the handover process? 25. Did handover occur? 26. Was there a formal inauguration ceremony? 27. Can you discuss why handover is important? 28. Is there a formal MOU for O&M?

Operation and Maintenance 29. Can you describe whose responsibility it is for O&M? 30. What is the role of the government in operation and maintenance? 31. What was the role of the government in the funding of the project? 32. Does the government support O&M costs? How are those allocations made? 33. What happens when the allocations run out? 34. What happens when there is a repair needed and there is no money? 35. What specifically has your role been with the sanitation system? 36. When there is a problem with the system, what do you do? 37. How often do you visit the sanitation system? 38. When you visit the system, what do you do?

199

39. How often do you talk with the community or operators?

Regulations 40. What regulations are there for sanitation systems here? a. What regulations are there for effluent water quality? b. What regulations are there for reuse of water, biogas, or compost? c. What permits are required for sanitation systems? i. How do you obtain a permit? d. Are there any inspection requirements? e. Are there requirements for continued use and maintenance? (Reporting or inspections) f. Are there any design requirements? g. Did regulations influence the decision to implement this technology within this community? How? 41. Do regulations differ for district, region, state, nation? 42. How did regulations influence this project? 43. What guidelines are there for the implementation or operation and maintenance of sanitation systems?

C2: Sustainability Survey Scripts To be distributed only after oral consent is obtained. Questions in blue are the questions being asked to the participants; all other text is explanatory for the enumerators/researchers.

1.0 Questions for Community Members

1.1 Indicator: Willingness-to-Pay Questions for UNICEF’s Sustainability Checks Framework (UNICEF 2017) to determine their two metrics: 1. Annual sanitation expenditures are calculated as a percentage of annual household income (monthly income, expenditures on utilities, other regular expenses) 2. Sanitation prioritization is qualitatively evaluated based on top three current and top three projected expenditures; if sanitation is included in either, it is considered a high priority for that household, otherwise is considered a low priority.

What is your household’s monthly income? What is your household’s current total savings? What is your household’s expenditure on public utilities (such as the amount for each water, electricity, and sanitation)? What are your household’s regular expenses? What are the top three items or services that your household spends the most on each month? If you received a raise of 1000 rupees, what are the three most important items or services that you would spend the raise on?

Questions from SEI WTP’s Questionnaire and Analysis Method (Fujita et al. 2005) to determine their metric: the most likely amount that households would be willing to pay for sanitation (improvements). There are three different question versions, which depends on the existing status

200

of the sanitation system in the community. The enumerator will pre-determine which “group” each community falls into and will ask only that group’s questions. The households will be selected randomly, and the first column of numbers within a group will be asked to the first household, the second column of numbers within a group will be asked to the second household, and the third column of numbers within a group will be asked to the third household, and repeated as such until the full sample is completed.

Group 1: Communities with failed sanitation systems (i.e., households are not connected to functional sanitation and do not currently pay sanitation fees); ask participants in Cases 3, 4, 7, 8, 9, 10, 13, 15, 17, 18:

Your household is not currently connected to the sanitation system, so you do not pay sanitation tariffs. If you were to receive “satisfactory sanitation services”, would you be for or against paying _(1st suggested price)_ rupees per month? Note that this amount would be in addition to your current monthly household expenditures.

If the respondent says yes, then ask: Would you be for or against paying an additional _(2nd suggested price when yes)_ rupees per month? If the respondent says no, then ask: Would you be for or against paying an additional _(2nd suggested price when no)_ rupees per month?

Group 2: Communities with successful sanitation systems that still experience some issues (i.e., households are connected to functional sanitation that experiences occasional problems such as sewer overflows during rainfall, and households pay monthly sanitation fees); ask participants in Cases 2, 12, 14, 16:

Your household currently pays ___rupees per month in sanitation tariffs. However, current sanitation services are not satisfactory because, for example, sewage overflows during rainfall. If you were to receive “satisfactory sanitation services” , would you be for or against paying an additional _(1st suggested price)_ rupees per month? Note that this amount would be in addition to your current monthly household expenditures.

Group 3:

201

Communities with fully successful sanitation systems (i.e., households are connected to functional sanitation that does not experience issues and pay monthly sanitation fees); as participants in Cases 1, 5, 6, 11, 19, 20:

Your household currently pays ___ rupees per month in sanitation. You are satisfied with current services in the sense that sewage from your household is transported to a safe place. However, there could be some additional improvements to the system, like treating the sewage even further to protect surface or groundwater or adding additional benefits like water recycling. If these improvements were made, would you be for or against paying an additional _(1st suggested price)_ rupees per month? Note that this amount would be in addition to your current monthly household expenditures.

Group Group 1 Group 2 Group 3 Sample Sample Sample Sample Sample Sample Sample Sample Sample 1, 4, 7, 2, 5, 8, 3, 6, 9, 1, 4, 7, 2, 5, 8, 3, 6, 9, 1, 4, 7, 2, 5, 8, 3, 6, 9, etc. etc. etc. etc. etc. etc. etc. etc. etc. 1st Suggested 30 40 55 15 30 40 15 20 25 Price (rupees) 2nd Suggested Price when 40 55 65 30 40 55 20 25 30 yes (rupees) 2nd Suggested Price when 15 30 40 10 15 10 10 15 20 no (rupees)

1.2 Indicator: Handwashing Questions for UNICEF’s Sustainability Checks Framework and the American Red Cross/Center for Disease Control’s Sustainability Framework for their metric of Handwashing Practice (Sabogal et al. 2014; UNICEF 2017). When have you washed your hands in the last 24 hours? How many times per day do you estimate that you wash your hands? For which activities do you wash your hands before? After?

1.3 Indicator: Community Support of Open Defecation Free (ODF) Status Questions for UNICEF’s Sustainability Checks Framework for their metric of community support of ODF Status (UNICEF 2017). Is there a local by-law or community agreement to not openly defecate? Do you feel pressure from other community members to not openly defecate?

202

1.4 Indicator: Maintenance of Open Defecation Free (ODF) Status Questions for UNICEF’s Sustainability Checks Framework for their metric of maintenance of ODF Status (UNICEF 2017). Did you defecate in the open in the past three days? Did you see anyone defecate in the open in the past seven days?

2.0 Questions for Municipalities

2.1 Quality of Open Defecation Free (ODF) Verification Process Questions for UNICEF’s Sustainability Checks Framework for their metric of quality of ODF verification process (UNICEF 2017). Is there a sanitation (ODF communities) database at the district? Is the ODF database operational and in use? How many years has the ODF database existed? Is there a person responsible for the ODF database on the district level? Has the community been declared ODF? When?

203

Appendix D: Observation Checklist

Some observation checklist items were adapted from the Small-Scale Sanitation Scaling-Up (4S) Project managed by Eawag-Sandec and funded by the Bill & Melinda Gates Foundation.

Citation: Eawag-Sandec. (2017). “4S: Small-Scale Sanitation Scaling-Up.” Eawag, (Nov. 4, 2017).

Observation Checklist Basic Community Information 1. Case study code 2. Community name 3. System address 4. Date 5. GPS location of system 6. Number of households in the community 7. Population of community 8. Implementing organization name 9. Implementing organization representative(s) 10. System manager name and contact info 11. Number of system operators 12. Name(s) of operator(s) 13. Community leadership structure 14. People working in the community? 15. Any festivals or meetings happening in the community? 16. Layout of households? 17. Infrastructure inequalities? 18. Available infrastructure 19. Available services 20. Space between households 21. Amount of solid waste in community 22. Drainage 23. Smells 24. Roads 25. Temples 26. Community Hall 27. School 28. Water Supply 29. Electricity 30. Bathing 31. Facilities and services nearby 32. Distance to government offices

204

System Performance 33. Excess sludge accumulation in any chamber or tank? 34. Excess solid waste in any chamber or tank? 35. Excess scum in any chamber or tank? 36. (For biogas) Is the gas pressure gage working? 37. Odors from sewers? 38. Odors in toilets? 39. Odors in any tank or chamber? 40. Odors at effluent? 41. Corrosion anywhere? 42. Clogging anywhere? 43. Anything broken? 44. Abnormal cracking? 45. Construction errors or deficiencies? 46. Signs of vandalism? 47. Manholes easy to open? 48. (For PGF) Health of plants in wetlands/PGF? 49. (For PGF) Stagnation in wetlands/PGF? 50. Algal growth?

Other performance measurements 1. Volume of biogas produced 2. Volume of biogas used 3. Volume of digestate produced 4. Volume of digestate or sludge used 5. Volume of feedstocks used 6. Energy used by system 7. Water used by system 8. Other resources used by system

Operator Resources 1. Safety equipment 2. Other tools 3. Training manuals 4. Other resources

Regulations 1. Obtained consent for operation? 2. Had visits from pollution control board? 3. Meeting discharge requirements?

Use 1. Number of people observed using the system from 5 am – 10 am 2. Number of people observed using the system from 4 pm – 8 pm 3. System income 4. Self-reported number of people using the system in the morning

205

5. Self-reported number of people using the system in the evening 6. Self-reported number of people using the system daily 7. Open defecation observations? How much open defecation still occurs in the community? 8. Is there any misuse observed? 9. Cleanliness of toilets? 10. Water supply availability?

Maintenance 1. Maintenance tasks required 2. Maintenance tasks performed 3. Maintenance errors

Community & Sanitation Priorities 1. Observations on each priority, including location/frequency of problems/current status

4S Technical System Evaluation Checklist

Please note: This form was to be filled out using KoboToolbox, an Android-based survey tool. In the survey tool, skip logic is employed to display questions relevant for each technology type. The full list of questions without skip logic is presented below.

Citation: Eawag-Sandec. (2017). “4S: Small-Scale Sanitation Scaling-Up.” Eawag, (Nov. 4, 2017).

Project code Coordinates of system location Date System configuration

What sewer or other conveyance technology is used to transport the wastewater to the treatment system? a) Conventional sewer b) Simplified sewer c) Solids-free sewer / Septic tank effluent gravity sewer d) Vacuum trucks e) Other f) Not sure, could not see it

Is the sewer and treatment system designed to deal with stormwater? a) Yes, stormwater overflow / bypass is part of the design b) Yes, conveyance and treatment capacity accounts for stormwater c) Yes, other d) No, there is a separate stormwater drainage network e) No, there is no solution for stormwater f) I don’t know

206

What pre-treatment stages are implemented in this system? a) Kitchen grease traps b) Oil skimmer c) Screen with manual cleaning d) Screen with mechanical cleaning e) Grit chamber f) Grease trap at plant g) Integrated grit and grease removal tank h) Other i) None

What are primary, secondary and tertiary treatment stages of the system? a) Primary sedimentation tank/clarifier b) Septic tank c) Improved septic tank (2-3 chambers) d) Imhoff tank e) Biogas digester / biogas settler f) Anaerobic baffled reactor (ABR) g) ABR with integrated settler h) Anaerobic filter (AF) i) AF with integrated settler j) Integrated ABR with AF k) Upflow Anaerobic Sludge Blanket Reactor (UASB) l) Horizontal flow constructed wetland / planted gravel filter (PGF) m) Vertical flow constructed wetland n) Free-water surface constructed wetland o) Trickling filter p) WSP - Anaerobic pond q) WSP - Facultative pond r) WSP - Maturation pond s) Aerated pond t) Activated Sludge Process (ASP) u) Sequencing Batch Reactor (SBR) v) Oxidation Ditch w) Extended Aeration (EA) x) Moving Bed Biofilm Reactor (MBBR) y) Membrane Bioreactor (MBR) z) Rotating Biological Contactor (RBC) aa) Secondary sedimentation tank/clarifier bb) Polishing pond cc) Vortex dd) Coagulation / Flocculation ee) Activated carbon adsorption ff) Depth filtration (packed-bed filtration) gg) Membrane filtration

207

hh) Chlorination ii) UV disinfection jj) Ozonation kk) Other

Is the sludge that is removed from the system being treated? a) Yes b) No c) Unknown d) Sludge is not removed

Where is the sludge treatment plant located? a) Within the wastewater treatment plant facilities b) At a different location

What are the treatment stages for sludge treatment? a) Anaerobic digestion b) Sedimentation / thickening tank or pond c) Unplanted drying bed d) Planted drying bed e) Co-composting f) Co-treatment with sludge at nearby WWTP g) Coagulation / Flocculation h) Belt filter i) Centrifuge j) Filter press k) Screw press l) Thermal drying a) Other b) Not sure, could not see it

Treatment and/or disposal of leachate/effluent from sludge treatment a) Constructed wetland b) Pond c) Back to wastewater treatment system d) Infiltration e) Discharge to natural water body f) Other g) None h) Not sure, could not see it

Use and/or disposal of treated wastewater a) River / stream / lake / pond – without use b) River / stream / lake / pond – with use for irrigation c) River / stream / lake / pond – with use for fish farming d) River / stream / lake / pond – with use of floating plants

208

e) Open drain f) Municipal sewer with treatment further downstream g) Municipal sewer without treatment further downstream h) Infiltration / Percolation pit i) Direct discharge to field for irrigation j) Direct discharge to land without productive use k) Collection tank for transport to discharge site l) Collection tank for irrigation m) Collection tank for toilet flushing n) Collection tank for drinking o) Other p) Not sure, could not see it

Is there an equalization or feeding tank to buffer peak flows? a) Yes b) No, there is a septic tank or similar which adequately does the job c) No

Is the system operational? a) Yes (fully or partly) b) No, it has completely stopped working or never worked

Use patterns

Number of users for which the system was designed Do the water levels in the tanks match the expected levels based on number of users? Do you observe people using the system? How many people are observed using the system? Are there differences in use based on the time of day? Do you observe wastewater flowing into the system? Please elaborate. Do you observe wastewater flowing out of the system? Please elaborate.

Appearance of primary settlers and clarifiers (incl. septic tanks and Imhoff tanks)

Any excessive sludge accumulation in the settler / clarifier? a) Yes (specify on next page) b) No c) Not sure, could not access it

What do you observe?

Any rising bubbles in the settler / clarifier? a) Yes b) No or not significant c) Not sure, could not access it

209

Any floating sludge or scum in the settler / clarifier? a) Yes (specify on next page) b) No or not significant c) Not sure, could not access it

What do you observe?

Appearance of anaerobic digesters (biogas reactors) and gas usage options

What usage option for biogas is installed? a) Cooking b) Lighting c) Water Heater d) Electricity production with generator e) No use (flaring) f) No use (release into the atmosphere) g) Other h) Not sure, could not see it

Other biogas use

Is the biogas actually used according to this design? a) Yes b) No c) Not sure

How is the biogas actually used?

Is there a gas scrubber? a) Yes b) No

Check the water trap (if any) in the gas pipe, what do you observe? a) It is empty b) It is full of water c) I cannot access it d) There is no water trap e) Other

What do you observe? Any details?

Check the gas pressure gauge, what do you observe? a) It is empty (no water) b) It is well calibrated c) It is not calibrated d) It shows pressure

210

e) It does not show pressure f) There is no pressure gauge g) Not sure h) Could not see it i) Other

What do you observe? Any details?

Any gas odours near the biogas reactor? a) Yes, there seems to be a leakage b) No c) Not sure, could not validate it

When you open the valve closest to the biogas digester, does gas come out? a) Yes b) No c) I don’t know, I cannot reach the valve

When you open the valve furthest away from the biogas digester (e.g. at the stove), does gas come out? a) Yes b) No c) I don’t know, I cannot reach the valve

Are there any problems with the biogas unit? a) None b) Corrosion c) Clogging d) Broken down e) No gas produced f) Other g) Not sure

What do you observe? Any details?

Appearance of anaerobic baffled reactors (ABR) and anaerobic filters (AF)

Number of access covers of ABR / AF

Water level in each compartment of ABR / AF a) Normal b) Too low (specify on next page) c) Too high (specify on next page) d) Not sure, could not see it

What do you observe? Any details?

211

Any signs of abnormal water level fluctuations on inner reactor walls, partition walls and vertical pipes of ABR / AF? a) Yes (specify on next page) b) No c) Not sure, could not see it

What signs of water level fluctuations do you observe?

Does the second or third ABR chamber have a thick scum layer? a) Yes b) No c) Not sure, could not see it d) N.a.

Can you observe a large amount of plastic waste (more than approx. 20 items) inside the second or third ABR reactor chamber? a) Yes b) No c) Not sure, could not see it

Appearance of constructed wetlands (CW) / planted gravel filters (PGF)

Health of the plants in the CW / PGF a) Numerous and green b) Scarce and sick-looking c) Even growth on the entire surface d) CW / PGF has dead / unplanted parts e) Plants require cutting f) Other g) Not sure, could not see it

What do you observe? Any details?

Any water at the surface of the CW / PGF? a) Yes (specify on next page) b) No c) Not sure, could not see it

What do you observe? Any details?

Is the swivel pipe (for water level regulation) of the CW / PGF set correctly during the field visit? a) Yes b) No c) I don’t know, I cannot access the chamber

212

d) There is no pipe for water level regulation

Does the wastewater flow homogeneously in the distribution pipes of the vertical flow CW? a) Yes b) No (specify on next page) c) Not sure, could not see it

What do you observe? Any details?

Any signs of rodents in the CW / PGF? a) Yes (specify on next page) b) No c) Not sure d) Could not see it

What signs of rodents do you observe?

Appearance of pumps

Are there any pumps in this sanitation system? a) Yes b) No c) Not sure

Are pumps functioning correctly? a) Yes b) No (specify on next page) c) Not sure, could not see it

What do you observe? Any details? What is the reason?

What are the technical specifications of the pumps?

What type of power backup exists? a) Diesel generator b) Gas generator c) Dual-fuel engine d) None e) Other f) Not sure, could not see it

Other type of power backup

Site appearance

General cleanliness of the site

213

a) Very clean b) Clean c) Dirty d) Very Dirty

General order at the site a) Tidy b) Messy

Any on-going construction or maintenance activity? a) Yes b) No

Please describe the on-going construction or maintenance activity, including reasons

Any presence of flies and mosquitoes? a) Yes, considerable number b) No, not significant

Any parts closed or restricted for site visit? a) Yes b) No

For what reason?

Any smell emanation in and around treatment plant? a) Very strong smell b) Strong smell c) Normal light smell / no smell

Potential smell nuisance for people nearby? a) Very likely b) Maybe / possible c) Unlikely or not possible (safe distance) d) Not sure

Potential noise disturbance for people nearby? a) Very likely b) Maybe / possible c) Unlikely or not possible (safe distance) d) Insignificant noise emissions e) Not sure

Any inadequate overgrowing by plants inside the treatment plant? a) Yes b) No

214

Design-related aspects

Estimate the total area occupied (footprint) [square feet]

Which elements of the system are prefabricated? a) None b) All (name producer on next page) c) Manholes (name producer on next page) d) Other (specify and name producer on next page)

Description of prefabricated elements and name of producer

Is it possible for a desludging vehicle to access the sludge compartments? a) Yes b) No c) Not sure

Is there access for sampling before/after each system component? a) Yes b) No (specify on next page) c) Not sure (specify on next page

Please describe where access for sampling is/may be difficult or not possible and why

Access possible for unauthorized persons? a) Yes b) No (fence, locking door, guard etc.) c) Not sure

Any changes or corrections implemented retrospectively? a) Yes (specify on next page) b) No c) Not sure

Which changes or corrections were implemented retrospectively? When and for what reason?

Does the installation you see (technology configuration & design) correspond with what you are told? a) Yes, everything appears to be according to the information available b) No, observations differ (specify on next page)

What differences do you observe?

Any peculiar design features, potential design or construction errors or other observations? a) Yes (specify on next page)

215

b) No

Please note your observations

Infrastructure condition

Any remarks regarding construction quality? a) Yes (specify on next page) b) No

What do you observe regarding construction quality?

Any visibility of relevant corrosion of concrete or metal parts or other damage? a) Yes (specify on next page) b) No or insignificant

What corrosion problems or damage do you observe? What implications does this have on the functionality?

Any signs of vandalism? a) Yes (specify on next page) b) No

What signs of vandalism do you observe?

Any abnormal water discharge somewhere? a) Yes (specify on next page) b) No

Where do you observe water discharge? How severe is it? What could be the reason and what implications does it have?

Any water logging? a) Yes (specify on next page) b) No

Where do you observe water logging? How severe is it? What could be the reason and what implications does it have?

Any abnormal water stagnation? a) Yes (specify on next page) b) No

Where do you observe water stagnation? What could be the reason and what implications does it have?

216

Any cracks that may indicate leakages or risks for future leakages? a) Yes (specify on next page) b) No (or insignificant cracks)

Where do you observe cracks? How severe are they? What implications do they have?

Any adverse conditions of manholes/control openings of sewers? a) None b) Absent / stolen c) Stuck (cannot be opened) d) Broken e) Severely corroded f) No handle g) Covered by dirt (invisible, cannot be opened) h) N.a. i) Not sure, could not see manholes/control openings j) Other

Other

Any adverse conditions of manholes/control openings of reactors and tanks? a) None b) Absent / stolen c) Stuck (cannot be opened) d) Broken e) Severely corroded f) No handle g) Covered by dirt (invisible, cannot be opened) h) N.a. i) Other

Other

Operational aspects

Do you see the reason why the system has failed and is not operational? Do you think it is a permanent or a temporary failure?

Are all parts of the system operational? a) Yes b) No c) Not sure

Which parts of the system are not / may not be operational? Do you see the reason why?

217

Can you observe people throwing greywater from kitchen, laundry, cleaning and bathing on the street / into storm drainage or using it for gardening? a) Yes b) No

What do you observe? Any details?

Any dry channels between treatment units where no water is flowing at the time of the visit but should flow under normal operation? a) Yes (specify on next page) b) No

What do you observe? Any details?

Any signs of abnormal water level fluctuation on the walls of reactors? a) Yes b) No c) Not sure, could not see the internal walls of the reactors

Where exactly do you see signs of water level fluctuation (reactor name and compartment)? What exactly do you observe?

Wastewater appearance: any peculiar colour? a) Yes b) No

At which treatment stage do you observe a peculiar wastewater colour? Please describe the colour appearance.

Odour of wastewater and sludge in different compartments: any peculiarities? a) Yes b) No, normal smell c) Not sure, could not verify it

Please describe the odour and the treatment stage at which you observed it

Any compartments with presence of algal growth? a) Yes b) No

Where exactly can you see algal growth (reactor name and compartment)?

Any compartments with solid waste appearance in scum, sludge, or on water surface? a) Yes, significant amounts of solid waste visible b) Insignificant amounts of solid waste visible c) No solid waste visible

218

d) Not sure, could not see the inside of all reactors

Where exactly can you see solid waste (reactor name and compartment)? In the scum, sludge and/or on water surface?

Any scum accumulation and/or discharge into subsequent treatment units? a) Yes b) No significant scum accumulation

Where exactly can you see scum accumulation and/or discharge (reactor name and compartment)?

Any excessive foam / froth accumulation and discharge into subsequent treatment units? a) Yes b) No significant foam / froth accumulation

Where exactly can you see foam / froth accumulation and/or discharge (reactor name and compartment)?

Is the central electrical board for control of plant operation on or off? a) On b) Off c) Broken d) Not sure, could not see it

Is the operating personnel using adequate safety and protection equipment? a) Yes b) No c) Not sure, could not see it

Which safety and protection equipment is lacking or inadequate?

Is there a logbook or any other form of documentation/reports where information regarding operation, maintenance and monitoring of the system is recorded? a) Yes b) No c) I don’t know

What kind of information is being documented? Please select only what you see. a) Performance monitoring data (lab results) b) Inflow and outflow volume c) Energy use d) Other operational data (e.g. pressure head at filters) e) Operational activities (e.g. removal of screenings) f) Maintenance activities (e.g. change of oil and grease, motor maintenance) g) Inspection results

219

h) General observations and notes i) Financial data j) Working hours and presence of operator k) Other (specify on next page) l) Not sure, I could not see it

Other documented information

How are the financial details such as costs and revenues recorded and documented? Please select only what you see. a) Excel b) Accounting software c) Logbook d) Regular reports e) Not recorded / documented f) Other g) Not sure, I could not see it

Other way how financial details are recorded and documented

Influent, effluent and sludge characteristics

Turbidity of effluent a) Very clear effluent b) Slightly turbid effluent c) Very turbid effluent d) Not sure, could not see it

Is the effluent discharge frothy? a) Yes b) No c) Not sure, could not see it

Extensive plant growth in discharge water body? a) Yes, remarkable plant growth b) Yes, but normal c) No plant growth d) Could not see discharge water body e) N.a. – no discharge water body

Any dead fish in discharge water body? a) Yes b) No

Do you think the system is currently hydraulically under- or overloaded (flow of water)? And do you think the system is currently organically under- or overloaded (pollution load)?

220

a) Hydraulically overloaded b) Hydraulically normal c) Hydraulically underloaded d) Organically overloaded e) Organically normal f) Organically underloaded g) Neither under- nor overloaded h) I don’t know

What could be the reason? a) More household connections than planned b) Fewer household connections than planned c) Stormwater entering the system d) Higher water consumption than expected e) Lower water consumption than expected f) Sewer system not correctly connected g) I don’t know h) Other

Other possible reason

Wastewater flow a) Normal: volume of inflow and outflow appear to be equal b) Outflow volume appears to be lower than the inflow volume c) Outflow volume appears to be higher than the inflow volume d) Not sure, could not see it

What could be the reason? Any bypass or loss possible? a) Probably bypass b) Probably loss c) Other reason d) Not sure

What other reason do you see for the lower outflow than inflow?

What could be the reason for higher outflow than inflow?

Water Quality Parameters. a) Dissolved oxygen in aerated / aerobic tanks b) Dissolved oxygen in effluent c) pH in influent d) pH in effluent e) COD in influent f) COD in effluent g) BOD in influent h) BOD in effluent

221

Interviewer’s notes and feedback

Further notes, observations and remarks, additional relevant information collected regarding the small-scale system

222 Appendix E: Published Supporting Information for Chapter 2

This appendix presents the published Electronic Supporting Information in the journal article: Davis, A.; A. Javernick-Will; and S. Cook. A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems. Environmental Science: Water Research & Technology, 2018, 4, 1451–1463.

223 E1.0 System Information and Respondent Demographics Table E1. Summary of community and sanitation system information. Community # Technology Type System Outcome System Manager Community Leader 1 DEWATS Successful Male Operator None 2 DEWATS + Biogas Successful None* Yes 3 DEWATS Failed None Yes 4 Baffled Septic Tank + Gravel Filter Failed None Yes

5 Baffled Septic Tank + Gravel Filter Failed None Yes

6 DEWATS + Biogas Failed WSHG Yes 7 Settling Tank + Single-Pass Successful Male Operator Yes Intermittent Sand Filter 8 DEWATS Successful None* Yes 9 DEWATS Failed WSHG Yes 10 DEWATS Successful Male Operator Yes 11 DEWATS + Biogas Successful WSHG Yes 12 DEWATS + Biogas Failed Male Operator Yes 13 DEWATS + Biogas Successful Male Operator Yes 14 Community Ecological Sanitation Successful None* None System 15 DEWATS Successful Male Operator Yes 16 DEWATS Failed None Yes 17 DEWATS Failed None Yes 18 DEWATS + Biogas Successful WSHG None 19 DEWATS Failed Male Operator Yes 20 DEWATS + Biogas Failed WSHG Yes *These systems did not have operators from the community but received maintenance assistance from the government or sanitation implementer.

224

Table E2. Participation demographics for all communities. Interviews, photovoice, and focus groups sought to identify community and sanitation priorities simultaneously. Interviews Focus Groups Photovoice

Sex Role Sex Role Sex Role Community Number of Community Community Number of Community Community Number of Community Community Male Female Mixed Operator WSHG Male Female Operator WSHG Male Female Operator WSHG # Interviews Leader Member Participants Leader Member Photovoice Leader Member 1 11 36% 45% 18% 0% 91% 9% 0% 10 50% 50% 0% 100% 0% 0% 10 40% 60% 0% 100% 0% 0% 2 11 45% 45% 9% 9% 91% 0% 0% 9 44% 56% 0% 100% 0% 0% 8 38% 50% 0% 100% 0% 0% 3 11 55% 45% 0% 9% 91% 0% 0% 8 50% 50% 0% 100% 0% 0% 9 56% 44% 11% 89% 0% 0% 4 12 33% 50% 17% 8% 92% 0% 0% 10 50% 50% 0% 100% 0% 0% 8 50% 50% 0% 100% 0% 0% 5 12 50% 42% 8% 8% 92% 0% 0% 8 50% 50% 0% 100% 0% 0% 9 44% 56% 0% 100% 0% 0% 6 12 42% 42% 17% 8% 72% 0% 20% 10 50% 50% 10% 70% 0% 20% 9 44% 56% 11% 67% 0% 22% 7 13 38% 46% 15% 8% 77% 15% 0% 9 56% 44% 0% 89% 11% 0% 8 50% 50% 0% 88% 13% 0% 8 12 42% 42% 17% 8% 92% 0% 0% 12 50% 50% 0% 100% 0% 0% 8 50% 50% 13% 88% 0% 0% 9 12 42% 42% 17% 8% 72% 0% 20% 11 45% 55% 0% 73% 0% 27% 9 44% 56% 11% 56% 0% 33% 10 11 45% 55% 0% 9% 73% 9% 0% 11 45% 55% 0% 100% 0% 0% 8 38% 63% 13% 75% 13% 0% 11 11 36% 45% 18% 9% 71% 0% 20% 10 50% 50% 0% 80% 0% 20% 9 44% 56% 0% 67% 0% 33% 12 10 40% 40% 20% 10% 70% 10% 0% 8 50% 50% 13% 75% 13% 0% 8 38% 50% 13% 75% 13% 0% 13 12 50% 42% 8% 8% 83% 8% 0% 8 50% 50% 13% 75% 13% 0% 8 50% 50% 13% 75% 13% 0% 14 12 50% 50% 0% 0% 100% 0% 0% 8 50% 50% 0% 100% 0% 0% 8 50% 50% 0% 100% 0% 0% 15 14 43% 43% 14% 7% 86% 7% 0% 9 44% 56% 0% 100% 0% 0% 9 44% 56% 11% 78% 11% 0% 16 13 38% 54% 8% 8% 92% 0% 0% 8 50% 50% 13% 88% 0% 0% 8 50% 50% 13% 88% 0% 0% 17 10 40% 30% 30% 10% 80% 0% 0% 9 44% 56% 0% 100% 0% 0% 8 50% 50% 0% 100% 0% 0% 18 11 45% 45% 9% 0% 80% 0% 20% 12 50% 50% 0% 75% 0% 25% 10 50% 50% 0% 60% 0% 40% 19 11 55% 45% 0% 18% 73% 9% 0% 9 44% 56% 0% 100% 0% 0% 9 56% 44% 22% 78% 0% 0% 20 11 45% 55% 0% 9% 64% 0% 18% 10 50% 50% 0% 70% 0% 30% 8 38% 63% 0% 63% 0% 38% Average 11.60 44% 45% 11% 8% 83% 3% 6% 9.45 49% 51% 2% 90% 2% 6% 8.55 46% 53% 6% 82% 3% 9%

225

Table E3. Participation demographics for AHP groups for all 20 communities. Sex Role Community Number of Community Community Male Female Operator WSHG # Participants Leader Member 1 10 40% 60% 0% 100% 0%NO 0%NW 2 8 63% 38% 0% 87% 13% 0%NW 3 12 58% 42% 0% 100% 0%NO 0%NW 4 10 50% 50% 0% 100% 0%NO 0%NW 5 8 50% 50% 0% 100% 0% 0%NW 6 9 56% 44% 0% 89% 11% 0%NW 7 9 56% 44% 0% 89% 11% 0%NW 8 12 42% 58% 0% 100% 0%NO 0%NW 9 11 45% 55% 0% 100% 0%NO 0%NW 10 9 44% 56% 0%NL 100% 0% 0%NW 11 9 56% 44% 0% 89% 11% 0%NW 12 11 55% 45% 0% 82% 0%NO 18% 13 10 50% 50% 0% 100% 0%NO 0%NW 14 12 42% 58% 0% 75% 0%NO 25% 15 12 42% 58% 0% 83% 0%NO 17% 16 11 36% 64% 0% 82% 0%NO 18% 17 10 50% 50% 0% 100% 0%NO 0%NW 18 10 40% 60% 0% 100% 0% 0%NW 19 8 63% 38% 0%NL 100% 0%NO 0%NW 20 10 40% 60% 0%NL 80% 0%NO 20% Average 10 49% 51% 0% 93% 2% 5%

226

E2.0 Most Effective Method

Table E4. Summary of the total number of unique community and unique sanitation priorities identified in each community. Numbers reflect the total from all three priority identification methods combined in each community. Community Total Number of Unique # Priorities Identified Community Sanitation 1 15 12 2 12 11 3 15 12 4 12 10 5 12 14 6 13 12 7 10 9 8 15 10 9 14 15 10 14 11 11 15 10 12 14 12 13 14 10 14 14 12 15 14 11 16 15 11 17 13 12 18 14 10 19 15 11 20 15 15 Average 14 12

Table E5. Comparison of method performance in all 20 communities for identifying the total number of unique community priorities and sanitation priorities for (a) interviews compared to focus groups, (b) focus groups compared to interviews, (c) focus groups compared to photovoice. Compared using one-way ANOVA (p-values are listed in parentheses): bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). (a) (b) (c) Interviews vs. Focus Groups Focus Groups Photovoice vs. Interviews vs. Photovoice Community Interviews Interviews Photovoice Total Priorities (0.095)* (0.000) (0.000)* Sanitation Interviews Interviews Photovoice Total Priorities (0.000) (0.015)* (0.003)*

227

Table E6. Comparison of method performance in all 20 communities for identifying the total number of unique physical priorities and unique abstract priorities. Compared using one-way ANOVA (p-values are listed in parentheses): bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). Note: p-values for the comparison between interviews and focus groups and between photovoice and focus groups were all less than 0.033. Physical Priorities Abstract Priorities Community Interviews Interviews Priorities (0.094)* (0.895)* Sanitation Interviews Interviews Priorities (0.652)* (0.000)

Table E7. Comparison of method performance in all 20 communities for identifying the total number of unique community priorities and sanitation priorities for (a) interviews compared to focus groups, (b) focus groups compared to interviews, (c) focus groups compared to photovoice. Compared using one-way ANOVA (p-values are listed in parentheses): bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). (a) (b) (c) Interviews vs. Focus Groups Focus Groups Photovoice vs. Interviews vs. Photovoice Community Most Interviews Interviews Photovoice Priorities Important (0.037)* (0.000) (0.009)* Sanitation Most Interviews Interviews Photovoice Priorities Important (0.000) (0.018)* (0.057)*

E2.1 Most Important Priorities E2.1.1 Analytical Hierarchy Process Methodology The Analytical Hierarchy Process (AHP) was conducted with a different focus group

(called the AHP group) in each community to develop rankings for each of unranked lists of community and sanitation priorities. The unranked lists of community and sanitation priorities were generated from the combined results of each community’s interviews, photovoice, and focus group results. Each community’s AHP group had 8 to 10 community members, who were selected to ensure a balance of age, sex, and socio-economic status within the community. The AHP group was provided with visual aids and verbal explanations (in their native language) of the AHP objectives and procedures. For each AHP pairwise comparison, the participants had to determine which of the two priorities was most important and then rank its relative importance by choosing an integer value between 1 and 9, where 9 indicates a priority is extremely more important than

228

the other and 1 indicates the priorities are equally important. Researchers documented the process with observation notes and intervened in discussions only as needed to ensure that all group members had the chance to voice their opinions and that consensus on the final choices was achieved.

An example pairwise comparison for sanitation priorities: An AHP group was asked to compare low cost and energy generation. An example response was, “Low cost is more important than energy generation.” Then they would be asked, “How much more important is low cost than energy generation?” An example follow-up response was, “Low cost has strong importance over energy generation” (or low cost is 5 times more important than energy generation). The scores chosen for each pairwise comparison populated an AHP comparison matrix, which was used to calculate the relative importance. Thus, the AHP facilitators would assign “low cost” a value of 5 and assign “energy generation” a reciprocal value of 1/5 in the AHP comparison matrix.

Example AHP Pairwise Comparison Matrix. Shaded boxes correspond to the example given in the text. Low Cost Energy Generation Cleanliness Low Cost 1 5 1/3 Energy Generation 1/5 1 1/7 Cleanliness 3 7 1

The pairwise comparisons continued until all priorities in the list had been compared to each of the others. Then, the normalized eigenvector of the AHP comparison matrix was calculated to determine the overall relative rank of each priority (i.e., the ratio scale of priorities).(Saaty 1977,

2008) To assure judgements were not random, a consistency ratio was calculated; the data’s consistency index was first calculated from the normalized eigenvector and then divided by a standardized evaluation index (i.e., the random consistency index, generated from numerous simulations of randomized ratings) to determine the consistency ratio.(Saaty 2008) If the ratio was

229

less than 0.10, then the judgements were considered consistent.(Saaty 2008) If the ratio exceeded

0.10, then the judgements were considered inconsistent (i.e., they could potentially have been generated in a random manner), and researchers facilitated discussions with the AHP groups to resolve contradictory judgements and obtain a consistent result. After a consistent ranked list was generated, researchers performed member checking, where the AHP group was asked to confirm that the rankings and relative importance.

E2.1.2 Determining the Top Priorities Threshold 100%

50% % % Priorities Identified

Interviews Photovoice Focus Groups 0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Number of Priorities

Figure E1. Results of average method performance across all 20 communities for identifying the top priorities. Based on the steadier performance of methods for n>5, it was determined that the top five highest ranked priorities are where methods differ greatest and thus in-depth comparisons are most important.

230

E2.2 Method Effectiveness and Time Required

Figure E2. Comparison of the number of unique priorities identified and the time required for data collection for interviews, photovoice, and focus groups. Results show (a) community priorities and (b) sanitation priorities for all 20 communities.

231

E3.0 Impact of Community Role and Respondent Sex on Method Effectiveness

E3.1 Respondent’s Community Role

Table E8. Results of a comparison of the most effective role from all 20 communities based on identification of the greatest number of the most important priorities. Results are for comparisons between (a) leader and operator, (b) leader and WSHG member, (c) community member and leader, (d) community member and operator, (e) community member and WSHG member, (f) WSHG member and operator. For Community Member (Member) and WSHG Member (WSHG), the results are for the average response for one individual. The p-values from one-way ANOVA are in parentheses. Bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). Role

Priority (a) (b) (c) (d) (e) (f) Method Category Leader Leader Member Member Member WSHG vs. vs. vs. vs. vs. vs. Operator WSHG Leader Operator WSHG Operator

Community Leader WSHG Member Member Member WSHG Priorities (0.574)* (0.635)* (0.000) (0.000) (0.000) (0.054)* Interviews Sanitation Leader WSHG Member Member Member WSHG Priorities (0.454)* (0.255)* (0.000) (0.000) (0.000) (0.053)*

Community Leader WSHG Member Member Member WSHG Priorities (0.149)* (0.188)* (0.000) (0.000) (0.000) (0.020) Photovoice Sanitation Leader WSHG Member Member WSHG WSHG Priorities (0.496)* (0.056)* (0.000) (0.005) (0.000) (0.024)

232

Figure E3. Results of community roles for identifying unique sanitation priorities through interviews for (a) community priorities and (b) sanitation priorities. Bars represent the average number of unique sanitation priorities identified by one participant from each role, except for Multiple Community Members, which presents the total number of sanitation priorities identified by multiple community members.

233

Figure E4. Results of the number of unique sanitation priorities identified by all community members in interviews (n = 7 to 12) and the average community member in an interview for (a) community priorities and (b) sanitation priorities. Error bars represent the maximum and minimum number of unique sanitation priorities identified by individual community members are also presented.

234

Figure E5. Results of the number of unique sanitation priorities identified in interviews by each community role (community leader, community members, system manager). Communities presented are the six communities (out of 20) that had three roles (leader, members, and manager; where a community had either an operator or WSHG). Results are normalized to community leaders and reflect the percent difference between the number of unique sanitation priorities identified by the community leader and members of each comparative role. Total Responses for Community Members and WSHG Members reflect results from all respondents in the role until theoretical saturation was reached (n=7 to 12 and n=3 to 5, respectively). Complete results of the comparison between roles for all communities are presented in Figure E3(a)†. Results for sanitation priorities are presented in Figures E3(b) and E5†. CL = Community leader; CM = Community member; OP = Operator; WSHG = Women’s Self-Help Group. *Community number not in numerical order.

235

Figure E6. Results for identifying sanitation priorities when speaking with community leaders for (a) community priorities and (b) sanitation priorities. Bars represent the percent of unique sanitation priorities identified by one community leader from interviews and photovoice.

236

E3.2 Community Member Sex

Table E9. Results of the effectiveness of respondent sexes for identifying the most unique priorities. Results are presented for a comparison of the total number of unique priorities identified by multiple members of each sex and the average number of unique priorities identified by one member of each sex. P-values from one-way ANOVA are in parentheses. Bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). Respondent Sex Priority Priority Method Average Male Multiple Males Category Importance vs. Average vs. Multiple Female Females

Total Female (0.000) Females (0.000) Community Priorities Most Female (0.000) Females (0.001) Important Interviews Total Female (0.005) Females (0.110)* Sanitation Priorities Most Female (0.000) Females (0.039) Important Total Female (0.049) Females (0.610)* Community Priorities Most Female (0.106)* Females (0.387)* Important Photovoice Total Female (0.317)* Females (0.015) Sanitation Priorities Most Female (0.137)* Females (0.213)* Important

E3.3 Most Effective Method for Each Community Role and Respondent Sex

Table E10. Results of the most effective methods to use with each community role. P-values from one-way ANOVA are in parentheses. Bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). Community Priorities Sanitation Priorities Total Most Important Total Most Important Photovoice Photovoice Interviews Interviews Community Leader (0.799)* (0.236)* (0.818)* (0.754)* Multiple Community Interviews Interviews Interviews Interviews Members (0.062)* (0.010) (0.006) (0.026) Community Average Interviews Photovoice Interviews Interviews Roles Community Member (0.234)* (0.150)* (0.002) (0.618)* Interviews Photovoice Interviews Interviews Operator (0.775)* (0.897)* (0.739)* (0.678)* Average WSHG Interviews Photovoice Interviews Interviews Member (0.608)* (0.907)* (0.868)* (0.846)*

237

Table E11. Results of the most effective methods to use with each sex. Results are presented for a comparison of the total number of unique priorities identified by multiple members of each sex and the average number of unique priorities identified by one member of each sex. P-values from one-way ANOVA are in parentheses. Bold denotes p-values less than the confidence-level used in the analysis (p<0.05) and * denotes p-values greater than the confidence-level used in the analysis (p>0.05). Community Priorities Sanitation Priorities

Most Most Total Total Important Important Photovoice Interviews Interviews Interviews Average Male (0.017) (0.929)* (0.317)* (0.056)* Average Interviews Interviews Interviews Interviews Respondent Female (0.324)* (0.000) (0.007) (0.000) Sex Photovoice Photovoice Interviews Interviews Multiple Males (0.002) (0.103)* (0.091)* (0.119)* Multiple Interviews Interviews Interviews Interviews Females (0.399)* (0.247)* (0.723)* (0.082)*

E4.0 Electronic Supporting Information References Saaty, T. L. (1977). “A Scaling Method for Priorities in Hierarchical Structures.” Journal of Mathematical Psychology, 15, 234–281. Saaty, T. L. (2008). “Decision making with the analytic hierarchy process.” International Journal of Services Sciences, 1(1), 83–98.

238

Appendix F: Published Supporting Information for Chapter 3

This appendix presents the published Supporting Information in the journal article: Davis, A.; Javernick-Will, A.; Cook, S. Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities. Environmental Science & Technology 2019, 53 (1), 401–411. https://doi.org/10.1021/acs.est.8b04761.

239

F1. Sanitation System Description Table F1. Summary of each case’s sanitation system.a Case Current Current Technology Current Status # Manager Functional Conventional Male 1 (100% Used, 91% Maintained, 3 (DEWATSb) Operator of 3 Regulations Met) Resource Recovery Functional 2 (DEWATS with Biogas Digester and Onsite Irrigation; Designed to (100% Used, 90% Maintained, 3 Nonec recover biogas and water) of 3 Regulations Met) Functional Conventional Male 3 (90% Used, 90% Maintained, (DEWATS) Operator 2 of 2 Regulations Met) Nonfunctional Conventional 4 (0% Used, 0% Maintained, None (Baffled Septic Tank + Gravel Filter) 0 of 3 Regulations Met) Nonfunctional Resource Recovery 5 (22% Used, 0% Maintained, None (Baffled Septic Tank + Gravel Filter; Designed to recover water) 1 of 2 Regulations Met) Resource Recovery Nonfunctional 6 (DEWATS with Biogas Digester and Water Recirculation for Toilet (48% Used, 7% Maintained, WSHG Flushing; Designed to recover biogas and water) 0 of 3 Regulations Met) Functional Conventional Male 7 (95% Used, 100% Maintained, 2 (Settling Tank + Single-Pass Intermittent Sand Filter) Operator of 2 Regulations Met) Functional Conventional 8 (90% Used, 100% Maintained, 2 Nonec (DEWATS) of 2 Regulations Met) Resource Recovery Nonfunctional 9 (DEWATS with Static Pile Composting and Onsite Irrigation; Designed (85% Used, 20% Maintained, WSHG to recover water and compost) 0 of 3 Regulations Met) Nonfunctional Conventional Male 10 (6% Used, 0% Maintained, (DEWATS) Operator 1 of 2 Regulations Met) Resource Recovery Functional 11 (DEWATS with Biogas Digester and Onsite Irrigation; Designed to (100% Used, 92% Maintained, 3 WSHG recover biogas and water) of 3 Regulations Met) Nonfunctional Resource Recovery Male 12 (32% Used, 7% Maintained, (DEWATS with Biogas Digester; Designed to recover biogas) Operator 0 of 3 Regulations Met) Resource Recovery Functional Male 13 (DEWATS with Biogas Digester and Static Pile Composting; Designed (100% Used, 100% Maintained, Operator to recover biogas and compost) 3 of 3 Regulations Met) Resource Recovery Functional (Community Ecological Sanitation System(Tilley et al. 2014b) with Male 14 (100% Used, 100% Maintained, Onsite Irrigation and Static Pile Composting; Designed to recover water Operator 3 of 3 Regulations Met) and compost) Functional Conventional Male 15 (91% Used, 91% Maintained, (DEWATS) Operator 2 of 2 Regulations Met) Nonfunctional Conventional 16 (31% Used, 3% Maintained, None (DEWATS) 1 of 2 Regulations Met) Nonfunctional Conventional 17 (88% Used, 11% Maintained, None (DEWATS) 0 of 3 Regulations Met) Resource Recovery Functional 18 (DEWATS with Biogas Digester and Onsite Irrigation; Designed to (100% Used, 100% Maintained, WSHG recover biogas and water) 3 of 3 Regulations Met) Nonfunctional Conventional Male 19 (82% Used, 60% Maintained, (DEWATS) Operator 1 of 3 Regulations Met) Resource Recovery Nonfunctional 20 (DEWATS with Biogas Digester and Water Recirculation for Toilet (0% Used, 0% Maintained, WSHG Flushing; Designed to recover biogas and water) 0 of 3 Regulations Met) a The same 20 cases (communities) were evaluated in Davis et. al 2018(Davis et al. 2018b) and use the same case numbers.

240

b DEWATS is a Decentralized Wastewater Treatment System that consists of a settling tank (conventional) or biogas digester (resource recovery), anaerobic baffled reactor, and planted gravel filter.(Tilley et al. 2014b) c These systems did not have operators from the community, but the government or sanitation implementer would schedule and pay for an external service provider (e.g., desludging company) as needed for maintenance assistance. F2. Priority Coding Dictionary Table F2. Coding dictionary for sanitation priorities. Code Definition The need for or importance of having bathing facilities available where the Bathing Facilities at Toilets shared community toilets are located. The need for or importance of having a well-planned sanitation project Better Sanitation Planning where it is clear that the implementers are invested and want the project to be successful. The need for or importance of having generating biogas for sale or Biogas productive use (e.g., cooking fuel, electricity generation) from the sanitation system. The need for or importance of having the shared community toilets located Central Location in a central area, equidistant to most community members. The need for or importance of having access to smaller, child-friendly Child-Friendly Toilets toilets. The need for or importance of having a sanitation system with a user- Comfortable interface (e.g., toilet seats, stalls) that is comfortable. The need for or importance of having adequate community involvement in Community Involvement in the sanitation project, in particular, by having community involvement in Sanitation Planning planning and design. The need for or importance of providing more training to community Community O&M Training members regarding both basic sanitation concepts and technical treatment system information. The need for or importance of generating compost for sale or productive use Compost (e.g., organic fertilizer) from the sanitation system. The need for or importance of connecting sewer lines directly to the municipal (government) sewer lines, which typically lead to a large and Direct Municipal Sewer centralized treatment system. (At a minimum, conveying untreated Connection wastewater outside of a community so that there is no community treatment system.) The need for or importance of having toilets that are easy to use and do not Easy to Use require training. Efficient & Functional The need for or importance of having better pumps that efficiently and Treatment System Pumps quickly pump treated wastewater from the treatment system. Functioning Treatment The need for or importance of having a functional treatment system that is System problem-free and adequately treats the wastewater. The need for or importance of having regular, sufficient, and good operation Good O&M and maintenance of the sanitation system. The need for or importance of having well-constructed sanitation infrastructure, including a fully constructed system, with construction that Good Quality Construction matches the design, and improvements made to reduce maintenance issues that result due to poor construction quality. Government Support for The need for or importance of having ongoing technical and financial O&M support for the operation and maintenance of the treatment system. The need for or importance of having positive health benefits as a result of a Health & Hygiene sanitation system.

241

Code Definition The need for or importance of having regular and sufficient income Income Generation generation from the sanitation system, such as by user fees and/or sale of recovered resources (biogas, compost, water). The need for or importance of having large sewer pipes to reduce blockages Increase Sewer Pipe Size and broken pipes. The need for or importance of exchanging household toilets connected to a Individual Septic Tanks shared treatment system for household toilets connected to individual household septic tanks. The need for or importance of the sanitation system providing job Jobs for Women opportunities for women. The need for or importance of having sufficient lighting inside the toilets Lights in Toilets and around the toilets and treatment system to reduce accidents, vandalism, and crimes and to increase safety. The need for or importance of having toilets or a treatment system that is Low Cost affordable, overall for the community (e.g., capital cost, maintenance costs) and for all users (i.e., affordable user fees). The need for or importance of reducing the burden and complexity of Low O&M Demands maintenance on the community for the sanitation system. The need for or importance of having a sanitation system that generates Micro-Loan Program sufficient income to be able to continuously offer micro-loans to community members. The need for or importance of constructing manholes at grade (instead of 2-4 Move Manholes to Grade feet above grade) to increase space in the community and reduce accidents. The need for or importance of moving household toilets outside of the main Move Toilets Away from home structure, either to increase space inside the house or increase the Kitchen/Prayer Rooms* distance between the toilet and the kitchen or puja (worship) room. *In Davis et. al 2018,(Davis et al. 2018b) Move Toilets Away from Kitchen/Prayer Rooms is referred to as “Move Household Toilets Outside”. The need for or importance of developing the space around the community Multi-Use Area (Park) toilets for community use such as a park, garden, meeting area, or nice sitting space. The need for or importance of eliminating open defecation by having a good No Open Defecation sanitation system in the community. The need for or importance of eliminating blockages in toilets and sewers, No Sewer Blockages backflow into toilets, and overflows from any part of the sanitation system. The need for or importance of having toilets and a treatment process that are No Smells odor-free. Open 24 Hours The need for or importance of having community toilets open continuously. The need for or importance of having a sanitation system that guarantees Privacy privacy for the users. The need for or importance of having a sufficient number of community Reduced Waiting Time toilets so that users do not have to wait in lines. The need for or importance of repairing system damage such as to fix broken Repair System Damage doors to toilets, cracked pipes, or broken access covers. The need for or importance of having a sanitation system that is safe to use Safety and Dignity and provides dignity for users, particularly women and girls. The need for or importance of having a sanitary napkin disposal system at all Sanitary Napkin Disposal toilets that women will use. The need for or importance of having security for the sanitation Security for Treatment infrastructure to reduce vandalism and damage, such as a fence or lockable System gate.

242

Code Definition Shade for Sanitation The need for or importance of providing shade for the operator while Caretakers collecting user fees. The need for or importance of eliminating vandalism, damage, and dangers associated with people who drink, harass caretakers, or intentionally destroy Stop Antisocial Elements* the sanitation system. *In Davis et. al 2018,(Davis et al. 2018b) Stop Antisocial Elements is referred to as “Stop Antisocial Elements at Toilets”. The need for or importance of having functional drainage in the bathing Stop Shower Drain Clogging rooms provided at the community toilets. The need for or importance of having regularly maintained and constantly Toilet Cleanliness clean toilets. Toilets in all Houses The need for or importance of having toilets in each household. The need for or importance of also treating greywater from kitchens in the Treat Kitchen Greywater treatment system. The need for or importance of wastewater being treated prior to disposal. This captures specific reference to the importance of adequate treatment Treatment of Wastewater processes, distinct from a value of sanitation or having a treatment system in general. The need for or importance of having the treatment system or wastewater discharge far enough away from the community so that problems with Treatment System Far Away smells, overflows, visual aesthetics, and proximity to households are reduced. The need for or importance of having visually pleasing toilet facilities such Visual Aesthetics as nice painting, tiles, etc. Water Reuse The need for or importance of reusing treated wastewater for beneficial use. The need for or importance of having low-flush or waterless toilets that Water Savings reduce the water used by the sanitation system. The need for or importance of having a regular and adequate water supply Water Supply for Toilets for toilet flushing. The need for or importance of having western toilets for comfort, ease of use Western Toilets (in particular for the elderly and disabled) and for making toilets modern. The need for or importance of the sanitation system to provide opportunities Women's Empowerment for learning skills, making decisions, and jobs for women.

243

Table F3. Coding dictionary for community priorities. Code Definition The need for or importance of having a place for bathing within individual Bathing Facilities in Houses households. The need for or importance of having infrastructure systems and government Better Services for Elderly and services available and designed to accommodate elderly and disabled Disabled community members, such as pensions, mobile health clinics, disability- accessible infrastructure, etc. Bus Facilities The need for or importance of having nearby access to public transportation. The need for or importance of having a clean community environment that is Community Cleanliness generally free of pollution, dirt, weeds, solid waste, and defecation. The need for or importance of having a well-maintained community hall where Community Hall meetings, festivals, and functions can be held by community members. The need for or importance of having unity amongst community members and Community Unity strong, non-corrupt community leadership. The need for or importance of having regularly accessible and affordable Cooking Fuel cooking fuel. The need for or importance of having a functional, clean, and well-maintained Drainage (rainwater) drainage system that is free of smells, blockages, and mosquito breeding. The need for or importance of reducing dust particles and sound nuisances Dust & Sounds from Husk from a coconut husk processing factory run out of one house in the Factory community. The need for or importance of having nearby, quality, and affordable access to Education education for children and young adults in the community. The need for or importance of having uninterrupted and affordable electricity Electricity supply available in all households. The need for or importance of removing pests such as mosquitos or rats from Eradicating Pests the community. The need for or importance of having a fence that surrounds the nearby lake to Fence Around Nearby Lake prevent solid waste from going into the lake and to protect children from injury at the lake. The general need for or importance of having a functional sanitation system in Good Sanitation System the community. This priority is expressed as a general value of sanitation. The need for or importance of having adequate services provided by the Government Support government, primarily related to basic services and needs such as education, infrastructure provision and maintenance, and income. The need for or importance of having designated and sufficient space nearby to Graveyard a community for burying deceased family and community members. The need for or importance of having good health and access to nearby and Health & Healthcare affordable healthcare, in particular, emergency services and hospitals. The need for or importance of having well-maintained and well-constructed House Improvements houses and/or the ability to upgrade and expand existing shelters to accommodate growing families and changing family needs. The need for or importance of having formal house ownership in the name of House Ownership each community family. The need for or importance of having individual household drinking water Household Piped Water supply access (instead of shared community water sources). The need for or importance of having adequate access to jobs for all willing Jobs & Income community members and for having increased income to provide for families.

244

Code Definition The need for or importance of having formal land ownership in the name of Land Ownership each community family. Library The need for or importance of having a library located within the community. The need for or importance of having a micro-loan program available for Micro-Loans for Women women to borrow funds to start business, pay for household expenses, and support their children’s education. The need for or importance of having more space within a community so that More Land houses are not so close together or that expanding families and newly married couples can build new or larger houses. The general need for or importance of seeing development and progress within Overall Community a community, such as increased access to infrastructure and recognition of the Development community as part of a municipality instead of town panchayat. The specific need for or importance of having a place for children to safely Park for Children play. The need for or importance of having regular and sufficient access to safe Potable Drinking Water drinking water. The need for or importance of having well-functioning and well-maintained Public Toilet public toilets for tourists and visitors. The need for or importance of having a government-run ration shop located Ration Shop nearby to the community where community members could obtain government-provided rations for rice, flour, spices, and other food staples. Remove Old Water Supply The need for or importance of removing old water supply tanks to eliminate Tank water and location of mosquito breeding. The need for or importance of having well-maintained and quality paved roads Road Improvements throughout the community, specifically free of dangerous potholes and poor road drainage. The need for or importance of making improvements to the physical School Improvements infrastructure of community schools such as improving school water supply, fixing fences for safety, maintaining school toilets, etc. Smells at Resource Recovery The need for or importance of reducing smells associated with composting and Park solid waste segregation practices. The need for or importance of having regular solid waste collection and a Solid Waste Management means to dispose of solid waste outside of the community. The need for or importance of ending the untreated solid waste and wastewater Stop Juice Factory Waste discharge from the nearby juice and meat processing factory. The need for or importance of having streetlights throughout the community to Street Lights increase safety. The need for or importance of having an unimpeded, clear, and safe escape Tsunami Escape Route route in the event of a tsunami disaster where all community members are able to escape safely. The need for or importance of having access to regular, sufficient, and Water Supply affordable water for general use such as bathing, cooking, irrigation, toilet flushing, etc. The need for or importance of having increased opportunities for learning Women's Empowerment skills, making decisions, and jobs for women.

245

F3. Priority Addressment Data Collection Example of Priority Addressment Data Collection: For the sanitation priority of Water

Supply at Toilets, observations evaluated the water tank levels and pump functionality, and interview questions included How many times is water unavailable for toilet flushing in one week?

Table F4. Participation demographics for interviews evaluating the level of priority addressment for all 20 cases. Sex Role Number of Community Case Male Female Operator WSHG Participants Member 1 7 57% 43% 86% 14% 0%NW 2 8 50% 50% 100% 0%NO 0%NW 3 6 50% 50% 100% 0%NO 0%NW 4 5 40% 60% 100% 0%NO 0%NW 5 6 50% 50% 100% 0%NO 0%NW 6 5 40% 60% 80% 0%NO 20% 7 6 50% 50% 83% 17% 0%NW 8 8 38% 63% 100% 0%NO 0%NW 9 5 60% 40% 80% 0%NO 20% 10 6 50% 50% 83% 17% 0%NW 11 6 50% 50% 67% 0%NO 33% 12 5 60% 40% 80% 20% 0%NW 13 7 43% 57% 86% 14% 0%NW 14 7 57% 43% 100% 0%NO 0%NW 15 6 50% 50% 83% 17% 0%NW 16 7 57% 43% 100% 0%NO 0%NW 17 5 60% 40% 100% 0%NO 0%NW 18 6 50% 50% 83% 0%NO 17% 19 5 40% 60% 80% 20% 0%NW 20 6 50% 50% 67% 0%NO 33% Average 6 50% 50% 88% 6% 6%

246

F4. Assumptions for Addressed Values Table F5. Changes to sanitation priorities’ addressed values from the Current Scenario to the Intended Design Scenario. Priorities whose addressed value increased Priorities whose addressed value increased Priorities whose addressed value did not change by one addressment level (0.33) to always addressed (1.0) Bathing Facilities in Toilets Better Sanitation Planning Biogasa Central Location Community Involvement in Planning Child-Friendly Toilets Direct Municipal Sewer Connection Community O&M Training Comfortable Increase Sewer Pipe Size Government Support for O&M Compostb Individual Septic Tanks Health & Hygiene Easy to Use Move Manholes to Grade Income Generation Efficient & Functional Treatment System Pumps Move Toilets Away from Kitchen/Prayer Rooms Jobs for Women Functioning Treatment System Shade for Sanitation Caretakers Low Cost Good O&M Treat Kitchen Greywater Low O&M Demands Good Quality Construction Treatment System Far Away Micro-Loan Program Lights in Toilets Western Toilets No Open Defecation Multi-Use Area (Park)c Privacy No Sewer Blockages Safety & Dignity No Smells Stop Antisocial Elements Open 24 Hours Women's Empowerment Reduced Waiting Timed Repair System Damage Sanitary Napkin Disposale Security for Treatment System Stop Shower Drain Clogging Toilet Cleanliness Toilets in All Housesf Treatment of Wastewater Visual Aestheticsg Water Reuseh Water Savings Water Supply at Toilets a Part of the intended design for Cases 2, 7, 10, 12, 13, 18, and 20; not included in the intended design for Cases 15 and 17. b Part of the intended design for Cases 9, 11, and 14; not included in the intended design for Cases 10, 12, and 20. c Part of the intended design for Cases 6 and 11; not included in the intended design for Case 8. d Part of the intended design for Cases 11, 13, and 14; not included in the intended design for Case 20. e Part of the intended design for Cases 13 and 18; not included in the intended design for Case 5. f Part of the intended design for Case 16, but 50 of 200 households did not receive the toilets. g Part of the intended design for Cases 11, 15, 18, and 20; not included in the intended design for Case 12. h Part of the intended design for Cases 2, 9, 13, 18, and 20; not included in the intended design for Cases 1, 3, 6, 17.

247

Table F6. Changes to community priorities’ addressed values from the Current Scenario to the Intended Design Scenario. Priorities whose addressed Priorities whose addressed Priorities whose addressed value did not value increased by one value increased change addressment level (0.33) to always addressed (1.0) Bathing Facilities in Houses Community Cleanliness Good Sanitation System Better Services for Elderly & Disabled Cooking Fuela Bus Facilities Electricitya Community Hall Eradicating Pests Community Unity Health & Healthcare Drainage Jobs & Income Dust & Sounds from Husk Factory Micro-Loans for Women Education Water Supplya Fence Around Nearby Lake Women’s Empowerment Government Support Graveyard House Improvements House Ownership Household Piped Water Supply Land Ownership Library More Land Overall Community Development Park for Children Potable Drinking Water Public Toilet Ration Shop Remove Old Water Supply Tank Road Improvements School Improvements Smells at Resource Recovery Park Solid Waste Management Stop Juice Factory Waste Streetlights Tsunami Escape Route a Increased by one addressment level only for nonfunctional resource recovery cases that were designed to recover biogas or water and then became hypothetically functional in the Intended Design Scenario.

248

Table F7. Changes to sanitation priorities’ addressed values from the Intended Design Scenario to the Added Resource Recovery Scenario. Priorities whose Priorities whose addressed value Priorities whose addressed addressed value Priorities whose addressed value did not change increased by one addressment value increased decreased by one level (0.33) to always addressed (1.0) addressment level (0.33) Low O&M Demands Bathing Facilities in Toilets Income Generation Biogas Better Planning Jobs for Women Compost Central Location Low Cost Lights in Toilets Child-Friendly Toilets Micro-Loan Opportunities Water Reuse Comfortable Women's Empowerment Water Savings Community Involvement in Planning Water Supply at Toilets Community O&M Training Direct Municipal Sewer Connection Easy to Use Efficient & Functional Treatment System Pumps Functioning Treatment System Good O&M Good Quality Construction Government Support for O&M Health & Hygiene Individual Septic Tanks Move Manholes to Grade Move Toilets Away from Kitchen/Prayer Rooms Multi-Use Area (Park) No Open Defecation No Sewer Blockages No Smells Open 24 Hours Privacy Reduced Waiting Time Repair System Damage Safety & Dignity Sanitary Napkin Disposal Security for Treatment System

249

Priorities whose Priorities whose addressed value Priorities whose addressed addressed value Priorities whose addressed value did not change increased by one addressment value increased decreased by one level (0.33) to always addressed (1.0) addressment level (0.33) Shade for Sanitation Caretakers Stop Antisocial Elements Stop Shower Drain Clogging Toilet Cleanliness Toilets in All Houses Treat Kitchen Greywater in System Treatment of Wastewater Treatment System Far Away Visual Aesthetics Water Supply at Toilets Western Toilets

250

Table F8. Changes to community priorities’ addressed values from the Intended Design Scenario to the Added Resource Recovery Scenario. Priorities whose Priorities whose addressed addressed value Priorities whose addressed value did not value increased by one increased change addressment level (0.33) to always addressed (1.0) Bathing Facilities in Houses Cooking Fuel None. Better Services for Elderly & Disabled Electricity Bus Facilities Jobs & Income Community Cleanliness Micro-Loans for Women Community Hall Women’s Empowerment Community Unity Water Supply Drainage Dust & Sounds from Husk Factory Education Eradicating Pests Fence Around Nearby Lake Good Sanitation System Government Support Graveyard Health & Healthcare House Improvements House Ownership Household Piped Water Supply Land Ownership Library More Land Overall Community Development Park for Children Potable Drinking Water Public Toilet Ration Shop Remove Old Water Supply Tank Road Improvements School Improvements Smells at Resource Recovery Park Solid Waste Management Stop Juice Factory Waste Streetlights Tsunami Escape Route

251

F5. System Use

Figure F1. Percent of each case’s population that used the sanitation system.

F6. Sanitation Priorities Results

Figure F2. Results of the AHP relative importance values for the most common sanitation priorities (i.e., priorities that were expressed by 10 or more cases). The large range of importance demonstrates that priority rankings, in addition to priorities, were case-specific.

252

Table F9. Comparison of all cases’ total case sanitation scores between scenarios. Compared using one-way ANOVA (p-values are listed in parentheses): bold denotes statistically different (p>0.05) and * denotes not statistically different. The scenario listed in each box is the scenario whose average score was greater than the comparative scenario. Resource Functional Nonfunctional Recovery Conventional

All Cases Cases Cases Cases Cases (a) (b) (c) (d) (e) Current Intended Intended Intended Intended Intended vs. Design Design Design Design Design Intended (0.000) (0.228)* (0.000) (0.014) (0.001) Design Current Added Added Added Added Added vs. Resource Resource Resource Resource Resource Added Recovery Recovery Recovery Recovery Recovery Resource (0.000) (0.096)* (0.000) (0.011) (0.000) Recovery Intended Design Added Added Added Added Added vs. Resource Resource Resource Resource Resource Added Recovery Recovery Recovery Recovery Recovery Resource (0.325)* (0.654)* (0.278)* (0.774)* (0.283)* Recovery

Table F10. Comparison of total case sanitation scores: Percent difference for the change in score from the Current and Intended Design Scenarios for each case. A positive percentage means that the Intended Design Scenario total case sanitation score was larger than the Current Scenario.

Nonfunctional Functional Conventional Nonfunctional Resource Functional Resource Conventional Sanitation Sanitation Recovery Recovery

Case # % Difference Case # % Difference Case # % Difference Case # % Difference 4 350% 7 6% 12 132% 11 22% 10 1808% 15 82% 20 658% 2 29% 16 759% 3 20% 5 221% 13 9% 17 197% 8 62% 6 168% 14 5% 19 227% 1 27% 9 111% 18 2%

253

F7. Community Priorities Results Table F11. Summary of community priorities from all 20 cases. R denotes community priorities that could be influenced by resource recovery systems; S denotes community priorities that could be influenced by sanitation systems. The most commonly expressed priorities (shared by 10 or more cases) are highlighted in yellow. Community Priorities # of Cases Community CleanlinessS 20 Good Sanitation SystemS 20 Drainage 19 Solid Waste Management 18 Jobs & IncomeS, R 18 Water SupplyR 17 Education 17 Health & HealthcareS 17 House Improvements 16 Potable Drinking Water 15 Government Support 11 Cooking FuelR 10 Road Improvements 9 Eradicating PestsS 9 ElectricityR 8 Community Hall 6 Streetlights 5 Overall Community DevelopmentS, R 4 Bus Facilities 3 House Ownership 3 Household Piped Water 2 More Land 2 Park for Children 2 School Improvements 2 Bathing Facilities in Houses 1 Dust & Sounds from Husk Factory 1 Fence Around Nearby Lake 1 Graveyard 1 Library 1 Public Toilet 1 Ration Shop 1 Remove Old Water Supply Tank 1 Smells at Resource Recovery Park 1 Stop Juice Factory Waste 1 Tsunami Escape Route 1 Micro-Loans for WomenS, R 1 Women's EmpowermentS, R 1 Better Services for Elderly & Disabled 1 Community Unity 1 Land Ownership 1

254

Figure F3. Results of the AHP relative importance values for community priorities that were expressed by 10 or more cases. The large range of importance demonstrates that priority rankings, in addition to priorities, were case-specific.

Table F12. Comparison of total case community scores: Percent difference for the change in score from the Current and Intended Design Scenarios for each case. A positive percentage means that the Intended Design Scenario total case community score was larger than the Current Scenario. Nonfunctional Functional Nonfunctional Resource Functional Resource Conventional Sanitation Conventional Sanitation Recovery Recovery Case # % Difference Case # % Difference Case # % Difference Case # % Difference 4 62% 7 0% 12 115% 11 22% 10 198% 15 0% 20 70% 2 8% 16 105% 3 0% 5 73% 13 0% 17 110% 8 0% 6 59% 14 0% 19 78% 1 0% 9 112% 18 0%

Figure F4. Comparison of the ability of the Current, Intended Design, and Added Resource Recovery Scenario to address community priorities.

255

Table F13. Results of comparison of total case community scores between scenarios for all cases. Compared using one-way ANOVA (p-values are listed in parentheses): bold denotes statistically different (p>0.05) and * denotes not statistically different. The scenario listed in each box is the scenario whose average score was greater than the comparative scenario. Resource Functional Nonfunctional Conventional All Cases Recovery Cases Cases Cases (a) Cases (b) (c) (e) (d) Current vs. Intended Intended Intended Intended Intended Intended Design Design Design Design Design Design (0.003) (0.827)* (0.000) (0.019) (0.009) Current vs. Added Added Added Added Added Added Resource Resource Resource Resource Resource Resource Recovery Recovery Recovery Recovery Recovery Recovery (0.000) (0.235)* (0.000) (0.003) (0.000) Intended Added Added Added Added Added Design vs. Resource Resource Resource Resource Resource Added Recovery Recovery Recovery Recovery Recovery Resource (0.212)* (0.347)* (0.330)* (0.508)* (0.147)* Recovery

F8. Protocol Uncertainty Assessment Results Since some cases expressed multiple resource recovery-related priorities, there is some uncertainty as to how many and which of these priorities would be addressed by adding resource recovery. The small, community-based sanitation systems might produce or recover resources that could not fully and simultaneously address multiple priorities. For example, a biogas digester may not produce enough biogas to satisfy all cooking fuel needs and also generate electricity. However, since it is not certain which of these priorities would be addressed, and the priorities have different

AHP weights, the priority addressment protocol assumes all related priorities are simultaneously addressed. Since this approach measures the maximum potential for priority addressment, an uncertainty analysis was performed. The uncertainty analysis investigated the implications of this approach and compared the results where all related priorities were impacted by adding resource recovery (default) to (1) where only the related priority with the highest AHP weight (of the related priorities) was impacted, called “Only Highest AHP Weight Priority”, and (2) where only the

256

related priority with the lowest AHP weight (of the related priorities) was impacted, called “Only

Lowest AHP Weight Priority” (Figure F6). Overall, total case scores were equal or decreased by an average of 2 ± 6% (Table F14), which was not significantly different.

Figure F5. Comparison of the three scenarios’ average total case sanitation (a) and community (b) scores from all 20 cases when: (i) all biogas-related priorities’ addressed values were influenced by the addition of biogas (green), (ii) only the biogas-related priority with the largest AHP weight was influenced by the addition of biogas (blue), and (iii) only the biogas-related priority with the smallest AHP weight was influenced by the addition of biogas (yellow).

257

Table F14. Comparison of the average and standard deviation of the percent differences between Added Resource Recovery Scenario total case scores when considering uncertainty. Community Priorities Sanitation Priorities Standard Standard Average % Average % Deviation of % Deviation of % Difference Difference Difference Difference

Only Highest AHP Weight -1.2% -7.2% -0.5% -1.1% Priority Biogas Only Lowest AHP Weight 1.4% 8.6% 1.6% 7.3% Priority Only Highest AHP Weight -2.5% -4.6% -0.1% -1.2% Priority Water Only Lowest AHP Weight 2.8% 13.2% 1.0% 4.0% Priority Only Highest AHP Weight 0.0% 0.0% -1.3% 0.2% Priority Compost Only Lowest AHP Weight 0.0% 0.0% 3.7% 2.0% Priority

F9. Supporting Information References Davis, A., Javernick-Will, A., and Cook, S. (2018). “A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems.” Environmental Science: Water Research & Technology, 4, 1451– 1463. Tilley, E., Ulrich, L., Lüthi, C., Reymond, P., and Zurbrügg, C. (2014). Compendium of Sanitation Systems and Technologies. Eawag.

258

Appendix G: Published and Expanded Supporting Information for

Chapter 4

This appendix expands upon the Supplementary Information in the journal article: Davis, A., Javernick-Will, A., and Cook, S. (2019). “The use of qualitative comparative analysis to identify pathways to successful and failed sanitation systems”. Science of the Total Environment. https://doi.org/10.1016/j.scitotenv.2019.01.291.

Please note: Sections G1 – G7 are included in the SI published in the journal article. Section G7 was expanded to provide more calibration details. Section G8 was not included in the published SI and provides additional analysis details. Section G9 is the updated reference list for the expanded SI.

259

G1. Case Summaries Table G1. For all 20 cases: sanitation technology, implementing organization type, operation and maintenance (O&M) manager, outcome, and case summary. Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality; - NGO conducted a limited priority assessment that did not include sanitation-specific priorities; Out-of-State Successful - Community was informed but not involved in planning or Indian NGO (100% Used, decision-making; with prior Individual 91% - Behavior change education activities occurred before and Tamil Conventional sanitation 1 Household 800 Male Operator Maintained, after implementation and included sanitation seminars Nadu (DEWATS) experience Toilets 3 of 3 and mapping to identify open defecation sites and effects; and no prior Regulations - Municipality was involved in planning, decided the site experience in Met) location, and approved the treatment technology; the community - A male operator from community performs daily O&M; - Sanitary engineers from municipality visit system bi- weekly and provide technical assistance; municipality also pays for all O&M costs, which is included in their annual budget. - NGO partnered with community leader to engage community and plan project to stop open defecation and improve environmental health; - NGO did not conduct any priority assessment; - In planning, community members attended meetings and contributed to some design decisions, such as changing Resource Local Indian the biogas digester shape to be sensitive to the protected Recovery Successful NGO with use of the dome shape in Muslim culture; (DEWATS (100% Used, prior - Limited behavior change education occurred, focused with Biogas Individual Implementing 90% sanitation only on the benefits of sanitation; 2 Karnataka Digester; Household 850 Organization Maintained, experience - Municipality gave permission for the project but was not Intended to Toilets Engineers 3 of 3 and no prior otherwise involved. NGO did not attempt to engage recover Regulations experience in municipality and instead engaged community; biogas and Met) community - O&M is performed by NGO because system is right across water) the street from the organization's office and they use it as a showpiece for prospective customers; - NGO pays for O&M; - Water is reused to irrigate a small vegetable farm, and the vegetables are distributed to community; biogas is used as cooking fuel for community’s kindergarten.

260

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality; - NGO did not conduct any priority assessment; - Community was informed but not involved in planning or decision-making; Out-of-State Successful - Behavior change education activities occurred before and Indian NGO (84% Used, after implementation and included sanitation seminars, with prior Individual 90% sanitation games, and discussions of healthy behaviors; Tamil Conventional sanitation 3 Household 800 Male Operator Maintained, - Municipality was involved in the planning of system, Nadu (DEWATS) experience Toilets 2 of 2 decided the site location, and approved the treatment and no prior Regulations technology; experience in Met) - A male operator maintains system, primarily turning the the community pumps on/off daily and removing sewer blockages; - Municipality provides O&M technical and financial assistance; - Community generates income for O&M costs from annual fees collected by the panchayat (community leadership organization), - NGO without prior experience in India implemented system to end open defecation and protect surface water; - NGO conducted a limited priority assessment that did not include sanitation-specific priorities; - Community was not involved in planning and NGO efforts in planning to create a community sanitation committee International Failed were unsuccessful; Conventional NGO with (0% Used, - Limited behavior change education occurred focused only (Baffled Shared sanitation 0% on the benefits of sanitation; 4 Karnataka Settling Tank Communit 1,010 experience None Maintained, - Municipality was uninvolved in planning, and formal + Gravel y Toilets and no prior 0 of 3 system handover was never given to municipality; Filter) experience in Regulations - A caretaker was appointed to clean the toilets but not to the community Met) perform O&M. After one year, community struggled with the toilet water supply and cleaning stopped; - Neither municipality nor NGO provided ongoing O&M or financial assistance for O&M; - Toilets were closed because of O&M problems, and people in community returned to open defecation.

261

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO received international donor funding to implement a prefabricated system and end open defecation; - NGO did not conduct any priority assessment; - Community was not involved in planning. NGO spoke with community leader about a possible sanitation project, but Resource unclear communication meant the leader was unaware Recovery Local Indian Failed the project was going to happen. NGO made no other (Baffled NGO without (22% Used, attempts to engage community in planning; Settling Tank Shared sanitation 0% Tamil - Behavior change education did not occur; 5 + Gravel Communit 800 experience None Maintained, Nadu - Municipality lacked resources to be involved in planning; Filter; y Toilets and no prior 1 of 2 - No operator was appointed; Intended to experience in Regulations - Community receives no technical or financial assistance recover the community Met) from municipality or NGO and lacks the capacity to repair water) system; - The toilets are extremely unclean, and most of community goes for open defecation; - Treated water was supposed to be used for irrigation, but instead discharges to a drainage canal. - NGO implemented system to upgrade a failed system; - NGO conducted a limited priority assessment that did not include sanitation-specific priorities; - In planning, community members were only informed of the project and did not contribute to decision-making, and NGO made no other attempts to engage community in planning; - Behavior change education occurred pre-implementation Resource and included sanitation seminars, theatrical plays to Local Indian Recovery Failed promote positive sanitation behaviors, demonstrations of NGO without (DEWATS (48% Used, healthy behaviors, and educational pamphlets; prior with Biogas Individual Women’s Self- 7% - Municipality lacked resources to be involved in planning; sanitation 6 Karnataka Digester; Household 1,015 Help Group Maintained, - A WSHG was primarily responsible for O&M, but WSHG experience Intended to Toilets (WSHG) 0 of 3 experienced internal conflict and lacked the technical and and with prior recover Regulations financial capacity to continue O&M; experience in biogas and Met) - Income was generated from monthly household fees, but community water) many community members could not pay or refused to pay because of poor O&M; community received no municipal support and only limited technical assistance from NGO; - When community began experiencing water scarcity, O&M problems increased, and most community members returned to open defecation; - Biogas was used for lights in the toilets, and water irrigated plants at the nearby park.

262

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality; - NGO did not conduct any priority assessment; - Community was informed but not involved in planning, and NGO made no other attempts to engage community in Out-of-State planning; Successful Conventional Indian NGO - Behavior change education occurred before and after (95% Used, (Settling with prior implementation and included sanitation seminars, games Individual 100% Tamil Tank + sanitation to identify unsanitary behaviors, and community mapping 7 Household 820 Male Operator Maintained, Nadu Single-Pass experience to identify open defecation sites and effects; Toilets 2 of 2 Intermittent and no prior - Municipality was involved in the planning of system, Regulations Sand Filter) experience in decided the site location, and approved the treatment Met) the community technology; - A dedicated male operator performs daily O&M. Community attributes their system’s success to the operator’s dedication; - Municipality pays the operator’s salary and provides technical assistance when larger problems occur with system. - NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality; - NGO conducted a limited priority assessment that did not include sanitation-specific priorities; - Community was informed but not involved in planning. NGO met with community leader to discuss the project, Out-of-State Successful but he did not contribute further to decision-making; Indian NGO (90% Used, - Behavior change education occurred before and after with prior Individual 100% implementation and included sanitation seminars, Tamil Conventional sanitation Municipality 8 Household 875 Maintained, educational pamphlets, and discussions of healthy Nadu (DEWATS) experience Engineers Toilets 2 of 2 behaviors; and no prior Regulations - Municipality was involved in the planning of system, experience in Met) decided the site location, and approved the treatment the community technology; received formal handover and have collected monitoring data; - Municipality has a highly motivated sanitary inspector, who oversees two engineers who perform system O&M; - Municipality pays the operators and large O&M costs; - Community generates income for O&M costs from annual fees collected by the community leader.

263

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO implemented system to end open defecation; - NGO conducted a limited priority assessment that did not include sanitation-specific priorities; - In planning, community members attended meetings and selected the site; - Behavior change education occurred pre-implementation and included sanitation seminars, discussions of social norms, demonstrations of healthy behaviors, and Local Indian theatrical plays to promote positive sanitation behaviors; Resource Failed NGO without - Municipality was uninvolved in planning because NGO did Recovery (85% Used, prior not engage them; (DEWATS; Individual 20% sanitation - Income was generated from monthly household fees; 9 Karnataka Intended to Household 800 WSHG Maintained, experience community received limited technical assistance from recover Toilets 0 of 3 and with prior NGO and no municipal support; water and Regulations experience in - A WSHG was in charge of O&M, but lacked financial and compost) Met) community technical capacity to overcome system damage and municipal opposition; - Water was used for groundwater recharge and irrigation of a small garden, and vegetables were sold; sludge was land applied on nearby farms; - Municipality constructed 120 new households nearby and connected them to system without making modifications. System is now over capacity and completely flooded; municipality has no plans to repair system. - NGO implemented system to replace damaged septic tanks; - NGO did not conduct any priority assessment; - Community learned of the project during construction, and Out-of-State NGO made no attempts to engage community in planning; Failed Indian NGO - Behavior change education did not occur; (6% Used, without prior - Municipality lacked the resources to be involved in Individual 0% Tamil Conventional sanitation planning; 10 Household 810 Male Operator Maintained, Nadu (DEWATS) experience - A male community member was in charge of O&M, but Toilets 1 of 2 and no prior once the pumps failed, he was unable to perform O&M; Regulations experience in - Community receives no O&M financial or technical Met) the community assistance; - Community has a very strong, negative view of system, indicating that it has “destroyed” their community because of persistent problems system has caused (blockages, smells).

264

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO implemented system to replace damaged community toilets; - NGO conducted an extensive priority assessment that included sanitation-specific priorities; - In planning, community members attended meetings, selected site, determined number of toilets, and designed Resource farm; Local Indian Recovery Successful - Behavior change education occurred pre-implementation NGO with (DEWATS (100% Used, and included sanitation seminars, educational pamphlets, prior with Biogas Shared 92% and community mapping to identify open defecation sites; Tamil sanitation 11 Digester; Communit 1,006 WSHG Maintained, - Municipality lacked resources to be involved in planning Nadu experience Intended to y Toilets 3 of 3 but have since contributed to O&M; and prior recover Regulations - A WSHG receives technical and financial assistance from experience in biogas and Met) municipality; community water) - Income is generated from pay-per-use fees and sale of vegetables from the farm; - Treated effluent is used for irrigation of small farm nearby; system has a biogas digester which was formerly functional and used for cooking fuel for a tea shop run by WSHG. Biogas is no longer in use due to broken pipes that transported the gas from the digester to the tea shop. - NGO implemented system to end open defecation; - NGO did not conduct any priority assessment; - Community was only informed of the project but was not actively involved in planning; - Limited behavior change education occurred focused only on the benefits of sanitation; - Municipality lacked resources to be involved in planning; Resource Local Indian Failed - A male community member was appointed for cleaning Recovery NGO without (32% Used, and O&M but only performs occasional cleaning and no (DEWATS prior Shared 7% O&M; with Biogas sanitation 12 Karnataka Communit 802 Male Operator Maintained, - Community receives no O&M technical or financial Digester; experience y Toilets 0 of 3 assistance; Intended to and with prior Regulations - Income is generated from pay-per-use fees and is used recover experience in Met) for micro-loans for community, so there are insufficient biogas) community O&M funds; - Community never used the biogas produced from system because of bad smells; - System has been damaged by vandalism and community lacks the finances to pay for repairs; - Municipality has threatened to revoke the land lease because of the poor O&M but has not yet acted.

265

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO implemented system as part of an agreement with municipality to add composting to the city’s solid waste management center; - NGO did not conduct any priority assessment; - Community members attended planning meetings; Resource - Behavior change education initiatives were led by Recovery Local Indian Successful community members before and after implementation and (DEWATS NGO with (100% Used, included sanitation seminars, discussions of social norms, with Biogas prior Shared 100% sanitation games, and community mapping to identify Tamil Digester; sanitation 13 Communit 860 Male Operator Maintained, open defecation sites; Nadu Intended to experience y Toilets 3 of 3 - Municipality commissioned the project and attended all recover and prior Regulations meetings; biogas, experience in Met) - A male community member performs O&M; water, and the community - Municipality provides technical and financial assistance; compost) - Biogas is used by two households nearby system for cooking fuel. Water is used for irrigation of banana plants. Sludge is mixed with organic waste from municipality and used to produce compost, which is sold for agriculture, generating more than 100% of O&M costs. - NGO implemented system as a pilot project for ecological sanitation technology with international donor funding; - NGO did not conduct any priority assessment; - In planning, community members attended meetings, selected the site, and approved the final design; Resource - Behavior change education occurred pre-implementation Recovery Local Indian Successful and included sanitation seminars and community mapping (Community NGO with (100% Used, to identify open defecation sites; Ecological prior Shared 100% - Municipality lacked the resources to be involved in Tamil Sanitation sanitation 14 Communit 845 Male Operator Maintained, planning but gave permission for the project; Nadu System; experience y Toilets 3 of 3 - A male community member with a sanitary engineering Intended to and no prior Regulations degree performs O&M; recover experience in Met) - Community receives ongoing technical and financial water and the community assistance from NGO; compost) - System generates income from resource recovery sales; - Fecal waste is mixed with ash to produce compost; anal cleansing water is treated in a gravel filter and used for irrigation for a small nearby farm; urine is collected, mixed with water, and used as fertilizer for another farm.

266

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO was commissioned by the Tamil Nadu Water Supply and Drainage Board to implement system to protect groundwater quality; - NGO conducted a limited priority assessment that did not include sanitation-specific priorities; - Community was informed but not involved in planning Out-of-State Successful because many community members lacked time for Indian NGO (91% Used, planning; with prior Individual 90% - Behavior change education activities were conducted Tamil Conventional sanitation 15 Household 850 Male Operator Maintained, before and after implementation and included sanitation Nadu (DEWATS) experience Toilets 2 of 2 seminars, and demonstrations of healthy behaviors; and no prior Regulations - Municipality was involved in the planning of system, experience in Met) decided the site location, and approved the treatment the community technology; - A male operator performs O&M; - Municipality provides technical and financial assistance; - Community generates income for O&M costs from annual fees collected by the panchayat (community leadership organization). - NGO implemented system to replace damaged septic tanks; - NGO did not conduct any priority assessment; - Community learned of the project during construction, and NGO did not make any attempts to engage community in Out-of-State planning; Failed Indian NGO - Behavior change education did not occur; (31% Used, without prior - NGO did not handover system to municipality, so Individual 3% Tamil Conventional sanitation municipality refuses to assist with O&M; 16 Household 805 None Maintained, Nadu (DEWATS) experience - An operator was not appointed, so no O&M has been Toilets 1 of 2 and no prior performed; Regulations experience in - Municipality also lacks the human and financial resources Met) the community to provide support for system; - Community members state that the construction quality is poor because the contractors are only interested in profit and rushed to complete the job, and community was not able to make meaningful changes during construction; - Most of community practices open defecation.

267

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO without prior experience in India implemented system to end open defecation and protect surface water; - NGO conducted a limited priority assessment that did not include sanitation-specific priorities; - Community was not involved in planning, and efforts to create a community sanitation committee were International Failed unsuccessful; NGO with (88% Used, - Limited behavior change education occurred focused only Individual sanitation 11% Conventional on the benefits of sanitation; 17 Karnataka Household 1,010 experience None Maintained, (DEWATS) - Formal handover never was given to municipality, who Toilets and no prior 0 of 3 was also uninvolved in planning, so they refused to experience in Regulations provide O&M technical and financial assistance; the community Met) - No operator was appointed so no O&M was performed; - NGO provided no technical or financial assistance; - System was never desludged, and the pumps broke very early on in the few short months of operation; - Toilet wastewater now discharges directly to the nearby lake. - NGO implemented system to replace damaged community toilets and improve environmental health; - NGO conducted an extensive priority assessment that included sanitation-specific priorities; - In planning, community members attended meetings, selected the site, requested child-friendly toilets, and helped design community kitchen; - Behavior change education was led by community Resource members before and after implementation and included Local Indian Recovery Successful sanitation seminars, discussions of healthy behaviors, NGO with (DEWATS (100% Used, discussions of social norms, theatrical plays to promote prior with Biogas Shared 100% positive sanitation behaviors, sanitation games, and Tamil sanitation 18 Digester; Communit 1,000 WSHG Maintained, community mapping to identify open defecation sites; Nadu experience Intended to y Toilets 3 of 3 - Municipality lacked resources to be involved in planning and prior recover Regulations but is now supportive; experience in biogas and Met) - A WSHG is in charge of O&M of system; the community water) - Municipality pays for major O&M costs and performs O&M for the biogas digester; - WSHG generates income for O&M from pay-per-use fees and employs over 30 women monthly; - Biogas is used for a free community kitchen with 20 cookstoves; - WSHG generated enough income from toilets to construct water storage tank to improve community’s water supply and to construct community hall.

268

Number Toilet Implementing Case Sanitation of O&M State Connect- Organization Outcome Case Summary # Technology Intended Manager ion Type Users - NGO implemented system to end open defecation; - NGO did not conduct any priority assessment; - Community learned of the project during construction, and NGO did not make any attempts to engage community in International Failed planning; NGO without (82% Used, - Behavior change education did not occur; Individual sanitation 60% - Municipality was not involved in planning. Case 18 Tamil Conventional 19 Household 1,015 experience Male Operator Maintained, technically falls under village panchayat authority, Nadu (DEWATS) Toilets and no prior 1 of 3 although neighboring communities are under municipal experience in Regulations authority. There is a lack of clarity of which government the community Met) agency is responsible for community, so community receives no assistance; - An unskilled male operator runs the pumps, and some households desludge the tanks on an emergency basis, when tanks start to overflow. - NGO implemented system to end open defecation; - NGO conducted an extensive priority assessment that included sanitation-specific priorities; - In planning, community members attended meetings, went on exposure visits, and the WSHG was actively involved in system design and site selection; - Behavior change education occurred post-implementation Resource and included sanitation seminars, sanitation games, Local Indian Recovery Failed demonstrations of healthy behaviors, and theatrical plays NGO with (DEWATS (0% Used, to promote positive sanitation behaviors; prior with Biogas Shared 0% - Municipality was not involved in planning but agreed to sanitation 20 Karnataka Digester; Communit 850 WSHG Maintained, lease the land for system; experience Intended to y Toilets 0 of 3 - A WSHG who received training and had technical and prior recover Regulations knowledge was in charge of O&M; experience in biogas and Met) - WSHG generated income from pay-per-use fees and the community water) resource recovery. Biogas was used to heat water which was sold for bathing. Treated water was sold for construction and sludge was sold for agriculture application; - Municipality revoked the land lease and appointed their own employees to manage system and its revenue. Municipality failed to perform O&M and eventually locked toilets.

269

G2. Respondent Demographics Table G2. Summary of respondent demographics for interviews conducted to characterize causal conditions and outcomes for all 20 cases. Sex Role Number Implementing Community Community Case # of Male Female Mixed Municipality Operator WSHG Organization Leader Member Interviews 1 24 43% 50% 7% 11% 17% 7% 43% 0% 32% 2 25 52% 44% 4% 8% 15% 8% 72% 4% 0% 3 23 48% 43% 9% 9% 18% 9% 61% 13% 0% 4 26 48% 52% 0% 11% 15% 7% 44% 0% 30% 5 24 57% 36% 7% 11% 13% 7% 64% 11% 0% 6 26 60% 40% 0% 8% 13% 0% 76% 8% 0% 7 25 58% 38% 4% 8% 14% 8% 75% 0% 0% 8 26 46% 46% 8% 12% 17% 8% 73% 0% 0% 9 27 46% 46% 8% 8% 17% 8% 50% 0% 27% 10 24 44% 44% 7% 11% 17% 7% 48% 0% 26% 11 28 48% 36% 8% 8% 17% 8% 60% 16% 0% 12 23 50% 35% 8% 8% 15% 8% 65% 12% 0% 13 28 54% 46% 0% 8% 15% 8% 63% 13% 0% 14 25 65% 35% 0% 9% 13% 9% 70% 4% 0% 15 26 56% 36% 8% 8% 14% 8% 76% 0% 0% 16 25 54% 38% 8% 8% 14% 8% 69% 8% 0% 17 24 54% 33% 13% 13% 15% 8% 71% 0% 0% 18 28 65% 35% 0% 9% 13% 9% 65% 9% 0% 19 23 58% 33% 8% 8% 14% 0% 71% 13% 0% 20 27 57% 46% 4% 11% 13% 0% 54% 0% 29% Average 25 53% 41% 5% 9% 15% 7% 63% 5% 7%

270

G3. System Performance and Regulations Table G3. Summary of system performance based on water quality analyses†. COD BOD pH Case System Discharge (mg/L) (mg/L) Number Outcome Type Regulation* Influent Effluent Regulation Influent Effluent Regulation Influent Effluent 1 Successful Surface Water 250 469 54 30 223 26 5.5-9.0 7.90 8.10 2 Successful Irrigation/UGD Not Regulated 488 23 100 232 11 5.5-9.0 7.11 7.23 3 Successful Coastal 250 403 194 100 192 92 5.5-9.0 7.33 7.61 Samples not collected - Samples not collected - Samples not collected - 4 Failed UGD Not Regulated 350 5.5-9.0 no wastewater in system no wastewater in system no wastewater in system Irrigation/ 5 Failed Not Regulated 2327 258 100 1108 123 5.5-9.0 6.98 7.43 Surface Water Samples not collected - Samples not collected - Samples not collected - 6 Failed Irrigation/UGD Not Regulated 100 5.5-9.0 no wastewater in system no wastewater in system no wastewater in system 7 Successful Coastal 250 275 91 100 131 43 5.5-9.0 7.01 7.41 8 Successful Surface Water 250 388 46 30 185 22 5.5-9.0 7.31 7.65 Samples not collected - Samples not collected - Samples not collected - 9 Failed Irrigation/UGD Not Regulated extensive flooding 100 extensive flooding 5.5-9.0 extensive flooding prevented access prevented access prevented access 10 Failed Coastal 250 402 283 100 191 135 5.5-9.0 6.50 7.20 11 Successful Irrigation/UGD Not Regulated 2688 91 100 1280 43 5.5-9.0 7.12 7.34 Samples not collected - Samples not collected - Samples not collected - extensive system extensive system 12 Failed UGD Not Regulated 350 extensive system damage 5.5-9.0 damage prevented damage prevented prevented access access access 13 Successful Irrigation/UGD Not Regulated 1897 62 100 903 30 5.5-9.0 6.80 7.20 14 Successful Irrigation/UGD Not Regulated 106 17 100 50 8 5.5-9.0 7.30 7.32 15 Successful Coastal 250 270 89 100 129 42 5.5-9.0 7.66 8.01 16 Failed Coastal 250 308 267 100 147 127 5.5-9.0 7.23 7.83 Samples not collected - Samples not collected - Samples not collected - 17 Failed Surface Water 250 30 5.5-9.0 no wastewater in system no wastewater in system no wastewater in system 18 Successful Irrigation/UGD Not Regulated 2987 61 100 1422 29 5.5-9.0 6.90 7.04 19 Failed Surface Water 250 490 383 30 233 182 5.5-9.0 6.40 7.30 Samples not collected - Samples not collected - Samples not collected - 20 Failed UGD Not Regulated 350 5.5-9.0 no wastewater in system no wastewater in system no wastewater in system †Influent and effluent samples were collected and analyzed in January – April 2017. *In India, COD is regulated only for surface and coastal water discharge (Central Pollution Control Board 2017).

271

G4. Hypothesized Causal Conditions Table G4. List of hypothesized causal conditions from literature and case knowledge that may influence sanitation system success or failure. Hypothesized Causal Condition Included Excluded* Domain (Constant) Addressed Sanitation Priorities x Behavior Change Education x Clear O&M Plan x Community Financial Contributions to Capital Costs x Community Participation in Construction x Community Participation in O&M Minimized Community Participation in Planning x Community Socio-Economic Status x Combined with Construction Oversight Construction Quality Construction Quality x Culture x Combined with External Technical Assistance Technical Support Combined with External Financial Assistance Sufficient O&M Funds Government Barriers x Combined with Income Generation Sufficient O&M Funds Municipality Involved in Planning x O&M Training Low Necessity Organization Embeddedness Minimized Organization Sanitation Experience Minimized Past Sanitation Experience (Community) Low Necessity Priority Assessment Low Necessity Regulations x Combined with Resource Recovery Income Sufficient O&M Funds Sanitation System Age x Sanitation System Capital Costs x Sanitation System Size x Combined with Skilled Operator Technical Support Sufficient O&M Funds x Technical Support x Technology Complexity x Technology Type Low Necessity Vandalism/Damage Low Necessity *Exclusion Reasons: Combined = causal condition was combined into another condition included in the analysis due to similarities/overlap. For example, Construction Oversight was combined with Construction Quality. Low Necessity = causal condition was removed because its necessity score was less than 0.3, a conventional cutoff for condition exclusion (Opdyke 2017), and because case knowledge indicated that it was not an important driver of success or failure. Minimized = causal condition was removed during the analysis because logically simplified pathways without the condition were more consistent (consistency ≥ 0.8), and case knowledge supported the condition’s exclusion from the pathway.

272

G5. Indirectly Calibrated Causal Conditions Seven of the nine causal conditions were calibrated indirectly, primarily with qualitative data, using procedures detailed by Basurto and Speer (2012). Causal condition definitions and calibration details are summarized below. Table G13 summarizes the calibrations.

Behavior Change Education. Behavior Change Education is defined as the use of behavior change theory to teach community members the benefits of sanitation and to reduce open defecation. Sanitation education founded in behavior change theory has been demonstrated to be important for ensuring the use of sanitation systems (Mosler 2012; Rosenquist 2005; Tilley et al.

2014a). Behavior Change Education has been applied and evaluated primarily in rural contexts, such as the use of community-led total sanitation campaigns to trigger household adoption of pit latrines (Venkataramanan et al. 2018). Despite the limited evaluation of Behavior Change

Education in the urban and peri-urban contexts, implementing organizations are using behavior change theory to develop their training programs (Scope 2012; Wegelin-Schuringa 2000).

Furthermore, research has shown Behavior Change Education to increase sanitation use and success (Cairncross et al. 2005; Jenkins and Cairncross 2010; Jenkins and Scott 2007) as well as have no impact on sanitation outcomes (Ramani et al. 2012; Venkataramanan et al. 2018), suggesting that the causal condition alone is insufficient.

In-set membership is when intensive sanitation education, based on behavior change theory, was provided to a majority of community members and discussed proper toilet use, basic sanitation concepts, health and hygiene, and the importance of sanitation (Table G5). Behavior change activities included seminars, discussions of healthy behaviors and/or social norms, educational pamphlets, theatrical plays, games, community mapping to identify open defecation sites (Mosler 2012). Out-of-set membership is when no sanitation education was provided.

273

Table G5. Indirect calibration for Behavior Change Education. Intensive sanitation education was provided to a majority of community members that 1 focused on proper toilet use, basic sanitation concepts, health and hygiene, and the importance of sanitation. Behavior change theory was the basis for the training. Moderate amount of sanitation education was provided to a majority of community 0.67 members that focused on proper toilet use and basic sanitation concepts such as health and hygiene but lacked an explicit foundation in behavior change theory. Limited sanitation education was provided to a minority of community members. 0.33 Sanitation education only focused on one concept such as proper toilet use. 0 No sanitation education was provided to community members.

Clear O&M Plan. Clear O&M Plan is defined as the presence of a plan where all required maintenance tasks are known, and all stakeholders agree on whose responsibility it is to perform and finance each task. Conflicting perceptions of ownership and responsibility can lead to lack of sanitation system O&M (Brikké and Bredero 2003; Chatterley et al. 2014). Eales et al. (2013) found that in the absence of a formal agreement, most communities did not view sanitation system maintenance as their responsibility, despite implementing organizations “transferring” responsibility. Additionally, the lack of clear O&M responsibility is thought to also negatively affect system finances (IDECK 2015). A Clear O&M Plan can be developed through a formal handover, a formally signed memorandum of understanding or through meetings that achieve the same result.

In-set membership is when O&M responsibility is known for all tasks and unanimously agreed upon by all stakeholders (Table G6). Out-of-set membership is when no O&M plan exists, no discussions of O&M responsibility occurred, and there is disagreement over responsibility for all O&M tasks.

Table G6. Indirect calibration for Clear O&M Plan. Formal handover occurred, which included a formal memorandum of understanding or written 1 maintenance plan to detail O&M responsibilities. O&M responsibility is known for all tasks and agreed upon by all stakeholders. Majority of maintenance tasks have clear assignments that were accepted. There may be some 0.67 disagreement over stakeholder responsibilities for smaller tasks, such as clearing sewer blockages. A formal, written O&M plan is missing. No O&M plan exists. Discussions occurred regarding O&M responsibility, but there is 0.33 disagreement over stakeholder responsibilities for most maintenance tasks. No O&M plan exists. No discussions for O&M responsibility occurred, and there is 0 disagreement over stakeholder responsibilities for all O&M tasks.

274

Community Participation in Planning. Community Participation in Planning is defined as regular and meaningful involvement of community members in planning, which includes attending meetings and helping make decisions such as site selection and appropriate technology selection. Community Participation in Planning is a major focus of many sanitation planning frameworks (Luthi et al. 2011; Palaniappan et al. 2008; Törnqvist et al. 2008; WaterAid 2011b).

However, participation has been shown to have both positive (Black 1998; Eales et al. 2013;

Prokopy 2005; Roma and Jeffrey 2010) and negative (Chowns 2015; Marks and Davis 2012) impacts on outcomes in water and sanitation infrastructure. This may be due to the limited power of community members in many planning processes (Arnstein 1969; Bao et al. 2013) or due to the presence or absence of other important causal conditions (Bouabid and Louis 2015; Palaniappan et al. 2008). For this work, the planning phase includes all project activities that occur prior to construction: community mobilization, goal setting, project initiation, project conceptualization, stakeholder meetings, division of responsibilities, site selection, and technology selection.

In-set membership is when community members have high citizen power (Arnstein 1969;

Bao et al. 2013; Wegelin-Schuringa 2000), are actively engaged throughout the entire planning process, and had decision-making power over multiple decisions such as goal-setting, site selection, and technology selection (Table G7). Out-of-set membership is when community members were entirely uninvolved in the planning process (Arnstein 1969; Bao et al. 2013;

Wegelin-Schuringa 2000).

275

Table G7. Indirect calibration for Community Participation in Planning. High citizen power. Community members were actively engaged throughout the entire planning process and were in charge of informing other community members through regular meetings or door to door visits. Community members attended planning activities beyond informational 1 meetings, such as exposure visits, trainings, or planning meetings for management and maintenance. Community members had decision-making power over multiple decisions, such as goal-setting, site selection, and technology selection. Moderate citizen power. Community members were informed about the sanitation project through regular meetings or door to door visits. Community members had limited decision-making power 0.67 (such as site selection only). Minority of community members were not involved and could not attend meetings. Tokenism. Community members were informed about the sanitation project through meetings or door to door visits, but meetings were not held regularly, and community members had difficulty 0.33 attending them. Community members were not involved in any decision-making or meetings beyond informational sessions and provided no input during project planning. Non-participation. The community is entirely uninvolved with planning for the sanitation system 0 and learned of the project at the time of construction.

Construction Quality. Construction Quality is defined as a sanitation system that is well- constructed based on high material quality and correct implementation. Construction quality has been shown to influence well- and poorly-managed sanitation systems, but a prior QCA study

(Chatterley et al. 2013, 2014). found that construction quality alone was insufficient to result in either outcome. Poor-quality construction has a negative impact on system performance and can occur in the absence of construction oversight (Chowns 2015; Mansuri 2004).

In-set membership is when the system was constructed with high-quality materials, has no construction errors, and construction quality has no negative impact on performance and maintenance (Opdyke et al. 2018) (Table G8). Out-of-set membership is when the system was constructed with low-quality materials and has significant construction errors (e.g., incomplete effluent discharge pipes, no concrete reinforcement, incorrectly graded sewer lines) that can negatively affect performance and increase maintenance.

276

Table G8. Indirect calibration for Construction Quality. System was constructed with high-quality materials and there were no errors in construction. 1 Construction quality does not affect system performance or maintenance requirements. System was constructed with regular materials and there were no errors or insignificant errors in 0.67 construction. Construction quality does not affect treatment system performance or maintenance requirements. System was constructed with regular materials, but there were significant errors in construction 0.33 that adversely impacted system performance and increased maintenance requirements. System was constructed with materials of lower quality than called for in design specifications 0 and there were numerous, significant errors in construction that adversely impacted system performance and increased maintenance requirements.

Government Barriers. Government Barriers is defined as deliberate actions taken by the municipality that prevent or disrupt sanitation system use, maintenance, or performance.

Government Barriers can arise when there is a lack of political will (Sansom 2011); an underlying political motivation, such as a desire for recognition, that could lead to the disruption of NGO-led initiatives (Brikké and Bredero 2003); corruption (Sansom 2011); conflicting interests (Harris et al. 2011); or government exclusion from planning (Harris et al. 2011). Government Barriers can have a strong influence on sanitation failure (Devas and Grant 2003), but it remains unclear if there exist pathways that are sufficient to overcome these barriers.

In-set membership is when the local municipality actively opposes the sanitation system through clear, deliberate actions that prevent continued use, maintenance, and performance (Table

G9). Out-of-set membership is when the local municipality has not actively opposed the sanitation system.

Table G9. Indirect calibration for Government Barriers. Local municipality has taken significant actions to actively oppose the sanitation system. 1 This opposition disrupted maintenance and performance and prevented the community from using the system. 0 Local municipality has not actively opposed the sanitation system.

Municipality Involved in Planning. Municipality Involved in Planning is defined as the local municipality having regular and meaningful involvement in planning, which includes attending meetings and helping to make decisions such as site selection and appropriate technology selection. Municipalities are considered to be essential stakeholders in the planning and

277

implementation of successful sanitation systems (Harris et al. 2011; Kooy and Harris 2012).

Municipalities in India have the authority to approve sanitation projects, lease land, and regulate systems (Ministry of Urban Development 2008).

In-set membership is when the municipality is actively engaged throughout the entire planning process and had decision-making power over multiple decisions such as goal-setting, site selection, and technology selection (Table G10). Out-of-set membership is when the municipality was entirely uninvolved in the planning process.

Table G10. Indirect calibration for Municipality Involved in Planning. Local municipality was actively involved in the sanitation system planning. Local municipality gave 1 permission for the project, was involved in the planning meetings, and had decision-making power over multiple decisions such as goal-setting, site selection, and technology selection. Local municipality was informed of the sanitation system during the planning phase and attended 0.67 most planning meetings. Local municipality helped to make one smaller decision, such as site selection. Local municipality was informed of the sanitation system during the planning phase and attended 0.33 some planning meetings but was not regularly engaged and did not make any decisions. Local municipality was not involved in the planning of the sanitation system. Local municipality did 0 not give permission, refused to be involved, or was not informed of the system until after it was completely planned and designed.

Technical Support. Technical Support is defined as adequate technical capacity for maintenance that is available through the presence of both a skilled operator and external maintenance assistance. Sanitation planning frameworks highlight the importance of well-trained operators for maintenance and system performance (IDECK 2015; Sakthivel et al. 2014; Tilley et al. 2014a). Research also demonstrates that unskilled operators are less likely to perform maintenance correctly and on-time, leading to reduced performance (Katukiza et al. 2012; Mwirigi et al. 2014). Additionally, when infrastructure is maintained by community-based groups or individuals, maintenance and performance improve when the community receives technical assistance (Chatterley et al. 2014; Eales et al. 2013). Conversely, lack of technical assistance has been shown to negatively influence technology acceptance (Uddin et al. 2014). Research further demonstrates that the source of assistance (municipality or implementing organization) is

278

inconsequential compared to the type and amount of assistance (Eales et al. 2013; Jordan et al.

2016; Uddin et al. 2014).

Technical Support is the minimum score of Skilled Operator and External Technical

Assistance, as both are needed for sanitation maintenance and performance. For Skilled Operator, in-set membership is when a proactive operator is regularly present who demonstrates extensive technical system knowledge and is capable of recognizing and rectifying system problems independently (Table G11). Out-of-set membership is when there is no operator.

Table G11. Indirect calibration for Skilled Operator. Skilled operator demonstrates extensive technical system knowledge; is capable of 1 recognizing and rectifying system problems independently; and is proactive. Operator has some technical knowledge but needs external assistance to diagnose and 0.67 recognize system problems. Unskilled operator exists and is not present daily at the system or is just responsible for 0.33 toilet cleaning. Operator is unreliable and cannot diagnose or recognize system problems. 0 No operator exists.

For External Technical Assistance, in-set membership is when the municipality or implementing organization provides extensive ongoing technical assistance continuously and is involved in all technical maintenance tasks (i.e., all tasks that require a skilled operator and have an influence on treatment performance) (Table G12). Out-of-set membership is when neither the municipality nor the implementing organization provides any maintenance assistance post- implementation.

Table G12. Indirect calibration for External Technical Assistance. Municipality/Organization provides extensive ongoing technical assistance continuously, 1 where the municipality/organization is involved in all technical maintenance tasks (i.e., all tasks that require a skilled operator and have an influence on treatment performance). Municipality/Organization provides moderate ongoing assistance for at least half of the 0.67 required maintenance tasks more than once annually. Municipality/Organization provides extremely limited technical assistance, where assistance 0.33 occurs at most once a year and completes less than half of the required maintenance tasks. Municipality/Organization provides no maintenance assistance to community after 0 construction.

279

Table G13 summarizes the seven indirect calibrations for the causal conditions. Table G13. Summary of causal conditions’ indirect calibrations. Causal Calibration Condition 0 No sanitation education was provided to community members. Limited sanitation education was provided to a minority of community members. Sanitation 0.33 education only focused on one concept such as proper toilet use. Behavior Moderate amount of sanitation education was provided to a majority of community members Change 0.67 that focused on proper toilet use and basic sanitation concepts such as health and hygiene but Education lacked an explicit foundation in behavior change theory. Intensive sanitation education was provided to a majority of community members that focused 1 on proper toilet use, basic sanitation concepts, health and hygiene, and the importance of sanitation. Behavior change theory was the basis for the training. No O&M plan exists. No discussions for O&M responsibility occurred, and there is 0 disagreement over stakeholder responsibilities for all O&M tasks. No O&M plan exists. Discussions occurred regarding O&M responsibility, but there is 0.33 disagreement over stakeholder responsibilities for most maintenance tasks. Clear O&M Majority of maintenance tasks have clear assignments that were accepted. There may be some Plan 0.67 disagreement over stakeholder responsibilities for smaller tasks, such as clearing sewer blockages. A formal, written O&M plan is missing. Formal handover occurred, which included a formal memorandum of understanding or written 1 maintenance plan to detail O&M responsibilities. O&M responsibility is known for all tasks and agreed upon by all stakeholders. Non-participation. The community is entirely uninvolved with planning for the sanitation 0 system and learned of the project at the time of construction. Tokenism. Community members were informed about the sanitation project through meetings or door to door visits, but meetings were not held regularly and community members had difficulty 0.33 attending them. Community members were not involved in any decision-making or meetings beyond informational sessions and provided no input during project planning. Moderate citizen power. Community members were informed about the sanitation project Community through regular meetings or door to door visits. Community members had limited decision- Participation 0.67 making power (such as site selection only). Minority of community members were not involved in Planning and could not attend meetings. High citizen power. Community members were actively engaged throughout the entire planning process and were in charge of informing other community members through regular meetings or door to door visits. Community members attended planning activities beyond informational 1 meetings, such as exposure visits, trainings, or planning meetings for management and maintenance. Community members had decision-making power over multiple decisions, such as goal-setting, site selection, and technology selection. System was constructed with materials of lower quality than called for in design specifications 0 and there were numerous, significant errors in construction that adversely impacted system performance and increased maintenance requirements. System was constructed with regular materials, but there were significant errors in construction 0.33 Construction that adversely impacted system performance and increased maintenance requirements. Quality System was constructed with regular materials and there were no errors or insignificant errors in 0.67 construction. Construction quality does not affect treatment system performance or maintenance requirements. System was constructed with high-quality materials and there were no errors in construction. 1 Construction quality does not affect system performance or maintenance requirements. 0 Local municipality has not actively opposed the sanitation system. Government Local municipality has taken significant actions to actively oppose the sanitation system. This Barriers 1 opposition disrupted maintenance and performance and prevented the community from using the system.

280

Causal Calibration Condition Local municipality was not involved in the planning of the sanitation system. Local municipality 0 did not give permission, refused to be involved, or was not informed of the system until after it was completely planned and designed. Local municipality was informed of the sanitation system during the planning phase and 0.33 attended some planning meetings but was not regularly engaged and did not make any Municipality decisions. Involved in Local municipality was informed of the sanitation system during the planning phase and Planning 0.67 attended most planning meetings. Local municipality helped to make one smaller decision, such as site selection. Local municipality was actively involved in the sanitation system planning. Local municipality gave permission for the project, was involved in the planning meetings, and had decision- 1 making power over multiple decisions such as goal-setting, site selection, and technology selection. Technical Support takes the minimum of the scores for Skilled Operator and External Technical Assistance Skilled Operator External Technical Assistance Municipality/Organization provides no 0 No operator exists. 0 maintenance assistance to community after construction. Unskilled operator exists and is not Municipality/Organization provides present daily at the system or is just extremely limited technical assistance, responsible for toilet cleaning. 0.33 0.33 where assistance occurs at most once a year Operator is unreliable and cannot and completes less than half of the required diagnose or recognize system maintenance tasks. Technical problems. Support Operator has some technical Municipality/Organization provides knowledge but needs external moderate ongoing assistance for at least 0.67 0.67 assistance to diagnose and recognize half of the required maintenance tasks more system problems. than once annually. Municipality/Organization provides Skilled operator demonstrates extensive ongoing technical assistance extensive technical system continuously, where the 1 knowledge; is capable of recognizing 1 municipality/organization is involved in all and rectifying system problems technical maintenance tasks (i.e., all tasks independently; and is proactive. that require a skilled operator and have an influence on treatment performance).

281

G6. Directly Calibrated Causal Conditions Addressed Sanitation Priorities. Addressed Sanitation Priorities is defined as a sanitation system that addresses a majority of the community’s sanitation priorities. There is consensus in sanitation research and practice that appropriate technology selection, where sanitation systems are designed to meet specific user needs and preferences (i.e., priorities), can foster greater user acceptance and system success (Breslin 2003; Hacker and Kaminsky 2017; Seymour 2014;

Sperling et al. 2016). While studies demonstrate the importance of appropriate technology selection (Black 1998; Murphy et al. 2009; Palaniappan et al. 2008), there has been little research that investigates the role of appropriate technology in combination with other important causal conditions. Additionally, since there is only limited research that evaluates how well sanitation technologies addressed user priorities post-implementation (Davis et al. 2019a), the influence of

Addressed Sanitation Priorities on system outcomes is unknown. Addressed Sanitation Priorities was calibrated directly (Figure G1) using total case sanitation scores, which reflect the extent to which priorities are addressed based on importance; these scores were calculated using the priority addressment protocol for each case (Davis et al. 2019a). In-set membership is when the total case sanitation score is greater than 0.8. Out-of-set membership is when the total case sanitation score is less than 0.2. The crossover point is when the total case sanitation score is 0.4, which reflects a break between the case scores.

282

(a) (b) Figure G1. (a) Anchor and crossover points for direct calibration for Addressed Sanitation Priorities. The green line represents the threshold for in-set membership; the yellow line represents the cross-over point; the red line represents the threshold for out-of-set membership. (b) Direct Calibration for Addressed Sanitation Priorities.

Sufficient O&M Funds. Sufficient O&M Funds is defined as funds that are available from income generation or external financial assistance equal to or in excess of the system’s O&M costs.

Aside from technical factors, economic factors are the most often cited factors under consideration for the causes of sanitation outcomes (Kaminsky and Javernick-Will 2012) and are shown to be important for well-planned and successful systems (Bouabid and Louis 2015; Tilley et al. 2014a;

Törnqvist et al. 2008). Studies indicate that financial difficulty contributed to sanitation failure

(Cronin et al. 2014; Eales et al. 2013). Other research suggests that income generation (Mwirigi et al. 2014), external financial assistance (Starkl et al. 2013b), and/or offset costs from resource recovery (Katukiza et al. 2012) can improve sanitation affordability and maintenance. Income generation occurred from pay-per-use or monthly user fees or from the sale of resources recovered

(e.g., tea made using biogas for cooking fuel, bathing water heated using biogas, vegetables grown using recycled water, or compost). External financial assistance came from either the implementing organization or the municipality. Sufficient O&M Funds was calibrated directly

(Figure G2) where the raw data was the total of monthly income generation and external financial assistance as a percentage of monthly O&M costs. In-set membership is when funds available for

O&M are equal to or exceed 100% of O&M costs. Out-of-set membership is when no funds are

283

available for O&M. The crossover point is when funds available for O&M equal 50% of O&M costs, which reflects a break between the case scores.

(a) (b) Figure G2. (a) Anchor and crossover points for direct calibration for Sufficient O&M Funds. The green line represents the threshold for in-set membership; the yellow line represents the cross-over point; the red line represents the threshold for out-of-set membership. (b) Direct Calibration for Sufficient O&M Funds.

G7. Outcome Calibrations Success and Failure. Success was defined as the presence of three criteria (Davis et al.

2018b, 2019a): (1) the system is used by at least 75% of the community; (2) at least 90% of maintenance tasks are performed correctly and on time; and (3) the system complies with local regulations for pH, COD, and BOD. Cases were classified as failed if they did not meet one or more of the success criteria. Success outcome scores were determined by taking the minimum of the fuzzy set scores for performance, use, and maintenance. Failure outcome scores were the negated (i.e., absence) of the success scores (i.e., 1 – success).

Performance. Performance was indirectly calibrated using a three-value fuzzy set (Table

G14). Performance refers to the adequate containment and treatment of wastewater, according to applicable regulatory standards and treatment system design. The three parameters regulated by the Central Pollution Control Board (CPCB) are chemical oxygen demand (COD), biological oxygen demand (BOD) and pH (Central Pollution Control Board 2017). Water quality parameters

284

are based on the discharge type: surface water, coastal water, underground drainage (UGD, which is the term for municipal sewer), and irrigation.

In-set membership was defined as complying with all applicable pH, BOD, and COD regulations while out-of-set membership was defined as failing to comply with all three regulations; an intermediate value of 0.3 was defined as a system failing to comply with only one regulation.

Table G14. Indirect Calibration for Performance. Outcome Calibration 0 System is not containing or treating waste (does not meet two or more water quality regulations). System contains waste but fails to meet one water quality regulation OR Performance 0.3 system meets all water quality regulations but fails to contain waste. 1 System contains and treats waste, meeting all water quality regulations.

Use. Use was directly calibrated (Figure G3). Use is defined as the percent of the intended population that uses the sanitation system correctly, daily, and exclusively. While the goal is to attain complete coverage and use of sanitation to minimize exposure to fecal pathogens, health studies indicate that the benefits of increased use increase significantly up to a threshold of 75% use (Andres et al. 2014; Harris et al. 2017). Conversely, studies have found negligible benefits when use is less than 25% (Andres et al. 2014; Harris et al. 2017).

In-set membership was defined as more than 75% of the system’s target population using the system correctly, daily, and exclusively (i.e., no open defecation) (Andres et al. 2014; Harris et al. 2017) while out-of-set membership was defined as less than 25% using it correctly, daily, and exclusively.

285

(a) (b) Figure G3. (a) Anchor and crossover points for direct calibration for Use. The green line represents the threshold for in-set membership; the yellow line represents the cross-over point; the red line represents the threshold for out-of-set membership. (b) Direct Calibration for Use.

Maintenance. Maintenance was directly calibrated (Figure G4). Maintenance refers to all of the tasks required for functional performance of the treatment system and the toilet facilities. Lack of preventative maintenance can increase costs, reduce performance, and/or inhibit use of the sanitation system (IDECK 2015).

In-set membership was defined as at least 90% of the total required maintenance tasks were completed correctly and on time (Brikké 2000; Eales et al. 2013) while out-of-set membership was defined as less than 25% completed correctly and on time.

(a) (b) Figure G4. (a) Anchor and crossover points for direct calibration for Maintenance. The green line represents the threshold for in-set membership; the yellow line represents the cross-over point; the red line represents the threshold for out-of-set membership. (b) Direct Calibration for Maintenance.

286

G8. Expanded QCA Analytical Procedures G8.1 Overview The remainder of this appendix provides a detailed summary of the analyses of the causes of sanitation success and failure for Chapter 4. This section provides additional details on analysis procedures including simplifying assumptions, necessity and sufficiency for each causal condition, subset/superset analyses, and the final identification of the solution pathways.

G8.2 Preliminary Minimization and Removal of Causal Conditions First, I assembled a list of hypothesized causal conditions from literature and case knowledge (Table G4). Thirty-two causal conditions is too many for 20 cases in QCA (Ragin

2008); we removed conditions based on: lack of variation across the cases within a condition

(known as domain conditions); correlations with other conditions, indicating that two conditions may be measuring the same item; or low necessity, indicating that the condition was less important for the outcome.

Nine conditions from the initial list were determined to be domain conditions because they were relatively constant across all 20 cases and were removed from analysis. These domain conditions included: Community Financial Contributions to Capital Costs (overall were nonexistent or negligible); Community Participation in Construction (generally minimal);

Community Socio-Economic Status (low income communities with high unemployment and high percentages of individuals below the poverty line and/or in India’s lowest castes); Culture (all cases were from Karnataka or Tamil Nadu, India); Regulations (all systems were regulated by the

Central Pollution Control Board with similar treatment standards); Sanitation System Age (all systems were implemented between 2008 and 2010); Sanitation System Capital Costs (capital costs ranged from 1500-3500 rupees/person ($21-50) and were nearly entirely subsidized);

287

Sanitation System Size (all systems were designed to serve 800-1000 users); and Technology

Complexity (all systems had a similar number of treatment tanks and required maintenance tasks).

Six potential causal conditions were combined with others due to correlations/similarities:

Construction Oversight was combined with Construction Quality because the presence of oversight often led to corrections in construction errors, and the lack of oversight meant that construction contractors used lower quality materials and deviated from designs; External

Financial Assistance, Income Generation, and Resource Recovery Income were combined into

Sufficient O&M Funds because case knowledge demonstrated that the availability of funds for

O&M was more important than the source of those funds; and External Technical Assistance and

Skilled Operator were combined into Technical Support because case knowledge indicated that these sub-conditions interchangeably satisfied the need for local capacity for maintenance.

Five hypothesized causal conditions were removed because their necessity scores were below 0.3 (Opdyke et al. 2018) and because case knowledge indicated that they were not main drivers of success or failure, including: O&M Training, Past Community Sanitation Experiences,

Priority Assessment, Technology Type, and Vandalism/Damage (Table G15). Closer examination of the causal condition O&M Training demonstrated that system operators could gain skills through other mechanisms besides formal training programs, such as informal experience or university sanitation education. Consequently, O&M Training was far less important than operators’ skills retention and was excluded in the analysis. Past Community Sanitation

Experiences, which reflected whether a community had previously had a successful system, a failed system, or no system, was removed because there was no relationship between previous experiences and current outcomes. There was a nearly equal spread of cases that previously had a failed system and now have a successful system and vice versa. Priority Assessment was removed

288

because there was no apparent relationship between whether an assessment was conducted and how well priorities were addressed. Instead, Addressed Sanitation Priorities was a much stronger determinant of continued system use and maintenance. Similarly, while Technology Type had low necessity, the technology itself was not a driver of success. Instead, causal conditions such as

Sufficient O&M Funds and Technical Support were more important, especially for more complex systems (i.e., systems with more required maintenance tasks). Finally, Vandalism/Damage was removed because it was closely connected to the sub-outcome of maintenance, where the fact that maintenance rectified (or did not) vandalism/damage was more important than whether the event occurred in the first place. The removal of low necessity causal conditions simplified the analysis.

Table G15. Necessity scores for hypothesized causal conditions for success that were excluded from the analysis because necessity scores were below 0.30. Necessity for Specified Outcome Condition Success Use Maintenance Performance Past Sanitation Experiences 0.30 0.35 0.30 0.32 Priority Assessment 0.29 0.26 0.34 0.34 O&M Training 0.27 0.20 0.24 0.25 Technology Type 0.21 0.18 0.19 0.20 Vandalism/Damage 0.20 0.30 0.29 0.24

Finally, three potential causal conditions were removed through a re-examination of case knowledge and theoretical evidence in preliminary iterations of fsQCA that demonstrated that these hypothesized causal conditions were not the most influential drivers of success or failure:

Community Participation in O&M, Organization Embeddedness, and Organization Sanitation

Experience. Community Participation in O&M was removed because while literature suggests the importance of community engagement in all project phases, participation in planning may have the greatest influence on technology selection and therefore community buy-in for use and maintenance. Additionally, case knowledge demonstrated that the presence of a skilled operator and external technical assistance were far more important than whether the community was

289

operating the system. Organization Embeddedness was removed because case knowledge indicated that while familiarity with the community could lead to better addressed sanitation priorities, this was not always the case, and addressed sanitation priorities was a stronger determinant of continued use and maintenance. Organizations unfamiliar with the communities were also able to establish strong partnerships with municipalities and communities, despite not having previously worked in these locations. Organization Sanitation Experience, in many cases, meant that implementing organizations recognized the importance of establishing clear O&M responsibility and of engaging local stakeholders. However, increased sanitation experience did not necessarily mean that organizations adequately ensured that these important causal conditions were present. Therefore, we chose to focus the analysis on the causal conditions like Clear O&M

Plan that were more closely connected to the outcomes themselves.

From these initial analyses steps, we reduced the number of causal conditions analyzed to nine, which is a reasonable number for twenty cases using QCA (Ragin 2008): Addressed

Sanitation Priorities, Behavior Change Education, Clear O&M Plan, Community Participation in

Planning, Construction Quality, Government Barriers, Municipality Involved in Planning,

Sufficient O&M Funds, and Technical Support.

G8.3 Truth Table Analysis Following the calibration of the causal conditions and outcomes (sections G5-G7 of this appendix) and the creation of the truth table (Chapter 4, Table 4.2), the truth table was analyzed using the “Truth Table Analysis” function in fs/QCA software (Ragin et al. 2017). Truth table analysis relies on the process of minimization. Minimization allows for the logical simplification of possible pathways. For instance, two rows in the truth table that result in the same outcome might differ by only one causal condition, and removing the differing causal condition produces a more simplified expression. The minimization process performs these stepwise comparisons for

290

all possible combinations and yields the simplified combinations that are minimally sufficient to produce the outcome. For the outcomes of sanitation success and failure, a consistency score of

0.8 was used as the minimum score required for a pathway to be considered for inclusion in the final solution (Ragin 2008). Additionally, the proportional reduction in inconsistency (PRI), which adjusts the consistency metric to account for causal conditions that might be both subsets of the outcome and the negated outcome (Ragin 2008), was also evaluated. PRI is similar to the concept of the proportional reduction in error (PRE) in statistics and is a more robust measurement of how consistently a pathway leads to an outcome. Pathways with large differences between consistency and PRI scores (e.g., 0.83 and 0.60, respectively) were also removed to reduce the influence of cases that are a subset of both the outcome and negated outcome (Opdyke et al. 2018). Once a preliminary solution was obtained, a subset/superset analysis was performed for each outcome in order to further reduce the number of causal conditions in each pathway while maintaining or increasing each pathway’s consistency and coverage (Ragin 2008). Finally, pathways were compared with theory and case knowledge to ensure the final solutions presented the most complete and simplified causes of success and failure.

Fs/QCA software produces complex, intermediate, and parsimonious solutions for all pathway analyses. Complex and parsimonious solutions were examined at each step of the analysis but were not ultimately presented. Complex solutions do not consider any simplifying assumptions, and therefore, do not incorporate important theoretical knowledge (Ragin 2008). In the complex solution, all possible pathways that are not represented by an observed case (i.e., remainders) are assumed to not exhibit the outcome analyzed and are thus not included in the solution for the (positive) outcome (Ragin 2008). Complex solutions are often longer and are difficult to draw useful conclusions due to their complexity and case-specificity. Parsimonious

291

solutions arise when all remainders are used to obtain the most simplified pathway possible that produces the outcome (Ragin 2008). However, the minimization in the parsimonious solution may not align with theory, making these solutions more difficult to justify (Ragin 2008). Finally, intermediate solutions provide a middle-ground that incorporates the use of simplifying assumptions (based on theory and case knowledge) to minimize the pathways (beyond the complex solutions) while also maintaining important case complexity (Kaminsky and Jordan 2017; Ragin

2008). Intermediate solutions are the most commonly presented solutions in QCA literature

(Chatterley et al. 2014; Kaminsky and Jordan 2017; Kunz et al. 2015; Marks et al. 2018; Opdyke et al. 2018; Peletz et al. 2018). Therefore, for this research, only intermediate solutions were presented.

Finally, in research focusing on social phenomena, such as how social and institutional conditions influence sanitation systems, the problem of limited diversity must be addressed. Social phenomena are naturally limited in their diversity; thus, it is very common to find it impossible to identify an empirical case that represents each possible combination of conditions; this set of all possible combinations is called the logic space (Ragin 2008). While theory can suggest what outcomes might occur for combinations that lack empirical cases, conclusions about those combinations cannot be drawn with as much certainty. As a result, researchers aim to reduce the logic space by increasing the number of cases and reducing the number of causal conditions analyzed (Ragin 2008). To reduce the logic space, we limited our analysis of the 20 cases to ten causal conditions, which is a well-accepted amount for QCA (Jordan et al. 2011).

Granted, it is optimistic to presume that increasing the number of cases and reducing conditions eliminates all causal combinations without empirical cases, called counterfactuals in

QCA; thus, they are evaluated using counterfactual analysis (Ragin 2008). Counterfactual analysis

292

is a hypothetical thought experiment where researchers use theoretical knowledge to make assertions about whether the presence or absence of causal conditions would lead to the outcome

(Kaminsky and Jordan 2017). The goal of counterfactual analysis is to determine which of the possible unobserved pathways could theoretically lead to the outcome and which unobserved pathways would likely not occur. For instance, based on theory, one would expect that a failed system could occur without either Sufficient O&M Funds or a Skilled Operator and that there are likely unobserved failed systems without both of those conditions. Thus, the researcher could use theoretical evidence to make the assertion that the absence of both Sufficient O&M Funds and a

Skilled Operator would be associated with failure. For the counterfactual analysis, we used “easy” counterfactuals which assume whether a causal condition’s presence or absence would be expected to be associated with each outcome (Table G16).

Table G16. Summary of simplifying assumptions for success and failure of sanitation systems. Outcome Condition Success Failure Addressed Sanitation Priorities Present Absent Behavior Change Education Present Absent Clear O&M Plan Present Absent Community Participation in Planning Present Absent Construction Quality Present Absent Government Barriers Absent Present Municipality Involved in Planning Present Absent Sufficient O&M Funds Present Absent Technical Support Present Absent

G8.4 Success Analysis G8.4.1 Necessity and Sufficiency of Causal Conditions for Success First, I calculated the necessity and sufficiency of each causal condition for the outcome of success (Table G17).

293

Table G17. Necessity and sufficiency of conditions for success. Condition Necessity Sufficiency No Government Barriers 1.00 0.54 Sufficient O&M Funds 0.99 0.67 Clear O&M Plan 0.94 0.81 Construction Quality 0.92 0.71 Technical Support 0.85 0.88 Addressed Sanitation Priorities 0.81 0.91 Municipality Involved in Planning 0.73 0.89 Behavior Change Education 0.72 0.64 Community Participation in Planning 0.59 0.55

G8.4.2 Preliminary Analysis of Success Pathways Next, I used the Truth Table Analysis function to run a preliminary analysis of the possible success pathways. This analysis yielded two potential success pathways (Table G18).

Table G18. Preliminary pathways to success. The solution presented is the intermediate solution obtained using fs/QCA software (Ragin et al., 2017). Raw Unique Combination Consistency Cases Coverage Coverage Addressed Sanitation Priorities*Clear O&M 2, 11, Plan*Community Participation in Planning*Construction 0.56 0.17 0.93 13, Quality*No Government Barriers*Sufficient O&M 14, 18 Funds*Technical Support 1, 3, Behavior Change Education*Clear O&M Plan*Construction 7, 8, Quality*Municipality Involved in Planning*No Government 0.58 0.19 0.95 13, Barriers*Sufficient O&M Funds*Technical Support 15, 18 Solution Coverage: 0.75 Solution Consistency: 0.93

To arrive at these preliminary success results, I used the follow input conditions and assumptions:

Preliminary Analysis of Success Pathways Input Conditions

 Addressed Sanitation Priorities  Behavior Change Education  Clear O&M Plan  Community Participation in Planning  Construction Quality  Government Barriers  Municipality Involved in Planning  Sufficient O&M Funds  Technical Support

294

Preliminary Analysis of Success Pathways Assumptions

 Addressed Sanitation Priorities (Present)  Behavior Change Education (Present)  Clear O&M Plan (Present)  Community Participation in Planning (Present)  Construction Quality (Present)  Government Barriers (Absent)  Municipality Involved in Planning (Present)  Sufficient O&M Funds (Present)  Technical Support (Present)

G8.4.3 Success Subset/Superset Analysis The preliminary intermediate solution for success presented in Table G18 had high consistency and high coverage; both metrics are high enough to accept the preliminary intermediate solution without further analysis. However, seven causal conditions per pathway means the pathways remain quite complex and therefore may be more difficult to understand and replicate. To determine if causal conditions could be removed, a subset/superset analysis was performed to search for pathways that have similar consistency but higher coverage scores. The preliminary intermediate solution was also examined in-depth using case knowledge to determine which of the causal conditions were the most important drivers of the success outcome. Together, these methods allowed for simplification of the success pathways and identification of the final solution. For the subset/superset analysis, pathways with similar consistency and higher coverage are presented (i.e., subsets of the original pathway). If these do not exist, five subsets with consistency above 0.8 are presented.

Success Pathway 1 (Preliminary): Addressed Sanitation Priorities*Clear O&M Plan*Community Participation in Planning*Construction Quality*No Government Barriers*Sufficient O&M Funds*Technical Support Goal: Identify subsets with consistency near 0.93 and coverage above 0.56. Subset Consistency Coverage clearom*~govbarr 0.92 0.94 clearom*~govbarr*suffomfund 0.93 0.94 ~govbarr*techsup 0.92 0.85

295

~govbarr*suffomfund*techsup 0.93 0.85 suffomfund*techsup 0.92 0.85 clearom*~govbarr*techsup 0.92 0.82 clearom*~govbarr*suffomfund*techsup 0.94 0.82 clearom*suffomfund*techsup 0.93 0.82 addsanpri*~govbarr*suffomfund 0.93 0.81 addsanpri*~govbarr 0.93 0.81 addsanpri*suffomfund 0.93 0.81 addsanpri 0.91 0.81 addsanpri*clearom*~govbarr*suffomfund 0.95 0.79 addsanpri*clearom*~govbarr 0.95 0.79 addsanpri*clearom*suffomfund 0.94 0.79 addsanpri*clearom 0.92 0.79 addsanpri*~govbarr*suffomfund*techsup 0.94 0.74 addsanpri*~govbarr*techsup 0.94 0.74 addsanpri*suffomfund*techsup 0.94 0.74 addsanpri*techsup 0.91 0.74 addsanpri*clearom*~govbarr*suffomfund*techsup 0.94 0.73 addsanpri*clearom*~govbarr*techsup 0.94 0.73 addsanpri*clearom*suffomfund*techsup 0.93 0.73 addsanpri*clearom*techsup 0.91 0.73 clearom*commpartplan*~govbarr*suffomfund*techsup 0.91 0.58 clearom*commpartplan*~govbarr*suffomfund 0.91 0.58 commpartplan*~govbarr*suffomfund*techsup 0.91 0.58 clearom*commpartplan*suffomfund*techsup 0.90 0.58 commpartplan*suffomfund*techsup 0.90 0.58 addsanpri*commpartplan*~govbarr*suffomfund 0.92 0.56 addsanpri*commpartplan*~govbarr 0.92 0.56 addsanpri*commpartplan*suffomfund 0.91 0.56 addsanpri*clearom*commpartplan*~govbarr*suffomfund*techsup 0.93 0.56 addsanpri*clearom*commpartplan*~govbarr*suffomfund 0.93 0.56 addsanpri*commpartplan*~govbarr*suffomfund*techsup 0.93 0.56 addsanpri*clearom*commpartplan*~govbarr*techsup 0.93 0.56 addsanpri*clearom*commpartplan*~govbarr 0.93 0.56 addsanpri*commpartplan*~govbarr*techsup 0.93 0.56 addsanpri*clearom*commpartplan*suffomfund*techsup 0.92 0.56 addsanpri*commpartplan*suffomfund*techsup 0.92 0.56 addsanpri*clearom*commpartplan*suffomfund 0.92 0.56 clearom*~govbarr 0.92 0.94 clearom*~govbarr*suffomfund 0.93 0.94 ~govbarr*techsup 0.92 0.85 ~govbarr*suffomfund*techsup 0.93 0.85 suffomfund*techsup 0.92 0.85 clearom*~govbarr*techsup 0.92 0.82 clearom*~govbarr*suffomfund*techsup 0.94 0.82 clearom*suffomfund*techsup 0.93 0.82

296

addsanpri*~govbarr*suffomfund 0.93 0.81 addsanpri*~govbarr 0.93 0.81 addsanpri*suffomfund 0.93 0.81 addsanpri 0.91 0.81 addsanpri*clearom*~govbarr*suffomfund 0.95 0.79 addsanpri*clearom*~govbarr 0.95 0.79 addsanpri*clearom*suffomfund 0.94 0.79 addsanpri*clearom 0.92 0.79 addsanpri*~govbarr*suffomfund*techsup 0.94 0.74 addsanpri*~govbarr*techsup 0.94 0.74 addsanpri*suffomfund*techsup 0.94 0.74 addsanpri*techsup 0.91 0.74 addsanpri*clearom*~govbarr*suffomfund*techsup 0.94 0.73 addsanpri*clearom*~govbarr*techsup 0.94 0.73 addsanpri*clearom*suffomfund*techsup 0.93 0.73 addsanpri*clearom*techsup 0.91 0.73 clearom*commpartplan*~govbarr*suffomfund*techsup 0.91 0.58 clearom*commpartplan*~govbarr*suffomfund 0.91 0.58 commpartplan*~govbarr*suffomfund*techsup 0.91 0.58 clearom*commpartplan*suffomfund*techsup 0.90 0.58 commpartplan*suffomfund*techsup 0.90 0.58 addsanpri*commpartplan*~govbarr*suffomfund 0.92 0.56 addsanpri*commpartplan*~govbarr 0.92 0.56 addsanpri*clearom*commpartplan*~govbarr*suffomfund 0.93 0.56 addsanpri*commpartplan*~govbarr*suffomfund*techsup 0.93 0.56 addsanpri*clearom*commpartplan*~govbarr*techsup 0.93 0.56 addsanpri*clearom*commpartplan*~govbarr 0.93 0.56 addsanpri*commpartplan*~govbarr*techsup 0.93 0.56 addsanpri*clearom*commpartplan*suffomfund*techsup 0.92 0.56 addsanpri*commpartplan*suffomfund*techsup 0.92 0.56 addsanpri*clearom*commpartplan*suffomfund 0.92 0.56

Success Pathway 2 (Preliminary): Behavior Change Education*Clear O&M Plan*Construction Quality*Municipality Involved in Planning*No Government Barriers*Sufficient O&M Funds*Technical Support Goal: Identify subsets with consistency near 0.95 and coverage above 0.58. Subset Consistency Coverage clearom*~govbarr*munplan*suffomfund*techsup 0.95 0.61 ~govbarr*munplan*suffomfund*techsup 0.95 0.61 clearom*munplan*suffomfund*techsup 0.95 0.61 munplan*suffomfund*techsup 0.95 0.61 bechangeduc*clearom*~govbarr*munplan*suffomfund*techsup 0.95 0.58 bechangeduc*clearom*~govbarr*munplan*suffomfund 0.95 0.58 bechangeduc*~govbarr*munplan*suffomfund*techsup 0.95 0.58 clearom*munplan*techsup 0.95 0.61

297

bechangeduc*~govbarr*munplan*suffomfund 0.95 0.58 munplan*techsup 0.95 0.61 bechangeduc*clearom*munplan*suffomfund*techsup 0.95 0.58 bechangeduc*munplan*suffomfund*techsup 0.95 0.58 bechangeduc*clearom*munplan*suffomfund 0.95 0.58 clearom*~govbarr*munplan*techsup 0.95 0.61 ~govbarr*munplan*techsup 0.95 0.61 bechangeduc*munplan*suffomfund 0.95 0.58 bechangeduc*clearom*~govbarr*munplan*techsup 0.95 0.58 bechangeduc*~govbarr*munplan*techsup 0.95 0.58 bechangeduc*clearom*~govbarr*munplan 0.95 0.58 bechangeduc*clearom*munplan*techsup 0.95 0.58 bechangeduc*~govbarr*munplan 0.95 0.58 bechangeduc*clearom*munplan 0.95 0.58 bechangeduc*munplan*techsup 0.95 0.58 bechangeduc*munplan 0.95 0.58

From the subset/superset analysis, Construction Quality can be removed because it does not appear in subsets of either pathway that reduce the number of causal conditions, increase coverage, and maintain similar consistency. No Government Barriers appeared in numerous subsets and had high necessity. However, this causal condition was added during the analysis for failure to improve the failure solution’s consistency and explain the failure pathways for two unique cases. For the successful cases, No Government Barriers was constant across all ten cases and therefore functioned like a domain condition. Furthermore, the more important government- related factors were the local municipality’s roles in planning and in post-implementation support—the majority of literature on government roles for sanitation focus on these causal conditions as well. Overall, No Government Barriers was determined to not be an important driver of success and was later removed from the success analysis. The removal of any additional causal conditions significantly reduced coverage below that of the preliminary success pathways and/or reduced consistency far below 0.93.

298

G8.4.4 Final Success Pathway Analysis Following the subset/superset analysis, the truth table was analyzed again without

Construction Quality and No Government Barriers. The revised intermediate solution is presented in Table G19. These two pathways were accepted as the final solution because of the high consistency and coverage and because the pathways were determined to be the most simplified and most complete explanations of the causes for success. The first success pathway had five causal conditions and explained success for five cases. The second success pathway had five causal conditions and explained success for seven cases, two of which (Cases 13, 18) were also explained by the first pathway.

Table G19. Final pathways to success. The solution presented is the intermediate solution obtained using fs/QCA software (Ragin et al., 2017). Raw Unique Combination Consistency Cases Coverage Coverage Addressed Sanitation Priorities*Clear O&M 2, 11, Plan*Community Participation in Planning*Sufficient O&M 0.56 0.17 0.93 13, 14, Funds*Technical Support 18 1, 3, 7, Behavior Change Education*Clear O&M Plan*Sufficient 0.58 0.19 0.95 8, 13, O&M Funds*Technical Support 15, 18 Solution Coverage: 0.75 Solution Consistency: 0.93

To arrive at these final success results, I used the follow input conditions and assumptions:

Final Success Pathway Analysis Input Conditions

 Addressed Sanitation Priorities  Behavior Change Education  Clear O&M Plan  Community Participation in Planning  Municipality Involved in Planning  Sufficient O&M Funds  Technical Support

Final Success Pathway Analysis Assumptions

 Addressed Sanitation Priorities (Present)  Behavior Change Education (Present)  Clear O&M Plan (Present)

299

 Community Participation in Planning (Present)  Municipality Involved in Planning (Present)  Sufficient O&M Funds (Present)  Technical Support (Present)

G8.5 Failure Analysis G8.5.1 Necessity and Sufficiency of Causal Conditions for Failure Next, I analyzed the pathways to sanitation failure. Like success, I calculated the necessity and sufficiency of each causal condition for the outcome of failure (Table G20).

Table G20. Necessity and sufficiency of conditions for failure. Condition Necessity Sufficiency Unaddressed Sanitation Priorities 0.92 0.83 No Municipality in Planning 0.92 0.78 No Technical Support 0.89 0.86 No Clear O&M Plan 0.79 0.94 Poor Construction Quality 0.64 0.89 Lack of Behavior Change Education 0.61 0.70 No Community Participation in Planning 0.55 0.58 Insufficient O&M Funds 0.53 0.99 Government Barriers 0.20 1.00 Lack of Community Participation in O&M Lack of Organization Sanitation

Experience Lack of Organization Embeddedness *causal condition was only included in the first preliminary analysis of failure pathways.

G8.5.2 Preliminary Failure Pathway Analysis I used the Truth Table Analysis function to run a preliminary analysis of the possible success pathways. This analysis yielded two potential success pathways (Table G21).

300

Table G21. Preliminary pathways to failure. The solution presented is the intermediate solution obtained using fs/QCA software (Ragin et al., 2017). Raw Unique Combination Consistency Cases Coverage Coverage 4, 5, Lack of Municipality in Planning*Unaddressed Sanitation 10, Priorities*No Technical Support*No Clear O&M Plan*Poor 0.58 0.19 1.00 16, Construction Quality*Lack of Behavior Change Education 17, 19 Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*No Clear O&M Plan*Lack 5, 6, 0.21 0.03 1.00 of Community Participation in Planning*Insufficient O&M 12, 19 Funds Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*Government Barriers*Poor 0.26 0.03 1.00 20 Construction Quality Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*Government 0.41 0.07 1.00 9 Barriers*Insufficient O&M Funds Solution Coverage: 0.70 Solution Consistency: 1.00

To arrive at these preliminary failure results, I used the follow input conditions and assumptions:

Preliminary Failure Pathway Analysis Input Conditions

Preliminary Failure Pathway Analysis Assumptions

 Addressed Sanitation Priorities (Absent)  Behavior Change Education (Absent)  Clear O&M Plan (Absent)  Community Participation in Planning (Absent)  Construction Quality (Absent)  Government Barriers (Present)  Municipality Involved in Planning (Absent)  Sufficient O&M Funds (Absent)  Technical Support (Absent)

301

G8.5.3 Failure Subset/Superset Analysis The preliminary intermediate solution for failure presented in Table G21 had high consistency and high coverage; both metrics are high enough to accept the preliminary intermediate solution without further analysis. To determine if causal conditions could be removed to further simplify the pathways, a subset/superset analysis was performed to search for pathways that have similar consistency but higher coverage scores. The preliminary intermediate solution was also examined in-depth using case knowledge to determine which of the causal conditions were the most important drivers of the failure outcome. Together, these methods allowed for simplification of the failure pathways and identification of the final solution. For the subset/superset analysis, pathways with similar consistency and higher coverage are presented

(i.e., subsets of the original pathway). If these do not exist, five subsets with consistency above 0.8 are presented.

Failure Pathway 1 (Preliminary): Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*No Clear O&M Plan*Poor Construction Quality*Lack of Behavior Change Education Goal: identify subsets with consistency near 1.00 and coverage above 0.58. Subset Consistency Coverage ~techsup*~munplan 1.00 0.82 ~munplan 1.00 0.82 ~addsanpri*~techsup*~munplan 1.00 0.81 ~addsanpri*~munplan 1.00 0.81 ~techsup*~clearom*~munplan 1.00 0.75 ~clearom*~munplan 1.00 0.75 ~addsanpri*~techsup*~clearom*~munplan 1.00 0.74 ~addsanpri*~clearom*~munplan 1.00 0.74 ~techsup*~munplan*~constrqual 1.00 0.62 ~munplan*~constrqual 1.00 0.62 ~addsanpri*~techsup*~munplan*~constrqual 1.00 0.61 ~addsanpri*~munplan*~constrqual 1.00 0.61 ~techsup*~clearom*~munplan*~constrqual 1.00 0.59 ~clearom*~munplan*~constrqual 1.00 0.59 ~addsanpri*~techsup*~clearom*~munplan*~constrqual 1.00 0.58 ~addsanpri*~clearom*~munplan*~constrqual 1.00 0.58

302

Failure Pathway 2 (Preliminary): Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*No Clear O&M Plan*Lack of Community Participation in Planning*Insufficient O&M Funds Goal: identify subsets with consistency near 1.00 and coverage above 0.21. Subset Consistenc Coverag y e ~techsup*~munplan 1.00 0.82 ~munplan 1.00 0.82 ~addsanpri*~techsup*~munplan 1.00 0.81 ~addsanpri*~munplan 1.00 0.81 ~techsup*~clearom*~munplan 1.00 0.75 ~clearom*~munplan 1.00 0.75 ~addsanpri*~techsup*~clearom*~munplan 1.00 0.74 ~addsanpri*~clearom*~munplan 1.00 0.74 ~commpartplan*~techsup*~munplan 1.00 0.49 ~commpartplan*~techsup*~clearom*~munplan 1.00 0.49 ~commpartplan*~munplan 1.00 0.49 ~commpartplan*~clearom*~munplan 1.00 0.49 ~addsanpri*~commpartplan*~techsup*~munplan 1.00 0.49 ~addsanpri*~commpartplan*~techsup*~clearom*~munplan 1.00 0.49 ~addsanpri*~commpartplan*~munplan 1.00 0.49 ~techsup*~munplan*~suffomfund 1.00 0.48 ~addsanpri*~techsup*~munplan*~suffomfund 1.00 0.48 ~addsanpri*~munplan*~suffomfund 1.00 0.48 ~munplan*~suffomfund 1.00 0.48 ~techsup*~clearom*~munplan*~suffomfund 1.00 0.44 ~clearom*~munplan*~suffomfund 1.00 0.44 ~addsanpri*~techsup*~clearom*~munplan*~suffomfund 1.00 0.44 ~addsanpri*~clearom*~munplan*~suffomfund 1.00 0.44 ~commpartplan*~techsup*~munplan*~suffomfund 1.00 0.41 ~commpartplan*~techsup*~clearom*~munplan*~suffomfund 1.00 0.41 ~commpartplan*~munplan*~suffomfund 1.00 0.41 ~commpartplan*~clearom*~munplan*~suffomfund 1.00 0.41 ~addsanpri*~commpartplan*~techsup*~munplan*~suffomfund 1.00 0.41 ~addsanpri*~commpartplan*~techsup*~clearom*~munplan*~suffomfun d 1.00 0.41 ~addsanpri*~commpartplan*~munplan*~suffomfund 1.00 0.41 ~addsanpri*~commpartplan*~clearom*~munplan*~suffomfund 1.00 0.41

Failure Pathway 3 (Preliminary): Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*Government Barriers*Poor Construction Quality Goal: identify subsets with consistency near 1.00 and coverage above 0.21. Subset Consistency Coverage govbarr 1.00 0.29

303

~techsup*govbarr 1.00 0.29 ~techsup*~munplan*govbarr 1.00 0.29 ~munplan*govbarr 1.00 0.29 ~addsanpri*govbarr 1.00 0.29 ~addsanpri*~techsup*govbarr 1.00 0.29 ~addsanpri*~techsup*~munplan*govbarr 1.00 0.29 ~addsanpri*~munplan*govbarr 1.00 0.29 govbarr*~constrqual 1.00 0.26 ~techsup*govbarr*~constrqual 1.00 0.26 ~techsup*~munplan*govbarr*~constrqual 1.00 0.26 ~munplan*govbarr*~constrqual 1.00 0.26 ~addsanpri*govbarr*~constrqual 1.00 0.26 ~addsanpri*~techsup*govbarr*~constrqual 1.00 0.26 ~addsanpri*~techsup*~munplan*govbarr*~constrqual 1.00 0.26 ~addsanpri*~munplan*govbarr*~constrqual 1.00 0.26 ~techsup*~munplan*~constrqual 1.00 0.62 ~techsup*~munplan 1.00 0.82 ~munplan*~constrqual 1.00 0.62 ~munplan 1.00 0.82 ~addsanpri*~techsup*~munplan*~constrqual 1.00 0.61 ~addsanpri*~techsup*~munplan 1.00 0.81 ~addsanpri*~munplan*~constrqual 1.00 0.61 ~addsanpri*~munplan 1.00 0.81

Failure Pathway 4 (Preliminary): Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*Government Barriers*Insufficient O&M Funds Goal: identify subsets with consistency near 1.00 and coverage above 0.21. Subset Consistency Coverage ~techsup*~munplan*~suffomfund 1.00 0.48 ~munplan*~suffomfund 1.00 0.48 ~addsanpri*~techsup*~munplan*~suffomfund 1.00 0.48 ~addsanpri*~munplan*~suffomfund 1.00 0.48 govbarr 1.00 0.36 ~munplan*govbarr 1.00 0.36 ~addsanpri*govbarr 1.00 0.33 ~techsup*govbarr 1.00 0.29 ~techsup*~munplan*govbarr 1.00 0.29 ~addsanpri*~techsup*govbarr 1.00 0.29 ~addsanpri*~techsup*~munplan*govbarr 1.00 0.29 ~addsanpri*~munplan*govbarr 1.00 0.29 ~suffomfund*govbarr 1.00 0.23 ~addsanpri*~suffomfund*govbarr 1.00 0.22 ~addsanpri*~munplan*~suffomfund*govbarr 1.00 0.22 ~techsup*~suffomfund*govbarr 1.00 0.21 ~techsup*~munplan*~suffomfund*govbarr 1.00 0.21

304

~munplan*~suffomfund*govbarr 1.00 0.21 ~addsanpri*~techsup*~suffomfund*govbarr 1.00 0.21 ~addsanpri*~techsup*~munplan*~suffomfund*govbarr 1.00 0.21 ~suffomfund*govbarr 1.00 0.21 ~techsup*~munplan 1.00 0.82 ~munplan 1.00 0.82 ~addsanpri*~techsup*~munplan 1.00 0.81 ~addsanpri*~munplan 1.00 0.81

From the subset/superset analysis, the only causal condition that can be removed is Lack of Behavior Change Education because it did not appear in subsets that reduce the number of causal conditions, increase coverage, and maintain similar consistency. The rest of the causal conditions that appeared in at least one of the four preliminary failure pathways were determined to be important explanations of failure from theory and case knowledge. The removal of any additional causal conditions significantly reduced coverage below that of the preliminary failure pathways and/or reduced consistency far below 1.00.

G8.5.4 Final Failure Pathway Analysis Following the subset/superset analysis, the truth table was analyzed again without Lack of

Behavior Change Education. The revised intermediate solution is presented in Table G22. The four failure pathways were accepted as the final solution because of the high consistency and coverage and because the pathways were determined to be the most simplified and most complete explanations of the causes for failure. The first failure pathway had five causal conditions and explained failure for six cases. The second failure pathway had six causal conditions and explained failure for four cases, two of which (Cases 5, 19) were also explained by the first pathway. The final two failure pathways each had five conditions and explained one case apiece. The inclusion of Government Barriers in the solution to failure was essential; solutions without this causal condition had very low coverage and/or consistency below 0.8.

305

Table G22. Final pathways to failure. The solution presented is the intermediate solution obtained using fs/QCA software (Ragin et al., 2017). Raw Unique Combination Consistency Cases Coverage Coverage 4, 5, Lack of Municipality in Planning*Unaddressed Sanitation 10, Priorities*No Technical Support*No Clear O&M Plan*Poor 0.58 0.19 1.00 16, Construction Quality 17, 19 Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*No Clear O&M Plan*Lack 5, 6, 0.21 0.03 1.00 of Community Participation in Planning*Insufficient O&M 12, 19 Funds Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*Government Barriers*Poor 0.26 0.03 1.00 20 Construction Quality Lack of Municipality in Planning*Unaddressed Sanitation Priorities*No Technical Support*Government 0.41 0.07 1.00 9 Barriers*Insufficient O&M Funds Solution Coverage: 0.70 Solution Consistency: 1.00

To arrive at these final failure results, I used the follow input conditions and assumptions:

Final Failure Pathway Analysis Input Conditions

 Addressed Sanitation Priorities  Clear O&M Plan  Community Participation in Planning  Construction Quality  Government Barriers  Municipality Involved in Planning  Sufficient O&M Funds  Technical Support

Final Failure Pathway Analysis Assumptions

 Addressed Sanitation Priorities (Absent)  Clear O&M Plan (Absent)  Community Participation in Planning (Absent)  Construction Quality (Absent)  Government Barriers (Present)  Municipality Involved in Planning (Absent)  Sufficient O&M Funds (Absent)  Technical Support (Absent)

306

G9. Supporting Information References Andres, L. A., Briceno, B., Chase, C., and Echenique, J. A. (2014). Sanitation and Externalities: Evidence from Early Childhood Health in Rural India. Policy Research Working Paper, The World Bank, South Asia Region. Arnstein, S. R. (1969). “A ladder of citizen participation.” Journal of the American Planning Association, 35(4), 216–224. Bao, P. N., Aramaki, T., and Hanaki, K. (2013). “Assessment of stakeholders’ preferences towards sustainable sanitation scenarios: Assessment on sanitation scenarios.” Water and Environment Journal, 27(1), 58–70. Basurto, X., and Speer, J. (2012). “Structuring the Calibration of Qualitative Data as Sets for Qualitative Comparative Analysis (QCA).” Field Methods, 24(2), 155–174. Black, M. (1998). Learning What Works: A 20 Year Retrospective View on International Water and Sanitation Cooperation. UNDP - World Bank Water and Sanitation Program, Washington, D.C. Bouabid, A., and Louis, G. E. (2015). “Capacity factor analysis for evaluating water and sanitation infrastructure choices for developing communities.” Journal of Environmental Management, 161, 335–343. Breslin, E. D. (2003). Demand response approach in practice: why sustainability remains elusive. WaterAid, Mozambique. Brikké, F. (2000). “Operation and Maintenance of rural water supply and sanitation systems.” A training package for managers and planners. IRC International Water and Sanitation Center, WHO. Brikké, F., and Bredero, M. (2003). Linking technology choice with operation and maintenance in the context of community water supply and sanitation. WHO and IRC WASH, Geneva, Switzerland. Cairncross, S., Shordt, K., Zacharia, S., and Govindan, B. K. (2005). “What causes sustainable changes in hygiene behaviour? A cross-sectional study from Kerala, India.” Social Science & Medicine, 61(10), 2212–2220. Central Pollution Control Board. (2017). “Environmental Standards: Water Quality Criteria.” Central Pollution Control Board, (May 17, 2016). Chatterley, C., Javernick-Will, A., Linden, K. G., Alam, K., and Bottinelli, L. (2014). “A qualitative comparative analysis of well-managed school sanitation in Bangladesh.” BMC Public Health, 14(6), 1–14. Chatterley, C., Linden, K. G., and Javernick-Will, A. (2013). “Identifying pathways to continued maintenance of school sanitation in Belize.” Journal of Water, Sanitation and Hygiene for Development, 3(3), 411–422. Chowns, E. (2015). “Is Community Management an Efficient and Effective Model of Public Service Delivery? Lessons from the Rural Water Supply Sector in Malawi: Is Community Management Efficient and Effective?” Public Administration and Development, 35(4), 263–276. Cronin, A. A., Ohikata, M., and Kumar, M. (2014). “Social and economic cost-benefit analysis of sanitation in Odisha State, India.” Journal of Water, Sanitation and Hygiene for Development, 4(3), 521. Davis, A., Javernick-Will, A., and Cook, S. (2018). “A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based

307

sanitation systems.” Environmental Science: Water Research & Technology, 4, 1451– 1463. Davis, A., Javernick-Will, A., and Cook, S. (2019). “Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities.” Environmental Science & Technology, 53(1), 401–411. Devas, N., and Grant, U. (2003). “Local government decision-making? Citizen participation and local accountability: some evidence from Kenya and Uganda.” Public Administration and Development, 23(4), 307–316. Eales, K., Blackett, I., Siregar, R., and Febriani, E. (2013). Review of Community-Managed Decentralized Wastewater Treatment Systems in Indonesia. WSP. Hacker, M. E., and Kaminsky, J. A. (2017). “Cultural preferences for the methods and motivation of sanitation infrastructure development.” Journal of Water Sanitation and Hygiene for Development, 7(3), 407–415. Harris, D., Kooy, M., and Jones, L. (2011). Analysing the governance and political economy of water and sanitation service delivery. Overseas Development Institute, London, UK. Harris, M., Alzua, M. L., Osbert, N., and Pickering, A. (2017). “Community-Level Sanitation Coverage More Strongly Associated with Child Growth and Household Drinking Water Quality than Access to a Private Toilet in Rural Mali.” Environmental Science & Technology, 51(12), 7219–7227. IDECK. (2015). Strengthening the operation & maintenance sector for servicing decentralised urban sanitation infrastructure in Bangalore: Landscape synthesis report. CDD Society and BORDA, Bangalore, India. Jenkins, M. W., and Cairncross, S. (2010). “Modelling latrine diffusion in Benin: towards a community typology of demand for improved sanitation in developing countries.” Journal of Water and Health, 08(1), 166. Jenkins, M. W., and Scott, B. (2007). “Behavioral indicators of household decision-making and demand for sanitation and potential gains from social marketing in Ghana.” Social Science & Medicine, 64(12), 2427–2442. Jordan, E., Javernick-Will, A., and Tierney, K. (2016). “Post-tsunami recovery in Tamil Nadu, India: combined social and infrastructural outcomes.” Natural Hazards, 84(2), 1327–1347. Kaminsky, J., and Javernick-Will, A. (2012). “Causes for Sustainable Maintenance and Operation of On-Site Sanitation Systems.” Proceedings of the 2012 Construction Research Congress, 21–23. Kaminsky, J., and Jordan, E. (2017). “Qualitative comparative analysis for WASH research and practice.” Journal of Water Sanitation and Hygiene for Development, 7(2), 196–208. Katukiza, A. Y., Ronteltap, M., Niwagaba, C. B., Foppen, J. W. A., Kansiime, F., and Lens, P. N. L. (2012). “Sustainable sanitation technology options for urban slums.” Biotechnology Advances, 30(5), 964–978. Kooy, M., and Harris, D. (2012). Political economy analysis for water, sanitation and hygiene (WASH) service delivery. Project Briefing, Overseas Development Institute, London, UK. Kunz, N. C., Fischer, M., Ingold, K., and Hering, J. G. (2015). “Why Do Some Water Utilities Recycle More than Others? A Qualitative Comparative Analysis in New South Wales, Australia.” Environmental Science & Technology, 49(14), 8287–8296. Luthi, C., Morel, A., Tilley, E., and Ulrich, L. (2011). Community-Led Urban Environmental Sanitation Planning (CLUES). UN Habitat and Eawag-Sandec.

308

Mansuri, G. (2004). “Community-Based and -Driven Development: A Critical Review.” The World Bank Research Observer, 19(1), 1–39. Marks, S. J., and Davis, J. (2012). “Does User Participation Lead to Sense of Ownership for Rural Water Systems? Evidence from Kenya.” World Development, 40(8), 1569–1576. Marks, S. J., Kumpel, E., Guo, J., Bartram, J., and Davis, J. (2018). “Pathways to sustainability: A fuzzy-set qualitative comparative analysis of rural water supply programs.” Journal of Cleaner Production, 205, 789–798. Ministry of Urban Development. (2008). National Urban Sanitation Policy. Government of India, New Delhi, India. Mosler, H.-J. (2012). “A systematic approach to behavior change interventions for the water and sanitation sector in developing countries: a conceptual model, a review, and a guideline.” International Journal of Environmental Health Research, 22(5), 431–449. Murphy, H. M., McBean, E. A., and Farahbakhsh, K. (2009). “Appropriate technology – A comprehensive approach for water and sanitation in the developing world.” Technology in Society, 31(2), 158–167. Mwirigi, J., Balana, B. B., Mugisha, J., Walekhwa, P., Melamu, R., Nakami, S., and Makenzi, P. (2014). “Socio-economic hurdles to widespread adoption of small-scale biogas digesters in Sub-Saharan Africa: A review.” Biomass and Bioenergy, 70, 17–25. Opdyke, A. (2017). “Resilient and Sustainable Infrastructure Systems: A Comparative Analysis of Post-Disaster Shelter Coordination, Stakeholder Participation, and Training.” University of Colorado at Boulder. Opdyke, A., Javernick-Will, A., and Koschmann, M. (2018). “A Comparative Analysis of Coordination, Participation, and Training in Post-Disaster Shelter Projects.” Sustainability, 10(11), 4241. Palaniappan, M., Lang, M., and Gleick, P. H. (2008). A Review of Decsion-Making Support Tools in the Water, Sanitation, and Hygiene Sector. The Pacific Institute, Oakland, California. Peletz, R., Kisiangani, J., Bonham, M., Ronoh, P., Delaire, C., Kumpel, E., Marks, S., and Khush, R. (2018). “Why do water quality monitoring programs succeed or fail? A qualitative comparative analysis of regulated testing systems in sub-Saharan Africa.” International Journal of Hygiene and Environmental Health, 221(6), 907–920. Prokopy, L. S. (2005). “The relationship between participation and project outcomes: Evidence from rural water supply projects in India.” World Development, 33(11), 1801–1819. Ragin, C. C. (2008). Redesigning social inquiry: fuzzy sets and beyond. University of Chicago Press, Chicago. Ragin, C. C., Patros, T., Strand, S. I., and Rubinson, C. (2017). “User’s Guide to Fuzzy-Set / Qualitative Comparative Analysis.” Ramani, S. V., SadreGhazi, S., and Duysters, G. (2012). “On the diffusion of toilets as bottom of the pyramid innovation: Lessons from sanitation entrepreneurs.” Technological Forecasting and Social Change, 79(4), 676–687. Roma, E., and Jeffrey, P. (2010). “Evaluation of community participation in the implementation of community-based sanitation systems: a case study from Indonesia.” Water Science & Technology, 62(5), 1028–1036. Rosenquist, L. E. D. (2005). “A psychosocial analysis of the human-sanitation nexus.” Journal of Environmental Psychology, 25(3), 335–346.

309

Sakthivel, S. R., Seshadri, A., Rahman, A., and Chariar, V. M. (2014). “Standardisation of Design and Maintenance of DEWATS Plants in India.” India Institute of Technology Technical Paper Series. Sansom, K. (2011). “Complementary roles? NGO-Government relations for community-based sanitation in South Asia.” Public Administration and Development, 31(4), 282–293. Scope. (2012). Ecological Sanitation in Trichy, India. Scope Trichy, Tamil Nadu, India. Seymour, Z. (2014). “Sanitation in developing countries: a systematic review of user preferences and motivations.” Journal of Water, Sanitation and Hygiene for Development, 4(4), 681– 691. Sperling, J., Romero-Lankao, P., and Beig, G. (2016). “Exploring citizen infrastructure and environmental priorities in Mumbai, India.” Environmental Science & Policy, 60, 19–27. Starkl, M., Brunner, N., and Stenström, T.-A. (2013). “Why Do Water and Sanitation Systems for the Poor Still Fail? Policy Analysis in Economically Advanced Developing Countries.” Environmental Science & Technology, 47, 6102–6110. Tilley, E., Strande, L., Lüthi, C., Mosler, H.-J., Udert, K. M., Gebauer, H., and Hering, J. G. (2014). “Looking beyond Technology: An Integrated Approach to Water, Sanitation and Hygiene in Low Income Countries.” Environmental Science & Technology, 48(17), 9965–9970. Törnqvist, R., Norström, A., Kärrman, E., and Malmqvist, P.-A. (2008). “A framework for planning of sustainable water and sanitation systems in peri-urban areas.” Water Science & Technology, 58(3), 563–570. Uddin, S. M. N., Muhandiki, V. S., Sakai, A., Al Mamun, A., and Hridi, S. M. (2014). “Socio- cultural acceptance of appropriate technology: Identifying and prioritizing barriers for widespread use of the urine diversion toilets in rural Muslim communities of Bangladesh.” Technology in Society, 38, 32–39. Venkataramanan, V., Crocker, J., Karon, A., and Bartram, J. (2018). “Community-Led Total Sanitation: A Mixed-Methods Systematic Review of Evidence and Its Quality.” Environmental Health Perspectives, 126(2), 026001. WaterAid. (2011). Sanitation Framework. WaterAid, London, UK. Wegelin-Schuringa, M. (2000). “Public awareness and mobilisation for ecosanitation.” IRC International Water and Sanitation Center, Delft, Netherlands, IRC International Water and Sanitation Center, Delft, Netherlands, Bonn, Germany.

310

Appendix H: Supporting Information for Chapter 5

A final version of this Supporting Information will be published in the journal article of Chapter 5 following the completion of this dissertation.

311

H1. Case Information

Table H1. Summary of sanitation systems used to apply the sanitation sustainability frameworks. Case Toilet Number of State Sanitation Technology Outcome # Connection Intended Users Successful Individual (100% Used, 1 Tamil Nadu Conventional (DEWATS) 800 Household Toilets 91% Maintained, 3 of 3 Regulations Met) Resource Recovery Successful (DEWATS with Biogas Individual (100% Used, 2 Karnataka 850 Digester; Intended to Household Toilets 90% Maintained, 3 of 3 recover biogas and water) Regulations Met) Successful Individual (84% Used, 3 Tamil Nadu Conventional (DEWATS) 800 Household Toilets 90% Maintained, 2 of 2 Regulations Met) Successful Conventional (Settling Tank (95% Used, Individual 7 Tamil Nadu + Single-Pass Intermittent 820 100% Maintained, Household Toilets Sand Filter) 2 of 2 Regulations Met) Successful Individual (90% Used, 8 Tamil Nadu Conventional (DEWATS) 875 Household Toilets 100% Maintained, 2 of 2 Regulations Met) Resource Recovery Successful Shared (DEWATS with Biogas (100% Used, 11 Tamil Nadu Community 1,006 Digester; Intended to 92% Maintained, 3 of 3 Toilets recover biogas and water) Regulations Met) Failed Resource Recovery Shared (32% Used, (DEWATS with Biogas 12 Karnataka Community 802 7% Maintained, Digester; Intended to Toilets 0 of 3 recover biogas) Regulations Met) Resource Recovery Successful (DEWATS with Biogas Shared (100% Used, 100% Maintained, 13 Tamil Nadu Digester; Intended to Community 860 3 of 3 recover biogas, water, and Toilets Regulations Met) compost) Resource Recovery Successful (Community Ecological Shared (100% Used, 100% Maintained, 14 Tamil Nadu Sanitation System; Intended Community 845 3 of 3 to recover water and Toilets Regulations Met) compost) Successful Individual (91% Used, 15 Tamil Nadu Conventional (DEWATS) 850 Household Toilets 90% Maintained, 2 of 2 Regulations Met) Failed Individual (88% Used, 17 Karnataka Conventional (DEWATS) 1,010 Household Toilets 11% Maintained, 0 of 3 Regulations Met) Resource Recovery Successful Shared (DEWATS with Biogas (100% Used, 100% Maintained, 18 Tamil Nadu Community 1,000 Digester; Intended to 3 of 3 Toilets recover biogas and water) Regulations Met) Note: Case numbers match (Davis et al. 2018b, 2019b; a).

312

H2. Indicators and Adaptations Table H2. Complete list of indicators and adaptations from all six sustainability frameworks. Indicator Category Definition Units Adaptation*** Framework Investment Costs (Economic) Refers to the monetary expense needed for the construction of the Composite Indicator WWTP. It involves several costly rupees/person- Investment Costs Economic No adaptation required Approach (Molinos- items such as land, construction, equivalents Senante et al., 2014) machinery and equipment, facilities and piping works Investment Costs Capital costs paid by households SEI Sustainability (individual & Economic (individual) and subsidized Rupees (total) No adaptation required* Criteria(Lennartsson et societal) (government, NGO) al., 2009) Lakh Rupees/MLD TechSelect 1.0 (Kalbar Life Cycle Costs Economic Net present worth (1 lakh= No adaptation required et al., 2012) 100,000 Rupees) O&M Costs (Economic) O&M Costs Maintenance costs paid by households SEI Sustainability (individual & Economic (individual) and subsidized Cost/person No adaptation required* Criteria(Lennartsson et societal) (government, NGO) al., 2009) Related to the management of the WWTPs, and they include the rupees/m3 Composite Indicator O&M Costs Economic following cost items: energy, staff, treated No adaptation required Approach (Molinos- reagents, waste management and water/year Senante et al., 2014) maintenance Capacity to Pay/Affordability Affordability of standard household latrine that is being promoted in the area and of % of households that report they can UNICEF material and afford latrine construction; % of Maximum value from either of the framework's two metrics Economic % Sustainability Checks services taking households that have access to finance listed (Jawara et al., 2017) into consideration mechanisms if needed the possible existence of in- kind or financing support for the poorest

313

Indicator Category Definition Units Adaptation*** Framework If users cannot afford to buy this technology, scalability will not be Indicator possible without subsidy. If users contains list of cannot afford to pay for the operation Technology questions to and maintenance costs (including the Assessment Affordability evaluate Economic cost of major rehabilitation, No adaptation required Framework (TAF) (user) positive, sustainable service levels will be (Olschewski and negative, highly unlikely without permanent Casey, 2015) neutral, or external financial assistance. Potential potential impact benefits can influence attitude of users towards investing in this technology Capacity to pay Household monthly sanitation fee as a % of household monthly SEI Sustainability Cost/person/yea (user, Economic Disposable income disposable income (where disposable income = monthly Criteria(Lennartsson et r municipality) income-monthly expenses)* al., 2009) Indicator Supportive financial mechanisms such contains list of Supportive Technology as subsidies very much assist uptake questions to financial Assessment by the poor communities but do not evaluate mechanisms Economic No adaptation required* Framework (TAF) guarantee sustainability or scalability positive, (regulator/investor (Olschewski and as they may not be in place for long negative, ) Casey, 2015) periods or at scale neutral, or potential impact Willingness to Pay Willingness to pay was measured using the contingent SEI Sustainability Rupees/househ Willingness to pay Social Reasonable % of available income valuation method, as detailed by (Fujita et al. 2005), in the Criteria(Lennartsson et old/month absence of a definition of "reasonable %"* al., 2009) Willingness to % of households that report sanitation % of households that, when asked, would spend money from a pay/prioritization UNICEF as a high priority; % of latrines that salary raise on sanitation as one of the top three items they of sanitation Economic % Sustainability Checks were repaired/rebuilt/upgraded in the would spend a raise on (based on the UNICEF questionnaire among areas of (Jawara et al., 2017) last year associated with the framework) expenditure Other System Benefits (Resource Recovery, Economic, etc.) The qualitative scale was based on the number of programs/projects that occurred as a direct result of sanitation system (indirect too difficult to measure accurately), such as micro-loan program, community hall, or water supply system Qualitative SEI Sustainability Local that were funded with income from sanitation user fees or Social No definition provided (Low, Medium, Criteria(Lennartsson et development resource recovery profits. The scale was defined based on High) al., 2009) differences between cases. Low = 0 projects/programs; Medium = 1-2 projects/programs that occurred because of sanitation; High = >3 projects/programs that occurred because of sanitation*

314

Indicator Category Definition Units Adaptation*** Framework Potential to recover nutrients was adapted based on observed cases: nutrients (compost) and nutrient potential (sludge, which could be used for land application or in composting but requires additional treatment): Low = No compost production possible % of incoming based on technology, intended design, or poor performance and to system of minimal amount of sludge produced that needs to be treated SEI Sustainability Potential for reuse NPKS Environmental No definition provided before reuse; Medium = Small volumes of compost production Criteria(Lennartsson et of nutrients (Converted to intended (small = <10m3/day for compost) or higher volumes al., 2009) Low, Medium, of sludge produced but treatment is needed before reuse (high = High) >10m3/day sludge) ; High = High volumes of compost production (high = >10m3/day compost) or high volumes of sludge produced and minimal or no treatment is needed before reuse (e.g., digester sludge). Potential to recover products was adapted based on observed cases: nutrients (compost), nutrient potential (sludge Concerns the potential of the production, which could be used for land application or in wastewater treatment technologies to composting) and energy (biogas). Very Low = No biogas or recover energy and/or phosphorus compost production possible based on technology or poor from wastewater. On one hand, Qualitative performance and negligible sludge production; Low = No organic matter in wastewater can be evaluation/num biogas or compost production is intended in design, but small considered not as an energy sink, but eric scale volumes of sludge is produced and would require additional as an energy source (Garrido et al. conversion Composite Indicator Potential to treatment for reuse (small = <10m3/day); Moderate = Small Environmental 2013). Although energy recovery is a (Very Low (1), Approach (Molinos- recover products volumes of biogas or compost production is intended in design difficult task in wastewater treatment Low (3), Senante et al., 2014) (small = <3 m3/day for biogas, <10m3/day for compost), or plants, in recent years significant Moderate (5), high volumes of sludge produced that would require additional progress has been done in this topic. High (7), Very treatment for reuse; High = High volumes of biogas, compost, On the other hand, some wastewater High (9)) or sludge production with minimal or no additional treatment treatment processes are able to remove required for reuse (e.g., digester sludge) OR small volumes of and recover phosphorus from both biogas and compost/treated sludge (high = >3m3/day wastewater biogas; >10m3/day compost/sludge); Very high = high volumes of both biogas and compost/treated sludge. Qualitative evaluation/num Very Low = Effluent BOD > 110 mg/L (>10% over regulatory eric scale The potential of the wastewater limit); Low = Effluent BOD 100-110 mg/L (over regulatory conversion Composite Indicator Potential for water treatment technologies to achieve an limit); Moderate = Effluent BOD = 90-100 mg/L (meets Environmental (Very Low (1), Approach (Molinos- reuse effluent with enough quality to be regulations); High = Effluent BOD = 80-90 mg/L (10-20% Low (3), Senante et al., 2014) reused below regulatory limit); Very High = Effluent BOD <80 mg/L Moderate (5), (>20% below regulatory limit) High (7), Very High (9)) % of the consumption of Volume treated water discharged by the sanitation system as a SEI Sustainability Potential for reuse the system Environmental No definition provided % of total water consumed by the community. Low = <5%; Criteria(Lennartsson et of water (Converted to Medium = 5-10%; High =>10% al., 2009) Low, Medium, High)

315

Indicator Category Definition Units Adaptation*** Framework If producers do not raise enough revenues to cover the cost of market introduction, promotion, technology Indicator development, supply chain contains list of Technology development and after-sales support, questions to Assessment Profitability this technology may fail to be scalable evaluate Economic No adaptation required Framework (TAF) (producer) or sustainable. In some cases, positive, (Olschewski and subsidies will be needed and may be negative, Casey, 2015) provided by third parties, e.g. NGOs neutral, or or governments, to enable the potential impact producer to create enough turnover and revenues System Performance SEI Sustainability Discharge: BOD Environmental Organic matter removal g/person/year No adaptation required* Criteria(Lennartsson et al., 2009) SEI Sustainability g/person/year Discharge: N, P Environmental Impact on eutrophication No adaptation required* Criteria(Lennartsson et of NP al., 2009) Eutrophication Life cycle assessment for kg TechSelect 1.0 (Kalbar Environmental No adaptation required Potential eutrophication potential PO4eq/p.e./year et al., 2012) Concerns the percentage of these Composite Indicator Nitrogen Environmental pollutants that are removed from the % No adaptation required Approach (Molinos- efficiency removal influent Senante et al., 2014) Concerns the percentage of these Composite Indicator Organic matter Environmental pollutants that are removed from the % No adaptation required Approach (Molinos- efficiency removal influent Senante et al., 2014) Concerns the percentage of these Composite Indicator Phosphorus Environmental pollutants that are removed from the % No adaptation required Approach (Molinos- efficiency removal influent Senante et al., 2014) If a sanitation technology is scaled up to use in multiple districts, there could Indicator Potential for be impacts on the environment and contains list of negative impacts Technology natural resources on a regional or questions to or benefits for Assessment national scale. Such impacts might evaluate natural resources Environmental No adaptation required Framework (TAF) include unsafe disposal of sludge from positive, on larger scale (Olschewski and septic tanks. If the performance of the negative, (regulator/investor Casey, 2015) technology is vulnerable to changes in neutral, or ) the environment, this will be another potential impact risk if the technology is used at scale Qualitative: based on system performance (effluent water Risk quality) and discharge type/location. High = Does not meet one Contamination of common property assessment or Risk of Infection: or more regulations; Medium = Meets all regulations and SEI Sustainability resources, such as lakes, rivers and qualitative Immediate Environmental discharges within the community (e.g., to drainage canal, Criteria(Lennartsson et coastal areas, located within the (Qualitative = Environment river); Low = Meets all regulations and discharges to al., 2009) community Low, Medium, underground sewer (i.e., covered receiving body) or far from High) the community (>1km)*

316

Indicator Category Definition Units Adaptation*** Framework

Qualitative: based on system performance (effluent water quality) and discharge type/location. High = Does not meet one Risk or more regulations; Medium = Meets all regulations and Contamination of common property assessment or discharges to open water body that is located outside of the SEI Sustainability Risk of Infection: resources, such as lakes, rivers and qualitative community (e.g., lake far away) or open water body that Environmental Criteria(Lennartsson et Downstream coastal areas, located outside the (Qualitative = extends outside of the community (e.g., river); Low = Meets all al., 2009) community Low, Medium, regulations and discharges to underground sewer or open water High) body within the community that does not extend beyond the community (i.e., without a downstream impact, such as a lake in the community)*

Sewage sludge is a by-product inevitably produced in wastewater treatment plants. Although it can be kg sludge/m3 Composite Indicator Sewage sludge used as an organic amendment, the Environmental treated No adaptation required Approach (Molinos- production processing, reuse, and disposal present water/year Senante et al., 2014) one of the most complex problems facing wastewater treatment sectors (Metcalf & Eddy 2004) Concerns the percentage of these Composite Indicator Suspended solids Environmental pollutants that are removed from the % No adaptation required Approach (Molinos- efficiency removal influent Senante et al., 2014) CDC/ARC System Number of Number of applicable regulations met by system according to Environmental Water quality results WASH(Sabogal et al., Performance regulations met effluent water quality analysis 2014) Odors Based on community complaints of smells (from addressed Qualitative priorities data collection). High = more than 75% of SEI Sustainability Odor Technical No definition provided (Low, Medium, respondents complained about bad smells; Medium = 25-50% Criteria(Lennartsson et High) of respondents complained about smells; Low = <25% al. 2009) complained about smells* Qualitative Wastewater treatment plants can be a evaluation/num The procedure and odor emissions factors (OEFs) associated source of objectionable odors. Several eric scale with each treatment module followed the process used by the studies have concluded that conversion citation in the definition, where Estimated Odor = System Composite Indicator complaints from people living near Odors Social (Very Low (1), Capacity*(Sum(OEF for each treatment module in the Approach (Molinos- wastewater treatment plants and/or Low (3), system)). The qualitative scale was defined based on the Senante et al. 2014) waste sites exhibit a great number of Moderate (5), groupings of the cases. Very High = >7.5E6; High = 5.5-7.5E6; odor-associated symptoms (Stellacci High (7), Very Moderate = 3.5-5.5E6; Low = 1.5-3.5E6; Very Low = <1.5E6 et al. 2010) High (9)) Health & Hygiene

317

Indicator Category Definition Units Adaptation*** Framework UNICEF Handwashing Life cycle assessment for global Health % No adaptation required Sustainability Checks facility warming potential (Jawara et al., 2017) % of household respondents reporting UNICEF Handwashing Health always washing their hands with soap % No adaptation required Sustainability Checks practice or ash at specific critical times (Jawara et al., 2017) Percentage of households with CDC/ARC Hygiene Behavior Health appropriate handwashing behavior % No adaptation required WASH(Sabogal et al., (stated goal = 50% increase) 2014) Percentage of population using CDC/ARC Hygiene Health hygienic sanitation facilities (stated % No adaptation required WASH(Sabogal et al., Promotion goal = 75% in use) 2014) Maintenance of ODF (Open UNICEF Defecation Free) % of ODF verified communities that Converted metric to community-level. % of community Health % Sustainability Checks Status (Meeting still meet all the national ODF criteria members that report no open defecation in the community (Jawara et al., 2017) National ODF Criteria) Maintenance of % of ODF verified communities UNICEF Converted metric to community-level. % of community ODF Status (No Health where no evidence of open defecation % Sustainability Checks members that use the system daily, correctly, and exclusively. OD Evidence) can be found (Jawara et al., 2017) Indicator The use of the sanitation technology contains list of Technology Potential for could have negative impacts on the questions to Assessment benefits or local environment, for example, evaluate Environmental No adaptation required Framework (TAF) negative impacts discharge of effluents from treatment positive, (Olschewski and (user) devices towards water bodies and negative, Casey, 2015) wells neutral, or potential impact (1) Participation of a large number Quality of ODF (70%) of households’ members; (2) A UNICEF Minimum value of the framework's two metrics for this verification Institutional large number of households and OD % Sustainability Checks indicator process areas around the village were visited (Jawara et al., 2017) for the verification process (%) Risk assessment or Qualitative. Scale definitions are based on literature findings of Increased health risk on the household SEI Sustainability Risk of Infection: qualitative health impacts based on % sanitation coverage and use (which Health level due to the need to handle excreta Criteria(Lennartsson et Household (Qualitative = also indicates amount of open defecation (exposure)). High = or to exposure to excreta al., 2009) Low, Medium, <60% use; Medium = 60-75% use; Low>75% use* High) Global Warming kg- Global Warming Life cycle assessment for global TechSelect 1.0 (Kalbar Environmental CO2eq/p.e./yea No adaptation required Potential warming potential et al. 2012a) r kg- SEI Sustainability Global Warming Life cycle assessment for global Environmental CO2eq/p.e./yea No adaptation required Criteria(Lennartsson et Potential warming potential r al., 2009)

318

Indicator Category Definition Units Adaptation*** Framework Energy Use Composite Indicator Energy kWh/m3 treated Environmental kWh/m3 treated water/year No adaptation required Approach (Molinos- consumption water/year Senante et al., 2014) SEI Sustainability Energy Environmental MJ/person MJ/person No adaptation required* Criteria(Lennartsson et al., 2009) Land Use Composite Indicator m2 land/person Land area required Environmental m2 land/person equivalents No adaptation required Approach (Molinos- equivalents Senante et al., 2014) TechSelect 1.0 (Kalbar Land Requirement Environmental m2 of land required for system m2/MLD No adaptation required et al., 2012) SEI Sustainability Land Environmental Land investment m2/person No adaptation required* Criteria(Lennartsson et al., 2009) Material Use Type and SEI Sustainability Chemicals Environmental Type and volume/person/year volume/person/ No adaptation required* Criteria(Lennartsson et year al., 2009) Type and The indicator was expanded to present one indicator per type of SEI Sustainability Construction Environmental Type and volume volume/number construction material (concrete, PVC sewer pipes based on Criteria(Lennartsson et materials /length diameter and quality, gravel, number of pumps, filter plants) al., 2009) SEI Sustainability Freshwater Environmental m3/person/year m3/person/year No adaptation required* Criteria(Lennartsson et al., 2009) Heavy metals, persistent organic SEI Sustainability Hazardous Environmental compounds, antibiotics/medical mg/person/year No adaptation required* Criteria(Lennartsson et substances residues, hormones al., 2009) Reliability Number of severe weather events and Occurrence of CDC/ARC natural disasters post-implementation Total number of natural disasters or Environmental No adaptation required WASH(Sabogal et al. (e.g., flooding, heavy rain, landslides, events events 2014) earthquakes, etc.) Very Low = System has never functioned because pumps were Qualitative required but not installed; Low = Extended time (months) evaluation/num before pump replacement severely inhibits use and eric scale Refers to the probability of performance, regardless of number of pumps; Moderate = 3+ conversion Composite Indicator mechanical failures and the impact of pumps and moderate time (up to two weeks) before pump Reliability Environmental (Very Low (1), Approach (Molinos- failures upon effluent quality replacement; High = 1-2 pumps and moderate to fast time (2 Low (3), Senante et al., 2014) (Eisenberg et al., 2001) days-2 weeks) before pump replacement; Very High = 0-1 Moderate (5), pumps and very fast time (<2 days) before pump replacement. High (7), Very None of the systems have redundant pumps, therefore more High (9)) pumps means more susceptibility to failure.

319

Indicator Category Definition Units Adaptation*** Framework Reliability of the system is defined as the possibility of achieving adequate performance for a specific period of time under specific conditions (Oliveira and Von Sperling 2008). 0 = System has never functioned because pumps were required (Singhirunnusorn and Stenstrom but not installed; 25 =Extended time before pump replacement 2009) considered two major aspects of (months) severely inhibits use and performance, regardless of reliability for the wastewater Continuous number of pumps; 50 = 3+ pumps and moderate time before treatment process: plant performance scale from 0 pump replacement (up to two weeks); 75 = 1-2 pumps and TechSelect 1.0 (Kalbar Reliability Technical and mechanical reliability. Reliability (worst) - 100 moderate to fast time (2 days-2 weeks) before pump et al., 2012) of the treatment system can be (best) replacement; 100= 0-1 pumps and very fast time (<2 days) assessed by the following: (1) the before pump replacement. None of the systems have redundant variability of treatment effectiveness pumps, therefore more pumps means more susceptibility to under normal and emergency failure. operation, (2) the probability of mechanical failures, and (3) the impact of failures upon effluent quality (Eisenberg et al. 2001) Products have to fulfil the expectations of users. If expectations Indicator are not met in relation to performance, contains list of Technology Reliability of design life, quality and ease of O&M, questions to Assessment technology and a sanitation technology may be evaluate Technical No adaptation required Framework (TAF) user satisfaction rejected, or users may not be willing positive, (Olschewski and (user) to pay for it. If this technology negative, Casey 2015) enhances social status, this may also neutral, or improve the willingness of users to potential impact pay for it % of households where latrines were UNICEF Resilient damaged or collapsed due to Sustainability Checks construction of Environmental weather/extreme events; % of latrines % No adaptation required (Jawara et al., 2017) latrines that were repaired or rebuilt within 1 (UNICEF 2017) month of filling up or being damaged Since lack of pump replacement (which leads to lack of water for toilet flushing or an overflowing system) is a common source of system failure (CITE) and the ability to replace a pump is indicative of the ability to replace any part of the system, the focus is mainly on pumps. Low = System is not Risk of failure, effect of failure, Qualitative functioning because pumps were not installed, have not been SEI Sustainability System robustness Technical structural stability; Shock loads, (Low, Medium, replaced in months, or an extreme event disrupted performance; Criteria(Lennartsson et effects of abuse of system High) Medium = 2-4 pumps and moderate time before pump al., 2009) replacement (up to two weeks) and there has not been an extreme event that disrupted performance and/or no disruptions are expected from extreme events; High = 0-1 pumps with time before pump replacement <2days, no disruptions from extreme events and no disruptions expected from extreme events*

320

Indicator Category Definition Units Adaptation*** Framework (1) % of districts where a functioning There is a monitoring system (able to collect, functional analyze and report on sanitation monitoring system UNICEF program) is in place; (2) % of in place that Institutional % No adaptation required Sustainability Checks communities with a functional triggers corrective (Jawara et al., 2017) monitoring system in place that action at lowest triggers corrective action at lowest level level Design Life The qualitative scale is based on groupings from the cases Qualitative SEI Sustainability where the shortest design life is 20 years and the highest design Durability/lifetime Technical Design life of sanitation alternatives (Low, Medium, Criteria(Lennartsson et life is 40 years. Low = 20 years; Medium = 30 years; High = 40 High) al., 2009) years* Lifespan of CDC/ARC Design lifetime of water and WASH Technical Number Adapted to be just sanitation infrastructure WASH(Sabogal et al., sanitation infrastructure infrastructure 2014) Durability is defined as the The qualitative scale is based on groupings from the cases technological life time, which is one where the shortest design life is 20 years and the highest design of the important criteria in selection of Continuous life is 40 years. Scale steps are based on equal distances a wastewater treatment technology. scale from 0 between groups. Finally, the highest value (100) is defined as TechSelect 1.0 (Kalbar Durability Technical The technology should have at least (worst) - 100 40 or greater years because the framework states that et al., 2012) 40-50 years of technological life with (best) technologies should last at least 40-50 years. 0 = <10 years; 25 minimal maintenance and spare part = 10-20 years; 50 = 20-30 years; 75 = 30-40 years; 100 = >40 requirements years Maintenance Adapted to be a sanitation committee focused on sanitation maintenance, sanitation fee collection & savings. Inactive/No Committee: Either no committee exists or a committee was previously formed but currently does not collect fees, lacks Presence of an active committee that bank account, does not repair systems in timely manner, and Qualitative repairs systems in timely manner, the lack of maintenance leads to extended service disruptions; CDC/ARC Presence of active (Inactive, Economic avoids service disruptions, diligently Active Committee: Committee is present and repairs systems WASH(Sabogal et al., water committees Active, Highly collects water fees, has water bank usually in a timely manner (delays do not exceed 2 weeks), 2014) Active) account avoids service disruptions, collects sanitation fees but does not have a bank account; Highly Active Committee: Committee is well-organized and present, repairs systems in timely manner, avoids service disruptions, diligently collects sanitation fees, and has sanitation bank account Presence of water % of households that declare having UNICEF to build, repair, or Environmental adequate access to water to clean % No adaptation required Sustainability Checks clean latrine latrine (Jawara et al., 2017) Complexity

321

Indicator Category Definition Units Adaptation*** Framework

Adequate Converted to community-level. % of all constructed assets of UNICEF operation and Technical % of latrines in good condition % the sanitation system (toilets, treatment tanks) that are in good Sustainability Checks maintenance of condition (Jawara et al., 2017) latrine The qualitative scale is based on number of required Simplicity might be a key factor in the Qualitative maintenance tasks, where scale steps reflect differences selection of wastewater treatment evaluation/num between cases (the least maintenance tasks required is 5, the systems, particularly in developing eric scale greatest is 15). Very High = >12 maintenance tasks and countries. A lack of skilled workers conversion Composite Indicator requires formal training for maintenance skills; High = 10-12 Complexity Social represents a major constraint when (Very Low (1), Approach (Molinos- maintenance tasks and requires formal training for maintenance decision makers choose to implement Low (3), Senante et al., 2014) skills; Moderate = 8-9 maintenance tasks and requires minimal a sophisticated treatment Moderate (5), maintenance skills; Low = 6-7 maintenance tasks OR requires (Singhirunnusorn and Stenstrom High (7), Very minimal maintenance skills; Very Low = <5 maintenance tasks 2009) High (9)) OR requires no maintenance skill

The qualitative scale is based on number of required maintenance tasks, where scale steps reflect differences between cases (the least maintenance tasks required is 5, the greatest is 15) and on familiarity with technology for Complexity of Qualitative construction. High = unfamiliar technology that requires formal SEI Sustainability construction and Technical No definition provided (Low, Medium, training for maintenance skills with >10 regular maintenance Criteria(Lennartsson et O&M (individual High) tasks; Medium = technology requires minimal maintenance al., 2009) and societal) skills (informal training, or knowledge gained from experience) with >10 maintenance tasks; Low = technology requires minimal maintenance skills (informal training, or knowledge gained from experience) OR <10 maintenance tasks*

Low = Influent and effluent cannot be accessed to take samples (e.g., effluent pipe is completely underground or has been damaged such that it is entirely inaccessible); Medium = there Qualitative is some difficulty to access influent or effluent, but samples can SEI Sustainability Ease of system Technical No definition provided (Low, Medium, be taken (e.g., effluent pipe is far away but accessible or Criteria(Lennartsson et monitoring High) effluent is hard to reach and requires additional equipment such al., 2009) as a long pole to take samples); High = Influent and effluent are easily accessible, and samples can be taken (e.g., effluent can easily be reached by any authorized person) Existence of a community-based body that is % of communities with an existing capable, dynamic, committee/association/individual UNICEF and supported by Social active in regular meetings and % No adaptation required Sustainability Checks local leaders providing continuous promotion of (Jawara et al., 2017) reinforcing sanitation sanitation or hygiene norms

322

Indicator Category Definition Units Adaptation*** Framework Existence of social UNICEF Existence of a local by-law and norm supporting Social Yes/No No adaptation required Sustainability Checks corresponding sanctions or reward the ODF status (Jawara et al., 2017) Indicator contains list of Producers and providers need specific Technology questions to Level of technical technical and business skills to ensure Assessment evaluate and business skills Technical they continue to provide pre- and No adaptation required Framework (TAF) positive, needed (producer) after-sales services at competitive yet (Olschewski and negative, profitable rates Casey, 2015) neutral, or potential impact Adjusted to be person-equivalents where one person = 8 hours of work per day. Since the definition did not further account for whether the operation workforce must be skilled or not, skills Manpower Number of people required for Number of were not included in the adaptation. However, for all cases TechSelect 1.0 (Kalbar Requirement for Technical operation people evaluated, a similar skillset was required (i.e., operators et al., 2012) Operation required training specific to the operation of the system but did not need further background education such as an engineering degree or formal operator certification)

Low = Not possible to use local competence (no skilled labor Possibility to use Qualitative available locally); Medium = Local population has construction SEI Sustainability local competence Technical No definition provided (Low, Medium, skills that are adequate for sanitation construction, but they lack Criteria(Lennartsson et for construction High) explicit sanitation construction experience; High = Local al., 2009) population has specific sanitation construction experience*

Low = there is not an operator for the system (community Possibility to use Qualitative SEI Sustainability unwilling or unable to perform maintenance); Medium= system local competence Technical No definition provided (Low, Medium, Criteria(Lennartsson et is either operated by municipality or NGO; High = operator is for O&M High) al., 2009) from community*

Indicator contains list of Local production of the sanitation Technology Potential for local questions to technology or spares might lead to Assessment production of evaluate Environmental income generation, but might need No adaptation required Framework (TAF) product or spares positive, specific inputs, which are difficult to (Olschewski and (producer) negative, provide on a constant basis Casey, 2015) neutral, or potential impact Qualitative Low = no operator; Medium = dedicated operator from outside SEI Sustainability Responsibility Social Definition of level of organization (Low, Medium, the community; High = highly organized and active Criteria(Lennartsson et distribution High) maintenance committee or dedicated community operator* al., 2009)

323

Indicator Category Definition Units Adaptation*** Framework Indicator Sanitation technologies might need contains list of specific skills for management, Technology Skillset of user or questions to operation and maintenance. For some Assessment operator to evaluate Technical users, specific training is needed to No adaptation required Framework (TAF) manage positive, ensure proper use. Follow-up training (Olschewski and technology (user) negative, of users should be available if skills Casey, 2015) neutral, or are to be retained and updated potential impact Flexibility Not analyzed because the indicator was intended for SEI Sustainability Compatibility with comparisons of alternatives that are upgrades/modifications of Technical No definition provided Qualitative Criteria(Lennartsson et existing system** an existing system and was not intended for brand new systems, al., 2009) as are all 20 of the study cases 0 = technology requires complete reconstruction to undertake Flexibility attempts to account for the additional hydraulic or organic load; 25 = technology requires indigenous nature of a given type of significant modification (e.g., addition of one or more treatment technology to undergo the upgrade tanks) to undertake additional hydraulic or organic load; 50 = easily. The technology should be Scale from 0 technology requires moderate modification (e.g., small TechSelect 1.0 (Kalbar Flexibility Technical sufficiently flexible that minimal (worst) - 100 expansion of a treatment tank) to undertake additional et al., 2012) changes will be required to the (best) hydraulic or organic load; 75 = technology requires only infrastructure of the plant to undertake modifications to operation requirements (e.g., increased additional hydraulic and/or organic pumping to undertake additional hydraulic or organic load); load 100 = technology requires no modification to take on additional hydraulic and organic load. Low = technology requires significant modification (e.g., addition of one or more treatment tanks, complete reconstruction) to undertake additional hydraulic and organic Qualitative load; Medium = technology requires moderate modification SEI Sustainability Flexibility/ Technical No definition provided (Low, Medium, (e.g., small expansion of a treatment tank or a change in Criteria(Lennartsson et Adaptability High) operation requirements such as pumping) to undertake al., 2009) additional hydraulic or organic load; High = technology requires no modification to take on additional hydraulic and organic load* External Support/Resources Converted to community-level. Instead of the % of (1) % of communities with post- communities within a district receiving follow-up support, the Existence of post- triggering follow up support activities, adaptation measures the amount of support provided within a triggering follow by NGOs, local government or both; Qualitative case: (1) Community receives post-triggering follow up support UNICEF up support (2) % of communities with a post- Institutional (Low, Medium, activities (Y/N); (2) Community has a post-ODF action plan Sustainability Checks activities and the ODF action plan; (3) % of districts High) (Y/N); (3) Local municipality has the capacity (human and (Jawara et al., 2017) type and quality of with the capacity (human and financial resources) to provide post-ODF follow up support these activities financial resources) to provide post- (Y/N). Low = Yes to one or fewer metrics; Medium = Yes to ODF follow-up support two metrics; High = yes to all three metrics

Follow-up from The presence of follow-up education Qualitative Low: Neither education nor technical assistance; Moderate: CDC/ARC outside Institutional and technical assistance post- (Low, Medium, Either education or technical assistance but not both; High: WASH(Sabogal et al., organizations implementation High) Both education and technical assistance 2014)

324

Indicator Category Definition Units Adaptation*** Framework The sector must possess sufficient Indicator Sector capacity for capacities to introduce sanitation contains list of validation, technologies, for example, the Technology questions to introduction of capacity to coordinate actors, to Assessment evaluate technologies, and Institutional document and share experiences, the No adaptation required Framework (TAF) positive, follow up capacity to carry out quality (Olschewski and negative, (regulator/investor regulation, monitoring and evaluation, Casey, 2015) neutral, or ) to carry out applied research and to potential impact provide back-up technical support Indicator Support The development and introduction of contains list of Technology mechanisms for technologies requires a lot of financial questions to Assessment upscaling resources over a long period when evaluate Technical No adaptation required Framework (TAF) technology there are hardly any revenues. Many positive, (Olschewski and (regulator/investor initiatives don’t manage to cross this negative, Casey, 2015) ) “Valley of Death” neutral, or potential impact Indicator contains list of Viable supply Viable supply chains are required for Technology questions to chains for product, a sanitation technology to be scalable Assessment evaluate spares, and Technical and used on a sustained basis. Supply No adaptation required Framework (TAF) positive, services chains can also enhance the feedback (Olschewski and negative, (producer) from users to suppliers Casey, 2015) neutral, or potential impact Legal Indicator Sanitation technologies introduced Alignment with contains list of should be aligned with national Technology national strategies questions to standards if they are to get support Assessment and validation evaluate Institutional from government institutions. Support No adaptation required Framework (TAF) procedures positive, from government institutions is (Olschewski and (regulator/investor negative, important to achieve scalability and Casey, 2015) ) neutral, or sustainability potential impact Qualitative SEI Sustainability Current legal Existing legal framework and Social (Low, Medium, No adaptation required* Criteria(Lennartsson et acceptability institutional capacity High) al., 2009) Indicator Registration of producers/providers contains list of Legal regulation and effective monitoring of their Technology questions to and requirements activities by regulatory authorities Assessment evaluate for registration of Institutional enhances quality assurance. It may No adaptation required Framework (TAF) positive, producers also help to raise awareness of (Olschewski and negative, (producer) standard prices for technologies and Casey, 2015) neutral, or services potential impact

325

Indicator Category Definition Units Adaptation*** Framework If services are to deliver optimal benefits and be sustainable, the roles and responsibilities of users, local Indicator governments, NGOs, external support contains list of Legal structures agencies, private service providers and Technology questions to for management of national government must be clear. Assessment evaluate technology & Institutional Responsibilities for financing, No adaptation required Framework (TAF) positive, accountability management and external support (Olschewski and negative, (users) must be clearly set out and Casey, 2015) neutral, or understood. Institutions must be in potential impact place to fulfil roles. Legislation should enable roles and responsibilities to be acted out Use Percentage of households with access CDC/ARC Sanitation Environmental to (presence of) sanitation facility % No adaptation required WASH(Sabogal et al., Coverage (stated goal = 100%) 2014) Use of sanitation UNICEF % of households with access to facilities (Access Social % No adaptation required Sustainability Checks improved, unshared latrines to Unshared) (Jawara et al., 2017) Use of sanitation UNICEF % of households accessing shared facilities (Access Social % No adaptation required Sustainability Checks latrines to Shared) (Jawara et al., 2017) Use of sanitation % of surveyed households that built a UNICEF facilities (Access Social new latrine and that still uses the % No adaptation required Sustainability Checks to New Latrines) latrine (Jawara et al., 2017) Use of sanitation % of surveyed households that have UNICEF facilities (Access Social rebuilt/upgraded their latrine in the % No adaptation required Sustainability Checks to Rebuilt last year (Jawara et al., 2017) Latrines) Appropriateness/Acceptability/Satisfaction The acceptability criterion deals with this aspect of acceptance of technology from the point of view of socio-economic culture. Acceptability Continuous also takes into consideration the issues scale from 0 Addressed Sanitation Priorities Score*100 to get scale from 0 TechSelect 1.0 (Kalbar Acceptability Social such as simplicity of wastewater (worst) - 100 to 100 et al., 2012) treatment technology and odor related (best) problems to community which are crucial attributes in the selection of wastewater treatment systems, particularly for developing countries Qualitative SEI Sustainability Appropriateness to Social Acceptable to use and maintain (Converted to Addressed Sanitation Priorities Score* Criteria(Lennartsson et local context quantitative) al., 2009)

326

Indicator Category Definition Units Adaptation*** Framework Availability/access Percentage of households that report UNICEF ibility/appropriate Social easy availability of sanitation % No adaptation required Sustainability Checks ness/attractiveness materials, products and services (Jawara et al., 2017) of sanitation Target users must express the real need or demand for the sanitation service level provided by this Indicator sanitation technology if management contains list of challenges are to be overcome in the questions to Technology future. Cultural taboos and habits can evaluate Assessment Demand for the Social cause users to reject a sanitation positive, No adaptation required Framework (TAF) technology (user) technology. If users feel a technology negative, (Olschewski and is inferior, they may reject it. If users neutral, or Casey, 2015) are unwilling to invest in a technology potential or pay for its operation and impact. maintenance, prospects for sustainability will be undermined Comfort, personal security, smell, Qualitative SEI Sustainability Convenience Social noise, attractiveness, adapted to needs (Low, Medium, No adaptation required* Criteria(Lennartsson et to different age/gender/income groups High) al., 2009) Qualitative evaluation/num eric scale The qualitative scale is based on number of motors (the only Refers to the production of conversion sources of noise within this set of cases) in a system, where Composite Indicator Noise Social undesirable sound in the surrounding (Very Low (1), scale steps reflect differences between cases. Very High = 4+ Approach (Molinos- area of the wastewater treatment plant Low (3), motors; High = 3 motors; Moderate = 2 motors; Low = 1 Senante et al., 2014) Moderate (5), motors; Very Low = 0 motors High (7), Very High (9)) The scale is based on the calibration of community participation in planning , a causal factor that was defined and The technology should promote public measured in a published study that used qualitative comparative Scale from 0 participation and make the community analysis (Davis et al. 2019b). The calibration used a 4-value TechSelect 1.0 (Kalbar Participation Social (worst) - 100 responsible for the success of the fuzzy set (0, 0.33, 0.67, 1) that applied a definition with clear et al., 2012) (best) implementation of the project differences between values to reflect meaningful differences between cases. Community Participation in Planning*100 to get scale from 0 to 100 Population growth or decline post- CDC/ARC Population growth Social implementation, as a percentage of % No adaptation required WASH(Sabogal et al., or decline population at implementation 2014)

327

Indicator Category Definition Units Adaptation*** Framework Qualitative evaluation/num Takes into account the opinion of the eric scale local perception affected by the conversion Addressed Sanitation Priorities Scores. Very Low = 0-0.2; Low Composite Indicator WWTP. It describes the phenomenon Public acceptance Social (Very Low (1), = 0.2-0.4; Moderate = 0.4-0.6; High = 0.6-0.8; Very High = Approach (Molinos- of social opposition to the Low (3), 0.8-1.0 Senante et al., 2014) construction of such facilities known Moderate (5), as “not in my back yard” High (7), Very High (9)) (1) Participation of a high % of community members from all categories including men, women, children, people with disabilities, UNICEF Quality of people from poorest households, Minimum value of the framework's two metrics for this Institutional % Sustainability Checks triggering process people from minority groups, decision indicator (Jawara et al., 2017) makers, opinion leaders, elderly, etc.; (2) % of community members recalling main messages of the triggering Since the definition was broad and unspecific, the indicator was adapted to be a measure commonly compared to perception: Complexity, compatibility, Qualitative satisfaction. Satisfaction was measured using a Likert scale, SEI Sustainability System perception Social observability - including aspects of (Converted to where 0 = very unsatisfied, 5 = very satisfied. Satisfaction was Criteria(Lennartsson et reuse quantitative) chosen for this indicator because Addressed Priorities Score al., 2009) was previously used to measure appropriateness to local context within this framework* Qualitative evaluation/num The qualitative scale is based on number of visually appealing eric scale aspects of the system, where scale steps reflect differences The disturbance caused by the conversion between cases. Visual impact was measured using a point Composite Indicator Visual Impact Social presence of the wastewater treatment (Very Low (1), system: one point per: lack of damage, far from community, Approach (Molinos- plant in the surrounding landscape Low (3), plants, visuals considered in design, and painted system. Very Senante et al., 2014) Moderate (5), High = 0 points; High = 1 point; Moderate = 2 points; Low = 3 High (7), Very points; Very low = 4+ points High (9)) Education/Behavior Change The qualitative scale is based on meaningful differences Ability to address Qualitative SEI Sustainability between the cases. Low = no education provided; Medium = awareness and Social No definition provided (Low, Medium, Criteria(Lennartsson et limited education provided but not on an ongoing basis; High = information needs High) al., 2009) education provided that extended beyond implementation

328

Indicator Category Definition Units Adaptation*** Framework There may be very weak demand from users for the service level provided by Indicator this technology without substantial contains list of Need for behavior change to perceptions, attitudes and Technology questions to change and social behaviors. This requires strong Assessment evaluate marketing Social community mobilization, social No adaptation required Framework (TAF) positive, (regulator/investor marketing and integration/alignment (Olschewski and negative, ) with existing traditions and incentives, Casey, 2015) neutral, or for example linking improved potential impact sanitation practices with agricultural practices Often without strong promotion, technologies, providers and supply chains will not be known to users and Indicator buyers. Good promotion is essential contains list of Technology Need for for scalability. Ongoing market questions to Assessment promotion and research must be carried out by evaluate Social No adaptation required Framework (TAF) market research producers and providers if sanitation positive, (Olschewski and (producer) technologies are to continue to meet negative, Casey, 2015) users’ needs. Poor user feedback neutral, or mechanisms can mean that design potential impact problems are not acted upon, affecting sustainability Sustainable behavior can be stimulated by tailoring the The scale is based on the calibration of behavior change technological design (e.g., education, a causal factor that was defined and measured in a pisciculture, biogas recovery, etc.) published study that used qualitative comparative analysis Promotion of Scale from 0 such that sustainable behavior is (Davis et al. 2019b). The calibration used a 4-value fuzzy set TechSelect 1.0 (Kalbar Sustainable Social (worst) - 100 promoted. Other ways to stimulate (0, 0.33, 0.67, 1) that applied a definition with clear differences et al., 2012) Behavior (best) sustainable behavior involve between values to reflect meaningful differences between increasing the end user’s awareness, cases. Behavior Change Education*100 to get scale from 0 to participation, and responsibility 100 (Balkema et al. 2002) Other Unique Environmental Indicators

Percentage of households with year- CDC/ARC Water Environmental round access to improved water % No adaptation required WASH(Sabogal et al., Infrastructure source (stated goal = 100%) 2014)

Other Unique Technical Indicators

329

Indicator Category Definition Units Adaptation*** Framework 0 = technology that is implemented for the first time (EcoSan Community Toilets); 25 = technology that has been implemented elsewhere but has limited applications and is not The features of the technological well-known in the sector (i.e., only implemented by one solution should be sufficiently Scale from 0 company); 50 = technology has been implemented elsewhere TechSelect 1.0 (Kalbar Replicability Technical familiar that it can be easily replicated (worst) - 100 by more than one company but is not commonly known by et al., 2012) in other places without reliance on (best) other organizations; 75 = technology has been implemented specific technical expertise elsewhere and is known in the sector but is not the most common; 100 = technology is extremely common and is known in the sector and to other non-implementer stakeholders (communities, government) *The SEI Sustainability Criteria provides a relative comparison between alternatives using --,-,0,+, and ++, but the framework lacked specific guidance for how to determine the differences between - and -- and + and ++. Therefore, the following process was followed for all indicators used by this framework: If the indicator uses a qualitative scale from low- >high, then each relative assignment corresponds to one step in the qualitative scale (e.g., if the 0 alternative is "medium", then a system that is "high" is + and a system that is "low" is -. If the 0 alternative is low, then a "medium" system is + and a high system is ++. If the indicator is quantitative, then a value that is +/-5% from the 0 alternative = 0; a value that is +/- 5-25% from the 0 alternative = + or -; a value that is >25% from the 0 alternative = ++ or -- **Indicator not analyzed because it was intended for alternatives that are modifications/upgrades of an existing system, not for new systems ***For the indicators that are measured based on a qualitative or quantitative scale, the scales are presented from worst to best. For example, for Risk of Infection, "High" indicates the worst risk of infection and is listed first; for Potential for Water Reuse, "High" indicates the best resource recovery scenario and is listed last.

330

H3. Summary of Indicator Types and Amount of Adaptation

Table H3. Summary of numbers of indicators by the selected frameworks. Total from all Composite UNICEF Summary of Indicator SEI Sustainability CDC/ARC Six TechSelect Indicator TAF Sustainability Categories Criteria WASH Frameworks Approach Checks # of Indicators 112 12 17 34 18 20 11 # of Economic Indicators 12 1 2 3 3 2 2 # of Health Indicators 7 0 0 1 0 4 2 # of Environmental 35 3 10 13 3 2 4 Indicators # of Social Indicators 27 3 5 8 3 7 1 # of Technical Indicators 21 5 0 9 5 1 1 # of Institutional 9 0 0 0 4 4 1 Indicators

Table H4. Summary of the level of adaptation required by the selected frameworks. Composite SEI UNICEF CDC/ARC Level of Adaptation TechSelect Indicator Sustainability TAF Sustainability WASH Approach Criteria Checks Total % of Indicators that Required Adaptation 64% 47% 53% 0% 45% 36% Indicator Adaptation Required Based on Need to Adapt Application (e.g., latrines -> small-scale sanitation) (High Adaptation = >50% of indicators; Low None None None Moderate Low (18%) Moderate Adaptation = 20-50% of indicators; Low Adaptation = <20% of (8%) (0%) (0%) (0%) (25%) indicators; None = 0% of indicators) Indicator Adaptation Required Based on Lack of Adequate Definition/Specificity of Measurement (High Adaptation = >50% of indicators; Moderate Adaptation = 20-50% of indicators; Low Adaptation = High Moderate High None Moderate Low (18%) <20% of indicators; None = 0% of indicators). This measures indicators that (58%) (47%) (53%) (0%) (25%) needed to be adapted in addition to the indicators above that were converted to small-scale/community-level. Yes - required definition for Does the aggregation or comparison method need to be modified or further Not No No differences No Not applicable. defined? applicable. between --,-,0,+,++

331

Table H5. Summary of the total amount of adapted indicators based on the sustainability pillars. Total # of Indicators # Not # %

(from all six Adapted Adapted Adapted frameworks) Economic 12 8 4 33% Health 7 4 3 43% Environmental 35 26 9 26% Social 27 13 14 52% Technical 21 5 16 76% Institutional 9 5 4 44%

332

H4. Framework Results

Table H6. Summary of the results from the UNICEF Sustainability Checks Framework (UNICEF 2017). * denotes failed systems; all

other systems were successful.

based based

w up up w

New

Shared -

Rebuilt

Improved -

e community e

triggering follo triggering

latrine

ODF status ODF

in place that in place

clean latrine clean

Affordability

Published Case Published

nitation facilities facilities nitation

Handwashing facility Handwashing

Handwashing practice Handwashing

quality of these activities of these quality

body to oversee sanitation oversee to body

Maintenance of ODF ofStatus ODF Maintenance ofStatus ODF Maintenance

s/attractiveness of sanitation s/attractiveness

Quality of triggering process of triggering Quality

Use of sanitation facilities facilities ofUse sanitation

Resilient construction of latrines construction Resilient

support activities and the type and and type and the activities support

Use of sanitation facilities facilities ofUse sanitation

Use ofUse sa of to pay/prioritization Willingness

Quality of ODF verification process verification of ODF Quality

Presence of water to build, repair, or or repair, build, to of water Presence

Use of sanitation facilities facilities ofUse sanitation

Existence of post Existence

sanitation among areas of expenditure areas among sanitation

Existence of social norm supporting the the supporting norm of social Existence

triggers corrective action at lowest level at lowest action corrective triggers

There is a functional monitoring system system monitoring a functional is There

Existence of an activ of an Existence

Adequate operation and maintenance of maintenance and operation Adequate Availability/accessibility/appropriatenes 18 90% 100% 0% 100% 0% 0% 100% 75% 100% 0% 20% Yes 100% 100% No 80% 80% High 80% Yes

11 95% 100% 0% 100% 0% 0% 100% 25% 100% 0% 5% Yes 100% 92% No 80% 80% High 100% Yes

13 85% 100% 0% 100% 0% 0% 100% 0% 100% 0% 15% Yes 50% 100% No 70% 65% High 85% Yes

7 90% 95% 100% 0% 0% 0% 95% 20% 95% 0% 35% Yes 20% 100% No 30% 30% Medium 60% Yes

14 85% 100% 0% 100% 0% 0% 100% 80% 100% 0% 0% Yes 0% 100% No 70% 70% Medium 65% Yes

2 90% 100% 100% 0% 0% 0% 70% 50% 70% 2% 30% Yes 75% 90% No 40% 70% Low 70% Yes

8 80% 90% 100% 0% 0% 0% 90% 35% 90% 5% 35% Yes 35% 100% No 35% 35% Medium 60% Yes

15 85% 91% 100% 0% 0% 0% 100% 35% 100% 5% 10% Yes 40% 90% No 30% 30% Medium 70% Yes

3 85% 84% 100% 0% 0% 0% 90% 45% 90% 5% 85% Yes 55% 90% No 30% 30% Medium 15% Yes

1 80% 100% 100% 0% 0% 0% 95% 45% 95% 0% 55% Yes 45% 91% No 35% 35% Medium 25% Yes

12* 35% 32% 0% 100% 0% 0% 0% 0% 0% 30% 40% No 25% 7% No 20% 20% Low 20% No

17* 75% 88% 100% 0% 0% 0% 70% 50% 70% 25% 55% No 70% 11% No 25% 25% Low 25% No

333

Table H7. Summary of the results from the CDC/ARC WASH framework (Sabogal et al., 2014). * denotes failed systems; all other systems were successful. Water Sanitation Hygiene Occurrence Follow-Up Population Presence of Lifespan of Case Infrastructure Coverage Behavior Hygiene of natural Education System growth or active water WASH Number (Target: (Target: (Target: Promotion disasters or and Performance decline committees infrastructure 100%) 100%) 75%) events Assistance Highly Active 18 95% 100% 60% 100% 1 20% 20 High 3/3 Committee Highly Active 11 65% 100% 20% 100% 2 24% 30 High 3/3 Committee Highly Active 13 75% 100% 35% 100% 2 9% 30 High 3/3 Committee Highly Active 7 90% 100% 30% 95% 2 7% 40 High 2/2 Committee 14 70% 100% 15% 100% 2 2% Committee 20 High 3/3

2 80% 100% 35% 100% 0 6% Committee 30 Moderate 3/3 Highly Active 8 90% 100% 35% 90% 3 9% 30 High 2/2 Committee Highly Active 15 100% 100% 20% 91% 2 6% 30 High 2/2 Committee 3 90% 100% 30% 84% 2 3% Committee 30 High 2/2

1 100% 100% 20% 100% 1 -13% Committee 30 High 3/3

12* 25% 30% 0% 32% 0 -6% No Committee 30 Low 0/3

17* 40% 88% 45% 88% 0 19% No Committee 30 Low 0/3

334

Table H8. Summary of the results from the SEI Sustainability Checks (Lennartsson et al., 2009). * denotes failed systems; all other systems were successful. 0 Alternative: Case Criteria Average Case 18 Case 11 Case 13 Case 7 Case 14 Case 2 Case 8 Case 15 Case 3 Case 1 Case 17* 12* Conventional DEWATS

Positive Criteria: +, ++ means relatively better than 0 Alternative; -, -- means relatively worse than 0 Alternative % of Investment Costs 1% -- ++ -- ++ -- ++ -- + ------++ that are Individual % of O&M Costs that are 260% - -- ++ ++ ------Individual Ability to address awareness and information Medium + + + 0 0 0 0 0 0 0 - - needs Appropriateness to local Medium + + + + + 0 0 - 0 - - - context Capacity to pay (user, 3632 + ++ 0 -- -- ++ - -- ++ + + ++ municipality) Compatibility with 0 0 0 0 0 0 0 0 0 0 0 0 0 existing system Convenience High ------0 -- 0 0 0 0 0 -- 0 Current legal acceptability Medium 0 0 0 0 0 0 0 0 0 0 0 0 Ease of system monitoring Medium 0 0 0 0 - 0 0 0 0 0 - - Flexibility/Adaptability Medium - - - 0 - - 0 0 0 - - - Local development Medium + + + 0 - 0 0 0 0 0 - - Possibility to use local competence for Low 0 0 0 ++ + 0 0 0 0 0 0 0 construction Possibility to use local High 0 0 0 0 0 - 0 0 0 ------competence for O&M Potential for reuse of Medium + + + 0 + + 0 0 0 0 - - nutrients Potential for reuse of High 0 - - 0 0 0 0 0 0 - - 0 water Responsibility distribution Medium + + 0 0 0 0 0 0 0 0 - - System perception 3.96/5 + + + + + 0 0 0 ------System robustness High - - - 0 0 - 0 0 0 0 -- -- Willingness to pay Medium + + 0 + - 0 + + + - + +

335

0 Alternative: Case Criteria Average Case 18 Case 11 Case 13 Case 7 Case 14 Case 2 Case 8 Case 15 Case 3 Case 1 Case 17* 12* Conventional DEWATS

Negative Criteria: -, -- means relatively better than 0 Alternative; +, ++ means relatively worse than 0 Alternative % of Investment Costs 99% 0 0 0 0 0 - 0 0 0 0 0 0 that are Societal % of O&M Costs that are 94% - ++ + 0 + + + - - + ++ ++ Societal Chemicals 0 0 0 0 0 0 0 0 0 0 0 0 0 Complexity of construction and O&M High 0 0 0 -- - 0 0 0 0 0 0 0 (individual and societal) Construction Materials (0.11m diameter PVC 504 ------++ ++ ++ ------++ Pipe) Construction Materials (0.16m diameter PVC 735 ------++ -- ++ ++ -- -- Pipe) Construction Materials 3.6 ++ + ++ - -- -- + ++ ------(0.5 hp Pumps) Construction Materials 50 + ------0 ++ + ++ -- -- - (Concrete) Construction Materials 174 ------+ ++ + ++ -- -- + (Concrete Manholes) Construction Materials 29 ------+ ++ ++ ++ -- -- ++ (Filter Plants) Construction Materials 41283 ------++ -- 0 - + - - -- ++ (Gravel) Discharge: BOD 516 + -- -- ++ -- -- 0 -- ++ -- ++ ++ Discharge: N, P 32 -- + ++ - - - - + + 0 ++ ++ Durability/lifetime 32 -- - - + ------Energy 6.09 -- 0 ++ ------++ ------Freshwater 2.89 - 0 - 0 -- 0 0 0 0 0 - 0 Global Warming 3000962 ------++ ------Hazardous substances 0 0 0 0 0 0 0 0 0 0 0 0 0 Investment Costs 8298 ------++ + + ------(individual & societal) Land 0.56 ------++ ++ - -- 0 - O&M Costs (individual & 147 - -- ++ -- + ++ -- - -- ++ ++ ++ societal)

336

0 Alternative: Case Criteria Average Case 18 Case 11 Case 13 Case 7 Case 14 Case 2 Case 8 Case 15 Case 3 Case 1 Case 17* 12* Conventional DEWATS

Odor Low + + + 0 0 + 0 0 0 0 ++ ++ Risk of Infection: Low 0 0 0 + 0 0 + + + + ++ ++ Downstream Risk of Infection: Low 0 0 0 0 0 0 0 0 0 0 ++ 0 Household Risk of Infection: Low 0 + + 0 + + + 0 0 0 ++ ++ Immediate Environment

Figure H1. Comparison of weighted and unweighted results from the TechSelect 1.0 Framework (Kalbar et al. 2012a). * denotes failed systems; all other systems were successful.

337

H5. Comparison of Framework Results Ranking of Cases from Monitoring and Relative Comparison Frameworks The two single-score aggregation frameworks, TechSelect and the Composite Indicator

Approach, provided a clear ranking of systems. For the four other frameworks, there was no clear system ranking. Therefore, ranks were assigned based on indicator information. To rank the results of the relative comparison frameworks, SEI Sustainability Criteria and TAF, the number of indicators with -, 0, and + (also -- and ++ for SEI Sustainability Criteria) were counted. The system with the most positive and the least negative impacts was ranked first. To rank the results of the qualitative monitoring frameworks, UNICEF Sustainability Checks and CDC/ARC WASH, systems were numbered from 1 (best) to 12 (worst) within each indicator. The individual indicator ranks were summed, and the system with the lowest sum was ranked first. There are limitations to this approach since not all indicators are equal in importance or magnitude, however, this approach was selected so the framework results could be illustratively compared to identify general trends.

Table H9. Comparison of rankings from 1 (best) to 12 (worst) across selected frameworks. * denotes failed systems; all other systems were successful.

Composite SEI UNICEF CDC/ARC Case TechSelect Indicator Sustainability TAF Sustainability WASH Number Ranking Approach Criteria Ranking Checks Ranking Ranking Ranking Ranking 7 6 1 4 2 7 2 18 8 7 1 1 1 4 11 1 10 2 5 2 7 13 11 2 3 8 3 1 15 3 5 5 10 8 3 14 10 3 7 3 4 10 1 2 9 10 6 5 6 2 9 4 8 7 6 5 3 4 6 9 4 10 8 8 5 8 6 9 9 9 12* 7 11 12 11 12 12 17* 12 12 11 12 11 11

338

Table H10. Definitions of sustainability adopted by the six selected frameworks. Framework Sustainability Definition Sustainable technology is “a strategy that enables men and women to rise out of poverty and increase their economic situation by meeting their basic needs, through TechSelect developing their own skills and capabilities while making use of their available resources in an environmentally friendly manner” (Kalbar et al. 2012a) Sustainable technology is “technology that does not threaten the quantity and quality Composite Indicator of resources and has the lowest costs with respect to the physical, socio-cultural and Approach economic environments” (Vleuten-Balkema 2003) Sustainable technology is “technology that does not threaten the quantity and quality SEI Sustainability of resources and has the lowest costs with respect to the physical, socio-cultural and Criteria economic environments” (Vleuten-Balkema 2003) TAF did not explicitly define sustainability but endeavored to assess “the applicability TAF of technologies, and of successful introduction, sustainable use, and the operation of technologies providing lasting services” (Olschewski and Casey 2015) “Sustainable development is development that meets the needs of the present without UNICEF compromising the ability of future generations to meet their own needs” (WCED Sustainability Checks 1987) Sustainability is “the long-term effectiveness of water and sanitation infrastructure and CDC/ARC WASH retention of hygiene education” (Sabogal et al. 2014)

H6. Supporting Information References Balkema, A. J., Preisig, H. A., Otterpohl, R., and Lambert, F. (2002). “Indicators for the sustainability assessment of wastewater treatment systems.” Urban Water, 4(2), 153–161. Davis, A., Javernick-Will, A., and Cook, S. (2018). “A comparison of interviews, focus groups, and photovoice to identify sanitation priorities and increase success of community-based sanitation systems.” Environmental Science: Water Research & Technology, 4, 1451– 1463. Davis, A., Javernick-Will, A., and Cook, S. (2019a). “The use of qualitative comparative analysis to identify pathways to successful and failed sanitation systems.” Science of The Total Environment, 663(1), 507–517. Davis, A., Javernick-Will, A., and Cook, S. (2019b). “Priority Addressment Protocol: Understanding the Ability and Potential of Sanitation Systems to Address Priorities.” Environmental Science & Technology, 53(1), 401–411. Eisenberg, D., Soller, J., Sakaji, R., and Olivieri, A. (2001). “A methodology to evaluate water and wastewater treatment plant reliability.” Water Science and Technology, 43(10), 91– 99. Fujita, Y., Fujii, A., Furukawa, S., and Ogawa, T. (2005). Estimation of Willingness-to-Pay (WTP) for Water and Sanitation Services through Contingent Valuation Method (CVM): A Case Study in Iquitos City, The Republic of Peru. Review, Japan Bank for International Cooperation, 59–87. Garrido, J. M., Fdz-Polanco, M., and Fdz-Polanco, F. (2013). “Working with energy and mass balances: a conceptual framework to understand the limits of municipal wastewater treatment.” Water Science and Technology, 67(10), 2294–2301. Kalbar, P. P., Karmakar, S., and Asolekar, S. R. (2012). “Technology assessment for wastewater treatment using multiple-attribute decision-making.” Technology in Society, 34(4), 295– 302.

339

Lennartsson, M., Kvarnström, E., Lundberg, T., Buenfil, J., and Sawyer, R. (2009). Comparing sanitation systems using sustainability criteria. Stockholm Environment Institute, SEI, Stockholm. Metcalf & Eddy. (2004). Wastewater Engineering: Treatment and Resource Recovery. McGraw- Hill, New York, NY. Molinos-Senante, M., Gómez, T., Garrido-Baserba, M., Caballero, R., and Sala-Garrido, R. (2014). “Assessing the sustainability of small wastewater treatment systems: A composite indicator approach.” Science of The Total Environment, 497–498, 607–617. Oliveira, S. C., and Von Sperling, M. (2008). “Reliability analysis of wastewater treatment plants.” Water Research, 42(4–5), 1182–1194. Olschewski, A., and Casey, V. (2015). “The Technology Applicability Framework. A Participatory Tool to Validate Water, Sanitation, and Hygiene Technologies for Low- Income Urban Areas.” Technologies for Development, Springer International Publishing, Cham, 185–197. Sabogal, R. I., Medlin, E., Aquino, G., and Gelting, R. J. (2014). “Sustainability of water, sanitation and hygiene interventions in Central America.” Journal of Water, Sanitation and Hygiene for Development, 4(1), 89–99. Singhirunnusorn, W., and Stenstrom, M. K. (2009). “Appropriate wastewater treatment systems for developing countries: criteria and indictor assessment in Thailand.” Water Science and Technology, 59(9), 1873–1884. Stellacci, P., Liberti, L., Notarnicola, M., and Haas, C. N. (2010). “Hygienic sustainability of site location of wastewater treatment plants.” Desalination, 253(1–3), 51–56. UNICEF. (2017). Sustainability Checks: Guidance to Design and Implement Sustainability Monitoring in WASH. UNICEF, 44. Vleuten-Balkema, A. J. van der. (2003). “Sustainable wastewater treatment: developing a methodology and selecting promising systems.” Dissertation, Eindhoven University, Eindhoven. WCED. (1987). Report of the World Commission on Environment and Development: Our Common Future. World Commission on Environment and Development, Oxford.

340

Appendix I: Reflections, Lessons Learned, and Suggestions

In this appendix, I reflect on my research experience, expanding upon the limitations and future research recommendations in Chapter 6 to discuss lessons learned and suggestions for future research.

Addressing Priorities The comparison of priority identification methods presented in Chapter 2 evolved out of my experience pilot testing the data collection protocols. As such, it was not initially conceived as a comparative method study. The pilot testing itself taught me much about effective interviewing techniques and the benefits of conducting focus groups to gain community rapport. The study overall may have been conducted more efficiently and comprehensively if I had originally set out to compare priority identification methods. Instead, as I was testing different methods to determine which would most effectively uncover communities’ experiences with sanitation, I discovered significant and unexpected variation in the usefulness of interviews, photovoice, and focus groups, leading to my modification of the data collection protocol to be able to compare these methods directly. Consequently, it would be interesting to expand this work to include more priority identification methods such as community mapping or transect walks (Freudenberger 2011) that allow participants to see and point out priorities or issues.

Time constraints during my fieldwork also meant that I was only able to conduct mixed- sex and mixed-role focus groups. While this is also often how focus groups are conducted in practice due to implementing organization time and resource limitations, there is a body of research that recognizes the biases introduced in heterogeneous groups. I do not think the focus groups saw the benefit of diverse perspectives that in some cases generates innovative or more comprehensive ideas; instead, the mixed focus groups appeared to suppress opinions and/or reach consensus

341

sooner, perhaps due to group dynamics rather than true representativeness of comprehensive priorities. This study would benefit from also investigating the effectiveness of female-only focus groups, male-only focus groups, and youth-only focus groups. Outside of comparing priority identification methods, I would most likely choose not to use mixed-sex, mixed-age, and mixed- community role focus groups in my data collection in general due to their relative lack of rich data.

Another area that could not be explored due to time and resource constraints is how well non-community members familiar with the community could identify local priorities. For instance, many of the resource-limited communities studied had relationships with local government officials (e.g., municipal engineers who came weekly to check the community water supply) or non-governmental organizations (e.g., social workers for elderly or disabled populations). These representatives might be able to provide insight to local priorities and are often solicited for their perspectives. As such, expanding the respondent roles to include non-community member roles as well as more community roles (e.g., teachers, shop owners) would provide additional information on how to best identify local priorities.

Additionally, it would be interesting to more closely investigate why priorities were indeed so context-specific. The slightly different demographics, histories, and infrastructure levels in each community likely shifted and shaped the priorities identified by communities. Some of the priorities identified were intuitive and understandable, such as the one community who mentioned

Graveyard because they were the only community who lacked government-sanctioned space to bury their loved ones. Others were less intuitive and warrant closer investigation, such as why two communities valued Women’s Empowerment or Micro-Loan Programs, which were not shared across all communities.

342

Finally, while photovoice did not identify as many priorities as interviews, I found it to be one of the most useful tools for generating rich discussions on sanitation experiences and encouraging respondents to participate in subsequent interviews. Among the study communities, digital cameras were somewhat of a novelty and the opportunity to have printed family photos in exchange for a respondent’s participation seemed to positively influence participation interest. I encourage others to consider the use of photovoice to augment traditional data collection methods and open the possibility for creative storytelling and unique participation in their research.

Pathways to Success and Failure In Chapter 4, I identified two pathways to sanitation success and four pathways to failure.

While results demonstrate that either pathway could (and did) lead to success, two cases were explained by both pathways. The most robust pathway to success, in my perspective, is to combine the two success pathways: engage both community and municipality stakeholders and achieve community buy-in through both behavior change education and addressing sanitation priorities.

The two cases that shared all of these elements generated the most income, had the highest community satisfaction and buy-in, the most reliable external support, and most likely, the best chances for long-term success.

It is also notable that not a single successful system was entirely community managed, but all received significant financial and technical external support. The WASH sector needs to move away from the idea that sustainability is achieved only when communities are independently financing and managing infrastructure. From developing to developed contexts, completely unsubsidized infrastructure systems rarely exist and only the smallest and most remote systems are managed without any kind of external support. Instead of entirely community-based sanitation systems that are managed exclusively by community beneficiaries, sanitation implementation models should focus instead on municipality-based sanitation systems. Sustainable municipalities

343

should have regular financing mechanisms for sanitation, the ability to recruit, train, and retain skilled and paid operators, and access to more resources than communities.

I began my research on sanitation success and failure in India during the central government’s Swachh Bharat Mission, or My Clean India Campaign, which aims to end open defecation by October 2019 through the widespread provision of latrines. The Mission unleashes an unprecedented amount of financial and human resources for sanitation in a country where talk of defecation has largely been taboo. The 20 communities included in my research had sanitation systems implemented years prior to the Mission launch, and my fieldwork spanned only a limited and infant time of the Mission. It would be incredibly fascinating to investigate how causes of sanitation success and failure evolve in the presence of the Mission. It would also be interesting to analyze municipality support for sanitation as an outcome and track changes in the causal pathways from before and after the Mission. Similarly, it would be interesting to investigate behavior change or sanitation education as outcomes to understand the impact of the Mission’s rhetoric on changing the national sanitation landscape. A question I was asked during my defense is what call would I put out for future sanitation research if I had the opportunity and power to do so? One of the most fascinating and important future research areas would be to investigate the drivers behind strong and sustained government support for sanitation systems as well as to understand how those drivers change depending on the level of government (i.e., local, regional, state, federal, international).

Similarly, another question I was asked is what study would I design if I had access to unlimited resources and large data sets? I would expand on the above proposal and examine a global dataset of institutional supports and barriers for multiple types of critical infrastructure (e.g., sanitation, water, electricity, roads). I would examine the frequency at which each driver of institutional support (e.g., external aid/donors, public pressure, sanitation champions) occurs within the

344

different infrastructure types and why. Also, since qualitative comparative analysis does not measure the frequency of occurrence of causal conditions, an interesting complement to this study could be surveying small-scale sanitation systems to determine how often systems do and do not have the important elements of success. For example, how many systems address local priorities?

How many systems have dedicated and sufficient O&M funds? How many systems lack municipality involvement in planning? This knowledge could further guide the most important areas to focus limited sanitation resources.

An original intent of my research was to replicate this study in resource-limited communities in the United States to elucidate cultural and institutional supports and barriers to sanitation across multiple, diverse contexts. Unfortunately, time and resource limitations constrained my ability to expand my research beyond India. The strength of these findings would greatly increase with the addition of resource-limited case studies from multiple contexts, in particular, regions where sanitation is under-studied. I expect that some of the major causal conditions for sanitation success would remain constant across contexts, including Technical

Support, Clear O&M Plan, and Sufficient O&M Funds. Likely, local stakeholder participation and community buy-in are also important, but it would be interesting to explore the different ways that these conditions are enacted in different contexts.

Measuring Sanitation Sustainability Initially, I intended for the sustainability chapter to present new indicators to measure social sustainability due to the greater emphasis on measuring economic and environmental pillars.

I am thankful for the many conversations I had with WASH practitioners that instead encouraged me to evaluate existing frameworks in-depth before proposing another new framework. This was one of the many lessons learned within my research.

345

From the framework evaluation, social and institutional indicators required some of the most adaptation and were very framework-specific. Future research could help by investigating which of these indicators specifically are the most effective. For instance, it would be useful to understand how to measure the ways that different levels of institutional support and regulations influence sanitation system impacts. A study that applies and compares institutional indicators in multiple contexts (e.g., United States and India) could help to refine indicator selection and measurement.

Another limitation of the evaluation of sustainability frameworks was the many adaptations that I proposed for ambiguous indicators. Future research should test the sensitivity of these assumptions on sustainability results and compare the proposed adaptations to possible other indicators or metrics. For example, I used the Addressed Priorities Score for several indicators related to sanitation system acceptance or satisfaction because this measured seemed to most closely match the frameworks’ definitions. It would be interesting to compare how results change if other metrics such as a Likert-scale measure of satisfaction are used instead. Similarly, the recommendations of this work could be extended by researching which low-resource measures could serve as proxies for more demanding data collection (e.g., finding a proxy for Willingness to Pay). Finally, to extend the impact of Chapter 5, it would be interesting to partner with some of the implementing organizations who have developed sustainability frameworks to test the viability and efficacy of the suggested improvements.

Resource-Limited Community Research Research in resource-limited communities at times can be difficult due to the obvious resource limitations. This research was made possible by numerous collaborations with universities, municipalities, implementing organizations, communities, and research assistants.

Throughout my fieldwork, I learned first-hand the importance of these collaborations with local

346

partners both for facilitating effective data collection and for illuminating essential cultural and institutional norms.

Most of my data collection and analysis was qualitative. I employed ethnographic elements in my research but did not conduct a true ethnography. During my fieldwork, I had time to shop at local markets, see museums, visit temples, share meals, attend celebrations, and make friends. This time provided valuable insight to my understanding of how and why municipalities and implementing organizations made decisions and added depth to my understanding of communities’ experiences with sanitation. During the nine months I spent in India for fieldwork, I saw my knowledge and research efficacy grow immensely. I think my research would have further benefited from even more extended time and more opportunities to integrate ethnographical observation and interviewing. While most fieldwork explicitly builds in time for when things go wrong, I also encourage researchers to build in extra time for unplanned observation and conversations. In addition to this, I found initial meetings with community leaders, panchayats, or informal discussions with groups of community members to be helpful to orient me to major community concerns and experiences and to provide me a small baseline of information from which I could refine my interview questions.

Throughout my research, and increasingly as time progressed, I learned the value of sharing my half-baked ideas and preliminary results more broadly within the WASH sector. For research to be practical and relevant, not only should it be built upon solid theoretical foundations and extend the literature, but it should also reflect the values and needs of its sector. These values and needs are not always explicitly voiced or included in organization documentation. Consequently, it was incredibly valuable to connect with WASH practitioners annually at the UNC Water &

Health Conference, at Stockholm World Water Week, and over email. My calibrations of the

347

causal conditions for the success and failure study were reviewed by numerous people throughout that study. The list of possible sanitation sustainability frameworks was also started by practitioner recommendations from what they had used or heard of most. I also learned much from WASH colleagues on how to conduct and present rigorous qualitative research. I strongly encourage new and existing researchers to build a network of practitioners who can collaborate on research and provide continual feedback. For me, some of these individuals and organizations include WSUP,

WaterAid, Sanergy, Sanivation, USAID, World Vision, CDD Society, BORDA, Eawag-Sandec,

Christoph Lűthi, Lukas Ulrich, Geeta Singha, Rohit Chandragiri, Marius Klinger, Jenna Davis, Liz

Jordan, Rita Klees, Sara Marks, Rachel Peletz, Kaity Mattos, Matt Bentley, Laura MacDonald, and Vidya Venkataramanan.

Lastly, One of the highlights of my fieldwork was the opportunity to work with a master’s student from TERI University’s water policy and sustainability program. I was connected to TERI

University through mutual colleagues from the USAID India Mission towards the end of my first fieldwork stint. I wish I had had the opportunity to work with more students from TERI University.

I encourage future researchers and fieldworkers to engage local students in the development and competition of fieldwork and research; their insights are invaluable.

348