National Survey and Segmentation of Smallholder Households in Uganda

WORKING PAPER

National Survey and Segmentation of Smallholder Households in Uganda Understanding Their Demand for Financial, Agricultural, and Digital Solutions Jamie Anderson, Colleen E. Learch, and Scott T. Gardner

April 2016

CONTENTS

A. INTRODUCTION AND KEY FINDINGS ��1 Introduction �� 1 Key Findings �� 2

B. ABOUT THE PROJECT ��5

C. FINDINGS ��10 1. Smallholder Household Dynamics in Uganda: Who They Are ��10 2. Smallholder Household Dynamics in Uganda: Income and Expenses ��23 3. Agricultural Risks and Mitigation among Smallholder Households in Uganda ��33 4. Mobile Phone Tools ��41 5. Financial Inclusion among Smallholder Households in Uganda ��46 6. Tools and Financial Inclusion: Segmentation ��56 7. Desires and Aspirations ��82

D. ANNEX 1: METHODOLOGY AND DESIGN ��90

E. ANNEX 2: RANDOM FOREST ALGORITHM ��95

III National Survey and Segmentation of Smallholder Households in Uganda

ACKNOWLEDGMENTS The authors wish to thank Caldwell Bishop, Mary Ann Fitzgerald, Lucy Mukemi, David Musiime, Denise O’Reilly, and Mamadou Thiam (InterMedia) for their valuable contributions and feedback during the preparation of this paper.

IV National Survey and Segmentation of Smallholder Households in Uganda

A. INTRODUCTION AND KEY FINDINGS Introduction between August and September 2015. This study sought to comprehensively Uganda is a fast-growing nation and map the many activities, interests, aspira- economy, showing year-over-year gross tions, barriers and pressures facing small- domestic product (GDP) growth1 since 2013 and a sharp recovery from previ- also explored nonagricultural household ous GDP declines. The growth comes holder households. The questionnaire- largely from public and private invest- ests, as well as challenges and aspirations. ment and demand, as well as expansion activities, financial practices and inter of agricultural activities,2 which is the - foundation of Uganda’s economy. It con- vations and insights from the national tributes 23 percent of the GDP, and over surveyThis report of smallholder shares the findings,households obser in 80 percent of the population is involved Uganda. It begins with an overview of in agricultural activities.3 Coffee, one of the research approach, core program the largest exports from Uganda, is al- - most entirely dependent on the small- nary phases of development, and topics holder farming sector.4 objectives, research questions, prelimi

Both the public and private sector in includingincluded intheir the household questionnaire. demograph It then- - ics,profiles farmographics, smallholder and farmers decision-making, in Uganda, as well as how farmers self-identify and infrastructureUganda have alsoin recent made significantyears. These in characterize their identity, and what vestments in the country’s financial motivates them to do the work they do. increasing access and use of more have benefited the country at large, This comprehensive exploration of the greater economic stability. Amidst this lives of smallholder farming households growth,formal financialsmallholder services farmers to have facilitate spe- sought to answer the following three - tially are more complex because of their ■■ questions:What does the community of practice livelihoodcific, unique,. Smallholder unmet needs farmers that face poten the need to know or do to support small- cost of inputs such as seed, fertilizer and holder farmer households build re- pesticide; bringing their goods to sale silient and productive livelihoods? along poor road networks; as well as the lack of or prohibitive cost of post-harvest ■■ - storage facilities for crops. And, they spond to the relevant needs and de- face these challenges within a largely in- siresHow of can smallholder financial mechanismshouseholds? re formal, cash-based economy where they ■■ What types of market strategies and approaches can cultivate the uptake or payment plans for their needs. find little to no access to credit, savings In close collaboration with GIZ, CGAP Theand report use ofexamines financial how mechanisms? smallholder conducted a nationally representa- households in Uganda manage their tive survey of smallholder households income and expenses, along with the

1 United Nations Development Programme’s index of human development in 2014 ranked Uganda 163 out or 188 countries for per capita GDP. 2015 Human Development Report (http://hdr.undp.org/en/2015-report). 2 As reported by African Development Bank Group (http://www.afdb.org/en/countries/east-africa/uganda/uganda- economic-outlook/) from African Economic Outlook (http://www.africaneconomicoutlook.org/en/). 3 https://www.cia.gov/library/publications/the-world-factbook/geos/ug.html. 4 .

http://www.new-ag.info/en/country/profile.php?a=2414

1 National Survey and Segmentation of Smallholder Households in Uganda

issues they face that threaten income household priorities. financially or emotionally, and drives smallholderand often lead sector, to financial exploring instability. household It Agriculture is not easy. It is a labor of then describes financial inclusion in the- intensity, risk, uncertainty and chal- clusion, including mobile phones and lenges, as farmers know from their own tools that are essential for financial in experience. It is also a labor of love. Smallholder farmers in Uganda are pas- awareness,national identification barriers, and documents, interests. The as sionate and committed to agriculture. sectionswell as adoptionof the paper of financial that follow products, then They take pride in their work, want to outline meaningful segments of the continue in it and try to look for oppor- smallholder population in Uganda, map- tunities to advance their work. The real- ping out groups of smallholder farmers ities of agriculture may cause a farmer that matter for fostering greater product to take pause and consider diversifying, adoption, and delving into their demand or opt for other, steadier streams of income, if ever they present themselves, explanation of the research methodology but a transition away from agricultural isfor included various infinancial Annex 1mechanisms.. A full activities is more likely to be driven by need, not a dislike for farming. This report has three main goals: Smallholder agriculture in Uganda is 1. Build the evidence base for those also work that is fairly contained to the household itself. These households en- assumptions and/or isolated obser- gage in agriculture relying mostly on vationsworking pair in agricultural with known, finance reliable so repthat- their own resources, with limited out- resentative data about the population. side assistance. What little support they do solicit typically comes by way of family and friends, or other members of the realities of smallholder farmers in community. More formal entities—such 2. UgandaConnect that readers could with otherwise the unique be - inputs providers, resellers, buyers, or roneously generalized from other otheras financial entities often institutions, a part of agricultural an agricul- smallholderoverlooked, farmer oversimplified, markets. or er tural value chain—play only a small part 3. Catalyze conversations about “what in the Ugandan smallholder ecosystem. next” for smallholder-farmer-centered strategies, products and approaches Important factors that facilitate agricultural, as well as Five fundamental characteristics of smallholder households in Uganda can Thehousehold actual survey finance. and full body of re- help the community of practice foster - their greater productivity and resiliency. withinsearch communitieswill support ofa numberpractice, of for finan both 1. Common dependence on agricul- thecial nearand agriculturaland long term inquiries. that arise ture: Agriculture provides the main income stream into the household, Key Findings and supports nearly all of the household activities. Majorities of fami- Agriculture is the lifeblood of the small- lies consume what they grow, trade holder household in Uganda. It provides goods for other necessities, and sell the primary sustenance of the house- their crops for income. One crop hold, gives the household strength, can be used in each of these ways,

2 National Survey and Segmentation of Smallholder Households in Uganda

but households still often fall short opportunity if smallholders could of their monthly needs. Most of the see it as a tool that is relevant for households live at or below the pov- them and their livelihood. Smallhold- erty line, and many live in extreme ers also lack some of the basic tools poverty. They work hard, have big aspirations and take pride in their accomplishments. (IDs)for digital. A large financial majority services, of smallhold such as- ernational farmers identification have used a mobile documents phone 2. There is promise with a new gen- in the past, so familiarity is relative- eration of smallholder farmers: The smallholder farming community not competing against other formal, includes both tenured, seasoned, ex- orly high.even Formalinformal, financial channels services. They are perienced farmers who have lived competing against a consumer’s in- through both the pains and abun- terest level and perceived relevance dant yields that come with farming, in using them. as well younger, newer farmers who bring with them more modern per- 5. Risky practices run counter to fi- spectives, vitality, and an ambitious nancially sound desires: In prac- mindset. Younger farmers are more tice, smallholder households do not educated, and some may also consid- have savings, do not have access to er leaving agriculture if a compelling funds in the event of an emergency, alternative arises. and do not have insurance or any other way to mitigate risk. Their as- 3. The smallholder farming ecosys- - tem lacks financial and agricul- cially astute, responsible, and even tural infrastructure: Smallholders prosperouspirations, however, mindset . reflect They awant finan to have limited connectivity to a value save, they want to insure their ac- chain, or any formal suppliers for tivities and they want to have more their agricultural activities. Involve- options for mitigating risk. The ap- ment with buyers or resellers is often in the context of loose value chains, yet diminished due to lack of access meaning that the transactions occur andpetite the for other financial challenging security circum has not- without a contract. Farmers also have stances they face. limited market access due to lack of transportation, and they know these Financial mechanisms constraints mean they might not get the best market price for their goods. Perhaps the most important element of

4. There is muted knowledge and respond to the relevant needs and de- perceived relevance of formal fi- siresidentifying of smallholders financial mechanismsis understanding that nancial mechanisms. Many small- what and how smallholder farmers pri- holder farmers operate without oritize. They are farmers at heart, and

- focus is on their farm. Smallholder farm- using formal financial services such- erstheir think profession about the defines farm them,needs soand their are nanceas bank institutions accounts, (MFIs), nonbank and finanmost driven by working toward a sustainable arecial notinstitutions even exposed (NBFIs), to theseor microfi insti- livelihood to provide for their family. tutions. The majority of smallholder farmers have never been in a bank. To that end, they have the greatest appe- Mobile money earns a high level of awareness and presents a potential them afford agricultural inputs such as tite for financial mechanisms that help

3 National Survey and Segmentation of Smallholder Households in Uganda

seed and fertilizer. There is also signif- farmers do not want to leave agricul- icant desire for products that can help ture. Instead they want to take what them afford school fees, which is a per- they do, and do it better. They want vasive issue in Uganda. practices and behaviors, and then to move away from risky financial Strategies stability. Many feel they are barred Five strategies emerge to cultivate fromfind a waymechanisms to build greater that caneconomic help them do just that. They feel they do mechanisms: - uptake and use of relevant financial tion, a mobile phone, enough money 1. Equip smallholder households tonot open have an the account, right type and of sometimes, identifica with identification cards and the smallholder farmers are simply not basic tools needed to become fi- aware how these mechanisms are nancially included, and educate relevant to them. Empathizing with them about the uses of mobile their struggles and appealing to what phones: smallholder farmers ultimately want is essential because brick-and-mortar for their lives can potentially break institutions Education are too distant in digital from finance home down intangible (as well as tangible) - barriers to product adoption. portation. Phones should be seen as moreand households than tools lackfor communication;adequate trans 4. Pair immediate needs with long- - term desires: The survey tested cial transactions or as information potential dual-mode products that channelsand also perceivedfor important as tools agricultural for finan combined both short- and long-term communications. Individuals within - households also need to have national clude insurance, loans that include bankingbenefits or to savings farmers. accounts, Loans that mobile in money accounts that include savings 2. IDBuild cards on required the inherent to open trustaccounts. of fi- and similar bundled products can go nancial mechanisms: Smallholder a long way in appealing to immediate households are keenly aware of mo- needs and conditioning a desirable bile money and trust mobile money long-term practice. Further, farmers services about as much as they do do not want to see their hard labor

institutions. The majority of small- pests, and therefore, want access to holders,banks and however, other nonbank have not financial been insurancesquandered or dueeven toconvenient bad weather and re or- inside a bank and may not even be liable information to mitigate those connected to an informal lending risks. or savings circle. Financial services providers looking to speak to this 5. Think “economies of scale”: Just group should capitalize on the basic under half of smallholder farmers levels of trust that smallholder farm- can be characterized as “farming for ers have in these institutions to build sustenance.” The economic value a meaningful value proposition. for investors and providers is found in both the size of this population 3. Empathize with life desires and and the lack of competition for its circumstances: Most smallholder attention.

4 National Survey and Segmentation of Smallholder Households in Uganda

B. ABOUT THE PROJECT

Working to build the evidence base on with a comprehensive attitudinal and smallholder farming households, GIZ behavioral research program in Janu- and CGAP sought to explore in more ary 2015. It consisted of a survey with - an accompanying household listing and al lives of smallholder households in a segmentation. The research sought to Ugandadetail . the This financial research and project agricultur began

answer three key questions.

“What do we need to know or “How can ﬁnancial “What type of market do to help smallholder farmer mechanisms respond to the strategies and approaches can households build resilient and relevant needs and desires of culvate uptake and use of producve livelihoods?” smallholders?” ﬁnancial mechanisms?”

Existing Research and Stakeholder or a business?), and do household Discussions. Building on other house- members, especially youth, see a fu- hold surveys in sub-Saharan Africa (e.g., agricultural censuses, Living Standards what is the level of comfort with digi- Measurement Study, FinScope, AgFiMS), ture in farming? On the financial side,- as well as the 2013 CGAP global segmen- nels and service providers? tation,5 this methodology and survey tal financial services and other chan instrument were designed to answer a ■■ Opportunities to improve financial inclusion for each segment of small- households in Uganda:6 holder households. number of questions about smallholder mechanisms does each segment of ■■ Understanding and segmenting smallholder households What demand, financial both smallholder households. What are in terms of customer needs (crop stor- the key characteristics of the small- age, transfer, build, secure, etc.) and holder sector at the national level products (e.g., credit, deposit, insur- (e.g., demographics, poverty sta- ance)? What informal and formal suite tus, hectares, crops and livestock, - segment currently use and where are lationships)? What segments of opportunitiesof financial mechanisms to add value doeswith eachnew smallholderlevel of intensification, households emerge? market re services and/or delivery channels?

■■ Attitudes and perceptions of - smallholder households. How do ed a series of deep dives into the exist- smallholder households perceive ingThe research first months in the of smallholder the project space includ to their agricultural activities (i.e., is it seen as a subsistence activity determine what questions had already been asked, identify their findings, and 5 See Christen, Robert Peck, and Jamie Anderson. 2013. “Segmentation of Smallholder Households: Meeting the Range of Financial Needs in Agricultural Families.” Focus Note 85. Washington, D.C.: CGAP, April. 6 CGAP retained the services of InterMedia to manage the survey in partnership with Ipsos Uganda. Additional national surveys and segmentations of the smallholder sector, led by CGAP, are also underway in Tanzania, Côte d’Ivoire, and Bangladesh.

A national survey and segmentation of smallholders in Mozambique was released in March 2016.

5 National Survey and Segmentation of Smallholder Households in Uganda

determine how to drive our objectives to well as World Bank Group colleagues, complement and expand on them. The including the Living Standards Measure- research team consulted several sources ment Survey (LSMS) team. in the process of building the program, including IFC, Dalberg, Finmark Trust, The extensive secondary research and FAO, GIZ, IFAD, and the World Bank. The secondary research brought a series of a gap in information about the actual needs,discussions desires with and stakeholders perceptions identified of small- with stakeholders. holder households. There seemed to be questions that informed discussions GIZ has been a close partner for this about smallholder households in Uganda research with smallholder households thatsignificant focused amounts on either of datatheir andagricultural insights because of its central role in advanc- - ing to date had taken a more comprehen- cularly among rural and agricultural siveactivities view orof theirthe household financial lives,. This but research noth householdsing financial. Thisinclusion coordination in Uganda, was parti im- project also sought to connect the agri- portant to inform the CGAP research, and - its results will contribute to GIZ’s market amine the interactions between the two. research and strategy. Several additional cultural data to the financial data to ex stakeholders and organizations also con- Identifying Target Group of Small- tributed valuable insights and consider- holder Households. Discussions with ations into the design of the research consultants and stakeholders in sub- project as key informants, and took part Saharan Africa and extensive desk re- in an informal technical working group to search7 concluded there is no clear review and guide the research. Some of agreement on the characteristics that these key organizations included African Development Bank, TechnoServe, FINCA heterogeneity of this client group. As a Uganda, Grameen Foundation, Uganda resultdefine ofa smallholder,both of these due lines in partof investi to the- Coffee Farmers Alliance, Opportunity gation, a matrix was developed for each International, Uganda Coffee Develop- of the key criteria that could be used ment Authority, KfW, Mercy Corps, the to distinguish smallholder households Bill & Melinda Gates Foundation, as from other households (see Table 1).

TABLE 1. Key criteria in defining smallholder households Key Criteria Considerations Market orientation Subsistence vs. market-oriented vs. hybrid Landholding size Threshold Labor input Family vs. hired Income Shared income from farming, multiple sources Farming system Technology, irrigation Farm management responsibility Owner, influence over how to farm Capacity Storage, management, administration Legal aspects Formal vs. informal Level of organization Member of group—producer, supply chain, service provider

October 2013. 7 Defining Smallholders: Suggestions for a RSB smallholder definitions; Roundtable on Sustainable Biomaterials;

6 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 1. Average size of agricultural holdings (in hectares) in Uganda

1.6 1.1 1.0 1.1 0.8

Total Uganda CentralEastern Northern Western

The desk research also found a range households in a way that was as inclu- - sive as possible, without diluting or distorting the population representa- variationsof definitions in their of a agricultural smallholder sectors house. hold across countries, reflecting the- two key criteria—landholding size and ers solely by their landholding size. The livestocktion. The count—as identification the measurestarting point used rangeSome differed governments greatly define across smallholdAsian and for identifying the target group for sam- African countries—from a maximum of ple selection. A series of self-identifying 2.5 hectares in India up to a maximum 46 hectares in Malaysia. In Uganda, re- ensure that each smallholder house- search from the Uganda Bureau of Sta- holdperception selected questions for the wasstudy also viewed asked ag to- tistics Census of Agriculture pointed riculture as a meaningful part of the to smaller average farm sizes, which household’s livelihood, income and/or consumption. our target group. In Uganda, the aver- agefurther farm reinforced size ranged the between justification 0.8 and of Before the survey, a household listing 1.6 hectares, depending on the region of exercise was conducted to identify po- the country (Figure 1).8 tential households to include in the survey sample. The listing exercise targeted The research team established a high smallholder households with the follow- watermark to identify smallholder ing criteria:

FIGURE 2. Listing criteria to identify relevant smallholder households

Household with up to 5 hectares OR Agriculture provides a meaningful Farmers who have less than: contribuon to the household AND 50 heads of cale; or livelihood, income, or consumpon 100 goats/sheep/pigs; or (self-idenﬁed) 1,000 chickens

Listing Operation and Methodology. to construct a reliable sampling frame. CGAP’s global implementation partner The listing operation was implemented for the national surveys of smallholder households, InterMedia, worked very partner in Uganda. closely with the Uganda Bureau of by Ipsos Uganda, InterMedia’s local field Statistics to conduct a household listing operation in randomly selected enumer- sample,9 ation areas from 15 July to 7 August 2015 urbanThen, and using rural a areas stratified, based on multi-stage the 2014 each region was classified into

8 Census of Agriculture; Uganda Bureau of Statistics, Ministry of Agriculture, Animal Industry and Fisheries, 2008/2009. 9 The methodology and design are detailed in Annex 1.

7 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 2. Framework for the smallholder questionnaire Household Agricultural Mobile Financial Section Demographics economics practices phones services Relationship Income Land Use (own Formal ownership or borrow) institutions Marital status Jobs Crops grown Types of Less than phones formal institutions Age Government Livestock Barriers Informal payments financial service providers Examples School attendance Saving Value chain Habits Importance of topics Income Investing Market Products Borrowing covered relationship Decision-making Emergency Water Products planning Financial situation Risk Labor mitigation Progress out of Inputs Poverty Index (PPI) Storage Coping

population census, and the sample was ■■ Considering the objectives and needs selected independently in each urban of the project. and rural stratum. The 216 (EAs) were randomly selected as primary sampling ■■ Accounting for stakeholder interests units with probability proportional to and feedback. the number of households in the EAs, and These foundations led to a framework then 15 smallholder households were for the survey instrument for sharing - across stakeholders, and to ensure the ity, which yielded a total of 3,240 small- research captured all of the necessary holderselected households in each EA in with the equal sample probabil. elements of a smallholder household. Questionnaire Design. - The framework was built around the naire design process began by using the sections shown in Table 2. secondary research and The stakeholder question discussions as core inputs into the Organization of the Survey. - measurements to shape the survey in- tionnaire was divided into three parts strument. This process also involved to capture the complexity inside The smallques- - doing the following: tions asked of all relevant individuals in defining the end goal of the research by theholder household, households, not just with one certain household ques ■■ Drawing from existing survey member (Table 3).10 It was designed instruments. in this way to capture the complete

10 The three questionnaires can be found in the user guide that accompanies the data set for this research.

8 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 3. Design of smallholder questionnaires Household survey Multiple-respondent Single-respondent questionnaire survey questionnaire survey questionnaire Head of the All household One randomly household, spouse, members over 15 years selected adult in the Target or a knowledgeable old who contributed to household respondent(s) adult the household income or participated in its agricultural activities • Basic information • Demographics • Agricultural on all household • Agricultural activities activities members • Household economics • Household Topics covered • Information about economics household assets • Mobile phones and dwelling • Formal and infor- characteristics mal financial tools portrait of the smallholder household, languages to ensure the integrity and as some members of a household may work on other agricultural activities in line with social and cultural customs. independently, without the full com- appropriateness of the questions were prehension of their involvement and Data collection took place from 16 responsibilities by members of the August to 7 September 2015, using household. computer-assisted data collection tools that regularly yielded data for analysis nine languages—Acholi, Ateso, Langi, Luganda,The questionnaire Lugbara, wasLugishu, translated Lutooro, into smallholderand quality household control to survey provide was timely im- Ngakaaramojong, and Runyankole—and plementedfeedback toby fieldIpsos staff. Uganda, The InterMe Uganda- then pretested and validated in all nine

dia’s local field partner.

9 National Survey and Segmentation of Smallholder Households in Uganda

C. FINDINGS11 1. SMALLHOLDER HOUSEHOLD DYNAMICS IN UGANDA: WHO THEY ARE Smallholder farmer households are decision-making role when it comes typically headed by men who have to the agricultural activities of the low levels of education yet maintain household. an optimistic viewpoint on their future. There is both maturity and youth in the Ugandan smallholder population. Nearly Smallholder households span Uganda, half of heads of households are under with the greatest density concentrated - in the Western and Eastern regions isthe under age ofthe 40 age (45 of percent).30 (Figure A 5) signifi. Just (23 percent) are concentrated in the overcant half portion, (55 percent) approximately are 40 one-fifth,or older, Northernof the country. region, Less where than national one-quarter statistics show the highest poverty levels,12 spectrum (601 years old). leaving the smallest population of small- and one in five is at the far end of the age holder farmers in the Central region Smallholder heads of households typi- (16 percent) (Figure 3), which is the cally manage their households, families most economically thriving part of the and livelihoods with limited formal edu- country.13 cation, rarely surpassing primary school - A man is three times as likely to be the ucation, and 64 percent did not contin- head of a smallholder farming house- ue(Figure their 6).education One-fifth past have primary no formal school ed. hold as is a woman (77 percent men vs. Only 16 percent advanced through sec- 23 percent women) (Figure 4). While ondary school. There is a sharp gender households are male-dominated, wom- difference in education levels (Figure 7). en do play an important, if not critical, Female household heads are even more

FIGURE 3. Region FIGURE 4. Gender of head of household

Central Female Western 16% 23% 31%

Eastern 30% Northern Male 23% 77%

Sample: Smallholder households, n52,870 Sample: Smallholder households, n52,870

11 Graphs and tables in the main body of the report include references to the unweighted base size, and, therefore, at times, will not look proportional to graphs that show subsets of other graphs. Due to rounding, not all percentages in charts total 100. 12 As reported by The African Development Bank (http://www.afdb.org/en/countries/east-africa/uganda/uganda- economic-outlook/) from African Economic Outlook (http://www.africaneconomicoutlook.org/en/). 13 Ibid.

10 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 5. Age of head of household FIGURE 6. Highest education attained by head of household 15-29 19% 45%

30-39 26% 24% 20% 40-49 23% 9% 2% 50-59 13% Never Some Primary Secondary Higher 60+ 19% aended primary educaon school

Sample: Smallholder households, n52,870 Sample: Smallholder households, n52,870 likely to have no formal education The smallholder household size and (47 percent, vs. 12 percent of men, have composition can vary across the popu- never attended school). lation, including some very small households of just one person (6 percent), - as well as those with eight or more hold heads are married or cohabiting, people (18 percent).14 Seven percent Three-quarters of smallholder house of smallholder households have 10 or separated, or widowed (Figure 8). The more members (Figure 10). The median genderand about of the one-fifth head of household are divorced, dif- fers by marital status; men lead sin- of households with double that number gle/never married and married homes household size is five, and the presence- while women lead divorced/separated/ cumstances and family life, and the im- widowed homes (Figure 9). may point to the general fluidity of cir

portance of risk mitigation. This fluidity FIGURE 7. Highest education attained FIGURE 8. Marital status of head of by gender of head of household household

Single / never married 4% 12% Divorced / 47% Separated / 48% Widowed 20% 34% 40% 18% Married / MaleFemale Cohabing 76% Primary or higher Some primary Never aended school

Sample: Smallholder households, n52,870 Sample: Smallholder households, n52,870

same dwelling unit, eat together from the same pot, and share most household expenses. Visiting relatives and domestic 14 workersFor the purposes are not considered of this survey, members “household” of a household was defined and as will, a group therefore, of related not be or included unrelated in personsthis study who. The live listing together manual in the in the user guide seems to contradict this: “Note, however, that domestic servants and other workers living and eating in the same household should be included as household members.”

11 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 9. Marital status by gender of FIGURE 10. Number of people in head of household household

7% 30% One 6% 83% Two 9% 93% 70% Three 13% Four 15% 17% Five 15% Single / never Married / Divorced / Six 14% married Cohabing Separated / Seven 11% Widowed Eight or more 18% Male Female

Sample: Smallholder households, n52,870 Sample: Smallholder households, n52,870

could be positive (e.g., a new breadwin- ner arrives to contribute to the house- have money only for food and clothes hold), but also a challenge (e.g., the very (Figuretheir households, 13). and another two-fifths young and very old who need special care and are not in a position to contrib- Smallholder farmers’ outlook on life and ute much to the household). their agricultural work is in stark contrast to their households’ circumstances. - Despite limited means and economic cant because many households fall be- vulnerability, most take the position that lowA large the poverty household line size of US$2 is also.50 a signifi day15 they work hard to be among the best at (Figure 11). Using the lower PPI score what they do (93 percent), their actions of US$1.25 a day, the comparison is determine their lives (86 percent), and their successes are due to hard work all smallholder households live on less (84 percent). They also align with the thaneven that starker: amount roughly per day one-quarter (Figure 12) of. perspective, “I always look for oppor- Smallholder farming households live tunities for improving my situation,” without much of a cushion to absorb suggesting a proactive rather than a re- active approach to their lives (Figure 14). of all smallholder households typically Far fewer farming households take a doadditional not have expenses. enough moneyRoughly for two-fifths food for more passive approach, believing it is

FIGURE 11. Poverty status of household FIGURE 12. Poverty status of household

$2.50 or Below above a day $1.25 a day 33% 26%

Below $1.25 or $2.50 a day above a day 67% 74%

Sample: Smallholder households, n52,870 Sample: Smallholder households, n52,870

15 From Progress out of Poverty Index 2013, Grameen Foundation; http://www.progressoutofpoverty.org/.

12 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 13. Household’s current financial situation (self-assessed)

41% 37%

19%

Not enough money for Enough money for food Enough money for food Aﬀord to buy certain food and clothes only and clothes and can save a expensive goods bit, but not enough to buy expensive goods

Sample: Smallholder households, n52,870 not wise to plan too far ahead because Land tends to be in small plots;16 roughly one’s luck might factor more heavily into two-thirds of Ugandan smallholder the future than planning (50 percent). households own less than two hectares (app. 2.5 acres) of land (Figure 15). An- Farm as source of income and other one-sixth of the households have subsistence

Typically, Ugandan smallholder farmers betweenSmallholders two andin Uganda five hectares. primarily grow individually own their plots of land, ei- food and staple crops, and there is a - collection of commonly grown crops. Maize and beans are the most com- ofther these through farms a leasefall under or certificate, customary or lawun mon, followed by cassava, sweet pota- (Tableder customary 4), which law. means Roughly there two-fifthsis usually toes and groundnuts (Figure 16). Only small percentages grow cash crops, State and communally owned farms are which tend to be coffee and sugar cane inno the official minority, documentation and mostly of concentrated ownership. (Figure 17). The median number of in the Northern region. crops per household is seven, and just

FIGURE 14. Do you agree or disagree with the following statements?

I always work hard to be among the best at what I do 93% 5%

My life is determined by my own acons 86% 11%

I always look for opportunies for improving my situaon 86% 10%

When I get what I want, it is usually because I worked hard for it 84% 12%

It is not always wise for me to plan too far ahead because many 50% 43% things turn out to be a maer of good or bad fortune

AgreeDisagree

Sample: Smallholder farmers, n52,771

estimates of land size usually lead to errors. Carletto, Gourlay, Winters. World Bank. “From Guesstimates to GPSstimates,” 16 JulyLand 2013 size. ishttp://www-wds a difficult thing .worldbankto measure. org/external/default/WDSContentServer/WDSP/IB/2013/07/30/000158349_20accurately. Many recent examinations of land measurement say that using farmer 130730084245/Rendered/PDF/WPS6550.pdf.

13 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 4. What is the form of ownership of your land?

Total Central Eastern Northern Western Individual ownership with 31% 47% 28% 19% 37% lease or certificate Individual ownership under 40% 24% 41% 52% 37% customary law Communal (resources are 8% 6% 5% 20% 4% shared) State ownership 0% 1% 0% 0% 1% Other 10% 7% 11% 7% 11% Don’t know 10% 15% 14% 2% 11%

Sample: Smallholder farmers who par`ticipate in agricultural activities, n55,203

FIGURE 15. How many hectares (ha) of agricultural land do you own?

None 1% Up to 0.5 ha 24% 0.5 to 1 ha 22% 1 to 2 ha 20% 2 to 3 ha 10% 3 to 5 ha 7% More than 5 ha 4% Don't know 11%

Sample: Smallholder households reported land size, n52,765

FIGURE 16. Which of the following crops FIGURE 17. Which of the following do you grow? Food and Staple Crops crops do you grow? Cash Crops

Maize 83% Coﬀee 21% Beans 80% Cassava 75% Sugar cane 7% Sweet potatoes 66% Groundnuts 56% Bananas 44% Coon 3% Millet 33% Sorghum 32% Tobacco 1% Irish potatoes 18% Simsim 17% Tea 1% Soya beans 12% Rice 10% Cowpea 6% Cocoa 0% Pigeon pea 5% Field peas 2% Sesame 0%

Sample: Smallholder farmers who participate in agricul- Sample: Smallholder farmers who participate in agricultural activities, n55,203 tural activities, n55,203

14 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 18. Crop type by percentage of consumption, sale, or trade

Cassava 91% Sorghum 72% 39% 36% (n=3,815) 11% (n=1,512) 10% Beans 91% Simsim 68% 54% 59% (n=4,262) 14% (n=795) 9% Bananas 86% Cowpea 65% 30% 30% (n=2,406) 5% (n=310) 6% Sweet potatoes 85% Rice 64% 20% 64% (n=3,394) 8% (n=483) 5% Maize 83% Soya beans 63% 62% (n=4,340) 11% (n=609) 4% 55% Irish potatoes 81% Sugar cane 63% 43% 36% (n=930) 2% (n=377) 2% Groundnuts 75% Pigeon pea 58% 47% 28% (n=2,809) 9% (n=218) 15% Millet 74% Coﬀee 6% 44% 78% (n=1,513) 9% (n=1,125) 2%

Consume Sell Trade Consume Sell Trade

Sample: Smallholder farmers, n55,203 Multiple responses allowed (% of people who grow each crop)

households consume at least a portion more than eight crops. of their crops. over one-quarter (27 percent) grow Households use their crops in multiple A single crop itself can also serve mul- ways, including for their own consump- tiple purposes. For example, 91 percent tion. Consumption rates the highest of of cassava farmers say they consume it, the three main uses (consume, sell, or - trade), particularly in the case of most mately one-tenth trade it. food or staple crops, and even for some just under two-fifths sell it and approxi cash crops (Figure 18). An estimated CGAP’s National Survey and Segmen- nine out of 10 smallholder farmer tation of Smallholder Households in

FIGURE 19. Number of crops grown FIGURE 20. Number of crops grown for consumption for selling

1-4 crops 68%

1-4 crops 50% 5 crops or more 49% 5 crops or more 19% Do not consume Do not sell what they grow what they grow 1% 13%

Sample: Smallholder farmers who grow crops, n55,088 Sample: Smallholder farmers who grow crops, n55,088

15 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 21. Number of crops grown FIGURE 22. Which of the following for trading crops that you grow is the most important to you and your family?

1-4 crops Cassava 25% 22% 5 crops Maize 24% or more Beans 16% 2% Bananas 6% Sweet potatoes 5% Coﬀee 5% Sorghum 4% Groundnuts 3% Simsim 2% Millet 2% Do not trade Irish potatoes 2% what they grow 76% Rice 1% Other 4%

Sample: Smallholder farmers participating in agriculture Sample: Smallholder farmers who grow crops, n55,088 who grow at least one crop, n55,088

17 also explored crop growth hold a greater utility for selling than they do for consumption, suggesting there is Mozambiquejust a few select crops that are of utmost some monetization that farmers factor importanceand use. In Mozambique, to the country’s there smallholder tend to be in when planning their land use. farmers. Most Mozambican smallholder households grow maize (88 percent) and the next most common crop is a distant raise livestock of any kind (Figure 23), second, cassava (55 percent), followed by andThree-fifths those who of raise smallholder livestock do farmers so for beans (47 percent). Maize is considered both consumption and sale. Indigenous the most important crop (66 percent) chickens (layers) are the most common among its growers; no other crop comes form of livestock, followed by indigenous close to the importance of maize. Only goats, goats raised for meat, broilers 10 percent of cassava growers consider (indigenous chickens) and indigenous it their most important crop, coming in cattle (Figure 24). second to maize. The majority of those who rear indig- Uganda shows a different dynamic, enous chicken do so for consumption with much more crop diversity. Uganda (Table 6); a similar dynamic is present has several crops that are important to for nonindigenous chickens. Cattle, pigs, smallholders. The vast majority of small- indigenous pigs and indigenous cattle, holder farmers in Uganda grow cassava, however, are reared mostly for income, maize and beans; the most important to not consumption, purposes by small- them and their families is a split between holder farmers. cassava and maize (Figure 22). Cassava and maize also play an important role Women have a significant role in in household consumption. Maize, decision-making beans, cassava and coffee also register as important crops for sale (Table 5) - among a smaller, more niche group of holder farmer households (Figure 4), farmers. There are several crops that yetMen headdecision-making over three-quarters on agricultural of small

http://www.cgap.org/publications/national-survey-segmentation-smallholder-households-mozambique

16 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 5. Which of the following crops that you grow do you consume the most/get the most money from selling? Consumption (n55,058)* Selling (n54,374) Cassava 33% 8% Maize 19% 24% Beans 16% 13% Bananas 10% 3% Sweet potatoes 8% 2% Sorghum 4% 2% Millet 2% 4% Irish potatoes 2% 2% Groundnuts 1% 8% Simsim 1% 7% Coffee ,1% 11% Rice ,1% 4% Soya beans ,1% 2% Other 1% 10%

*Responses add up to less than 100 percent due to rounding Sample: Smallholder farmers, n55,517 (% of smallholder farmers participating in agriculture who grow and consume/sell at least one crop) activities is not as gendered. In fact, to make the decision solely, without the women’s roles as decision-makers are other. important within smallholder households. In some cases, the largest share of Dedicated to agriculture and agricultural decisions are jointly made by looking to expand their activities men and women (Figure 25), including when households are headed by men. Smallholder farmer households in- When they are not joint decisions, in clude both tenured and newcomer some cases men are slightly more likely farmers as well as novices in the sector. farmers, reflecting both experienced FIGURE 23. Do you have any livestock, farmed for more than 10 years. As many herds, other farm animals, or poultry? (41Close percent) to three-fifths range in (58 experience percent) from have very new (less than two years, 5 percent) to moderately experienced (six to 10 years, 19 percent) (Figure 26). No 40% It is mostly the youngest generation (below 29 years old) that is newer Yes to farming, versus individuals adopt- 60% ing farming as a livelihood late in life (Figure 27).

Consistent across households, farming

Sample: Smallholder farmers who participate in agricul- emerges as a life choice and part of an tural activities, n55,203 identity, which can offer some insights

17 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 24. Which of the following do you rear?

Indigenous chickens –layers 32% Indigenous goats 29% Goats –meat 25% Indigenous chickens –broilers 20% Indigenous cale 19% Chickens –broilers 19% Pigs 18% Cale –beef 15% Chickens –layers 14% Indigenous pigs 11% Cale –dairy 10% Ducks 7% Sheep 5% Indigenous sheep 3% Bees (number of hives or boxes) 3% Goats –dairy 2% Turkeys 2% Fish (number of ponds) 0%

Sample: Smallholder farmers who have any livestock, herds, other farm animals or poultry, n53,184 Multiple responses allowed

TABLE 6. How many of each of the following do you. . .? Rear and get an income from* Rear for consumption* Pigs 79% 23% Indigenous pigs 76% 14% Cattle—dairy 71% 44% Goats—meat 70% 31% Indigenous goats 68% 29% Indigenous chickens—layers 59% 78% Indigenous chickens—broilers 59% 67% Indigenous cattle 59% 32% Cattle—beef 59% 30% Chickens—broilers 58% 74% Chickens—layers 54% 76% Sheep 54% 20% Indigenous sheep 53% 13% Bees (number of hives or boxes) 51% 50% Turkeys 50% 34% Goats—dairy 49% 16% Ducks 46% 61% Fish (number of ponds) 17% 17%

*Percentages show incidence within those rearing that specific type of livestock who are getting an income and total who are rearing for consumption. This captures overlap between the two categories. Sample: Smallholder farmers who have any livestock, herds, other farm animals or poultry, n53,184 Multiple responses allowed

18 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 25. Generally, who makes decisions on the following agricultural activities?

When to harvest 35% 32% 28% 3% Both When and where to sell crops 34% 26% 31% 6%

Quanty of crops to sell 34% 26% 31% 6% Wife/ girlfriend Planng me 32% 28% 33% 3%

Purchase of farm inputs 32% 27% 33% 6% Husband/ boyfriend Where to borrow money to support 31% 22% 33% 12% agricultural producon Not Quanty of livestock to sell 31% 20% 32% 15% applicable / Don't know When to sell livestock 30% 20% 31% 16%

Sample: Smallholder households, n52,870

into the motivations of this popula-

Eighty-eight percent intend to keep workingtion, despite in agriculture their dire financial (Figure state.28). FIGURE 26. How many years have you This intent carries across farming ten- been farming? ure and both genders. In fact, roughly 1% 5% seven in 10 of the newest smallholder Less than 2 years farmers (farming less than two years) believe they will continue farming as 19% 2-5 years well (Figure 29). Their dedication to 6-10 years agriculture is high, even despite their 58% - 17% More than 10 ly nine-tenths of those who self-report years thatfinancial they situation “don’t have (Figure enough 30), as money near Don't know for food” want to continue working in agriculture. Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296 Agriculture is not only what feeds the household, it is a livelihood that farmers enjoy. Nearly all agree with the statement, “I enjoy agriculture,”

FIGURE 27. How many years have you been farming? By age of respondent

28% 68% 23% 82% 86% 94%

47% 18% 11% 12% 7% 3% 5% 5% 2% Aged 15-29 Aged 30-39 Aged 40-49 Aged 50-59 Aged 60+ Five or less yearsSix to 10 yearsMore than 10 years

Sample: Smallholder farmers, n52,296 (% of smallholder farmers who participate in household’s agricultural activities and in each age category)

19 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 28. Do you intend to keep Farming realities introduce a working in agriculture? three-way clash: dedication, Don’t know commitment faces off against risk 2% and dire financials, as well as some No willingness to alter one’s livelihood 10% Commitment to agriculture, the enjoyment it brings, and the desire to expand

Yes thirds would take full-time employ- 88% mentare met if withoffered conflicting (66 percent), thoughts. further Two- illustrating just how hard it is to earn income through agriculture, as well as Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296 just how opportunistic a farmer has to be to support their family and home- stead (Figure 31). (90 percent), a large majority of smallholders want to expand their work The younger generation of farmers (86 percent), and many (59 percent) are (aged 15–29) shares those same senti-

(Figure 31). Farmers are fairly united take full-time employment if offered onsatisfied what theywith view what as they their have legacies achieved and itments. (Figure Just 32), over and three-quarters just over four would in their desires for the next generation. 10 feel they would not want to do any Just over two-thirds (68 percent) think other type of work. Those that are of agriculture as the legacy they leave 30 years or older do not feel as strong- their children and a similar proportion ly about taking full-time employment, want their children to continue in agriculture (66 percent). Agriculture is so (Figure 33). hard work, and smallholders know the as only three-fifths say they would do are not in agreement with their chil- agriculture and the willingness to take drenrealities. continuing Just over in athe quarter family of business, farmers full-timeThis conflict employment between iscommitting real. Farming to and others simply are not sure what is hard work, it can be risky, and it still will be right for their children’s future can leave the family in need. That said, (Figure 31).

deeper analysis shows the conflict might FIGURE 29. Do you intend to keep working in agriculture? By number of years in farming

8% 10% 12% 28%

90% 87% 86% 72%

More than ten years Six to ten yearsTwo to ﬁve yearsLess than two years Yes No

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296

20 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 30. Do you intend to keep working in agriculture? By household’s current financial situation (self-assessed)

10% 8% 15% 3%

88% 90% 83% 89%

We don't have enough money We have enough money We have enough money We can aﬀord to buy certain for food for food and clothes only for food and clothes and expensive goods can save a bit, but not enough to buy expensive goods Yes No

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296

FIGURE 31. Do you agree or disagree with the following statements?

I enjoy agriculture 90% 10% I want to expand my agricultural acvies by looking at new products and/or markets 86% 10% I regard my agricultural acvies as the legacy I want to leave 68% 28% for my family I would take full me employment if I were oﬀered a job 66% 31%

I want my children to connue in agriculture 66% 29%

I am sasﬁed with what my agricultural acvies have achieved 59% 39% AgreeDisagree

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296

FIGURE 32. Do you agree or disagree with the following statements?

I would take full-me employement if I were 76% 22% oﬀered a job I am sasﬁed with what my agriculture acvies 58% 41% have achieved

I would not want to do any other kind of work 44% 54% AgreeDisagree

Sample: Smallholder farmers aged 15–29 who participate in household’s agricultural activities, n5860

FIGURE 33. Do you agree or disagree with the following statements?

I would take full-me employement if I were 59% 38% oﬀered a job I am sasﬁed with what my agriculture acvies 60% 38% have achieved

I would not want to do any other kind of work 45% 53%

AgreeDisagree

Sample: Smallholder farmers aged 301 who participate in household’s agricultural activities, n51,436

21 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 7. Do you agree or disagree with the following statements? “Want to get out of farming” Index Criteria I would not want to do any other kind of work Disagree I would take full time employment if I were offered a job Agree I regard my agricultural activities as the legacy I want to Disagree leave for my family

Sample: Smallholder farmers, n5310

the legacy I want to leave for my family.” opportunity, versus a true intent to This indexed subgroup of smallhold- leavereflects farming one’s .individual Only a small experience group, conand- er farmers is very similar to the entire sisting of less than 12 percent of the sample across many demographics. The survey sample (n5310), shows a more two distinguishing demographics for hardened intent to leave farming. This this subgroup are age and region of the group (Table 7), while relatively small, country; this subgroup is younger (un- said they “would like to do other kinds der the age of 40) than the population of work,” “would take full-time employ- of smallholder farmers (Table 8) and is ment if I were offered a job,” and “do concentrated in the Western region of not regard my agricultural activities as the country.

TABLE 8. Do you agree or disagree with the following statements? All smallholder farmers Smallholder farmers who want Group n52,771 to “get out” n5310 Age 15–39 65% 76% Age 401 35% 24% Central region 15% 18% Eastern region 33% 27% Northern region 23% 12% Western region 29% 43%

Sample: Smallholder farmers

22 National Survey and Segmentation of Smallholder Households in Uganda

2. SMALLHOLDER HOUSEHOLD DYNAMICS IN UGANDA: INCOME AND EXPENSES Farming activities determine household income receive money from family and friends (Figurejobs that 35) yield. Over wages. one-third Roughly of one-fifthfarmers The majority of Ugandan smallholder get income from rearing livestock, poul- households generate income from agri- - culture, and most of that household in- atively stable across demographics. come comes from growing and selling try, fish or bees. Income sources are rel crops. It is likely the household under- Smallholder farmers who contribute to takes other activities to supplement in- the income of their households consis- come; however, the household does not tently shared that growing and selling report income from these activities at a crops are the most important, most re- rate that comes close to the income from liable and most enjoyable farming ac- farming. tivities (Table 9). By comparing these three concepts, data show that a large - portion of smallholder farmers in Ugan- mary job (i.e., where they spend the most ofFour-fifths their time) of farmersis farming report (Figure their 34); pri source with the one they like getting the 80 percent grow something to sell, and mostda equate and with the mostthe one important that is the income most reliable. A much smaller percentage said or bees. the money they earn from rearing live- 35 percent rear livestock, poultry, fish, The percentage of Ugandan smallholder rewarding. households who earn income through stock, poultry, fish, or bees is the most other means is fairly limited. Only a Aside from crop production and live- small portion (5 percent) are also busi- stock, a few smallholder households ness or shop owners, and only about earn income from other agricultural ac- one-tenth of farmers have occasional tivities or sources (Figure 36). Very small

FIGURE 34. What is your primary job? FIGURE 35. Do you generate income from any of the following sources? 1% 3% 4% 8% Growing something and selling it, such as crops, fruits, or vegetables 80% 5% Rearing livestock, poultry, ﬁsh, or bees and selling it or its byproducts 35% Geng money from family or friends 21% 79% Earning wages from occasional job 12% Running own business in retail or manufacturing 8% Running own business by providing services 5% Earning wages or salary 5% Farmer from regular job Geng a grant, pension, or subsidy Laborer 1% Business owner of some sort Professional (e.g. doctor, teacher) Other 6% Shop owner Other Sample: Smallholder farmers, n55,517 Sample: Smallholder farmers, n55,517 Multiple responses allowed

23 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 9. Which of the following income sources is...? Income sources Most important Like getting the most Most reliable Growing something and 67% 65% 64% selling it, such as crops, fruits, or Rearing livestock, poultry, 6% 7% 7% fish, or bees and selling it or Getting money from family 6% 6% 6% or friends Earning wages from occa- 5% 5% 5% sional job Running own business in 4% 4% 4% retail or manufacturing (selling or making goods) Earning wages or salary 4% 4% 4% from regular job Running own business by 2% 3% 3% providing services Getting a grant, pension, ,1% ,1% ,1% or subsidy of some sort Other 4% 4% 4% Don’t know 1% 1% 1%

Sample: Smallholder farmers, n55,517 percentages of farmers are involved in does, nearly half of farmers pick up the the processing of agricultural products, money in cash, in person, while one-third resell or rent land to other farmers for receive it via direct deposit to a bank growing crops. account (Figure 38). Comparatively, the nationally representative Financial More broadly, beyond agriculture, only Inclusion Insights survey of Ugandan 1 percent of smallholder households re- adults in 2015 showed that 6 percent of ceive payments from the government, Ugandan’s adults receive G2P payments such as pension, disability or welfare through a bank, and 2 percent through (Figure 37). Of the small percentage that mobile money.18

FIGURE 36. Are there any other ways that you get income?

4% 4% 3% 2%

Buy/get agricultural products Buy/get agricultural Provide a service to farmers or Rent land to farmers from farmers and process it/ products from farmers/ processors of farming products for farming purposes change it to another form processors and sell it (e.g., renng ploughs, tractors, (e.g., maize to ﬂour) other equipment)

Sample: Smallholder farmers, n55,517 Multiple responses allowed

18 Intermedia Uganda Financial Inclusion Insights (FII) Tracker survey Wave 3 (N53,000, 151), July-August 2015.

24 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 37. Do you receive income FIGURE 38. How do you usually get from any of the following? this government payment?

Personal pick-up in cash 45% Occasional sale of my 39% Direct deposit to a bank belongings account 33% Remiances/monetary or Digital card 6% other help from family 21% Deposit to your m-money 4% or friends account Deposit to another person's Occasional paid assignments, 4% 11%11% m-money account labor for hire Western Union/Money 3% Gram Other donor/NGO beneﬁts 2%% Personal pick-up in check 2%

Government beneﬁts Courier delivery 1% (pension, disability, 1%% Deposit to an agent’s 0% welfare, etc.) m-money account

Sample: Smallholder farmers who receive income from Sample: Smallholder farmers, n55,517 government benefits, n582 Multiple responses allowed Multiple responses allowed

Self-reported expenses are within smallholder families barely break even income; still, smallholders struggle each month and typically have to spend to get by more than they earn.19 Smallholder farmer households largely In Uganda, most households need live below the poverty line, earning 100,000 UGX or less to survive under US$2.50 a day, or, in extreme each month (Figure 39). More than poverty, earning under US$1.25 a day - more than what they need each month. holds said their expenses are 20,000 UGX three-quarters of households bring in (US$6)(Figures or 11 less and each 12). month One-fifth. Twenty-four of house percent said they need between 20,001 FIGURE 39. What is the minimum and 50,000 UGX (US$6 and US$15), and amount your household needs to sur- another 24 percent said they need be- vive per month (for personal expenses)? tween 50,001 and 100,000 UGX (US$15 and US$30). The rest, approximately 20,000 UGX or less (US$6 or less) 20% one-third of smallholder households, 20,001 - 50,000 UGX 24% per month to manage their household (US$6 - US$15) (Figurerequire 39) 100,001. UGX (US$30) or more 50,001 - 100,000 UGX 24% Household income in Uganda usually (US$15 - US$30) surpasses expenses, leaving majorities 100,001 - 200,000 UGX 17% of farming households in a slightly bet- (US$30 - US$59)

200,001 or more This phenomenon should not be taken (US$59 or more) 15% forter financialgranted. Thesituation CGAP month national to month.survey of smallholder households in Mozam- Sample: Smallholder households who gave a minimum amount for households’ survival n52,742 bique shows a different tendency, where http://www.cgap.org/

19 CGAP National Surveys & Segmentation of Smallholder Households in Mozambique, 2015–2016; publications/national-survey-segmentation-smallholder-households-mozambique

25 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 40. What is the minimum amount your household needs to survive per month (for personal expenses) by household’s average monthly income across all sources of money that your household receives?

4% 20,000 UGX or less (US$6 or less) 5% 85% n=477

20,001 - 50,000 UGX (US$6 - US$15) 7% 4% 84% n=613

50,001 - 100,000 UGX (US$15 - US$30) 14% 4% 76% n=632

100,001 - 200,000 UGX (US$30 - US$59) 15% 5% 77% n=511

200,001 or more (US$59 or more) 30% 4% 60% n=509

Don't make enough Breaking even Surplus

Sample: Smallholder households, n52,742

It is households that need more to sur- 20 these vive, however, that are more vulnerable to falling short each month. For smallholdersframework where in Mozambique, households make - smallerexpenses expenses fit into a moretraditional regularly spending and ing 200,001 UGX or more per month larger expenses less often (Figure 41).21 fallinstance, short .three This inis 10more households than double requir the For instance, grocery expenses, trans- percentage of those earning between portation and bills (including utilities, 50,001 and 100,000 UGX each month (Figure 40). While lower income house- expenses. holds are vulnerable, we see more bud- rent or airtime) are more frequent geting within means, and a self-reported Conversely, other larger expenses are lower incidence of falling short. The self-reported data also suggest there is a including educational expenses, emer- made on an infrequent basis (if at all), little bit of extra money each month for gency expenses, or home repairs. the lower-income households that can There is little to no difference between the spending behaviors of male and fe- potentially increasing this group’s at- male smallholder farmers in Uganda. tractivenessbe redirected as into a consumer a financial segment account,. Households in urban areas have these Certainly, the experience of falling short expenses more often than those in rural is far more pronounced in the higher in- areas do. come groups. This research surfaced that transactions Expenses for smallholder households in that one would expect to be made with some regularity, such as utility bills, were the size of the expense. Compared with not made in the recent past. Very few Uganda vary in frequency, depending on

20 Ibid.

21 Expense question displayed in Figure 42 did not include agricultural inputs, such as seed and fertilizer, specifically, and instead focused on broad-based household needs. Farming was only specific as a part of investments.

26 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 41. How often do you make each of the following expenses?

Grocery purchases 50% 30% 18% Bills: ulity bills, airme, rent, etc. 21% 25% 51% Transportaon 20% 67% 11% Medicine, medical payments, hospital charges 8% 79% 11% Investment in business, farm or future 2% 55% 39% Educaonal expenses, school fees 1% 63% 35% Home repairs 1% 37% 58% Emergency expenses 1% 33% 55% Make a large purchase, such as TV, house, etc. 0% 18% 78%

At least once a week Less o en Never

Sample: Smallholder households, n55,517

farming households had paid utility bills Risky money management practices in the past 30 or 90 days. Only about a with few options to mitigate an emergency - positedquarter moneyhad received (Figure money 42). from family or friends, and another quarter had de themselves in at-risk situations. That The limited outward expenditures of Smallholder farmers in Uganda find- households, compiled with the known nancial circumstances, and self-reported expenses, and the recognized distress incomefarmers outweighsendeavor to expenses avoid troubling for most fi that surrounds household budgets ap- suggests otherwise, however, even still, peared as a perplexing phenomenon farmers profess to struggle in a variety within the data. Households have lit- of ways. tle resources, typically bring in limited funds, and are still obligated to pay The self-reported characterization of school fees and household costs, yet are monthly spending matches what self- - reported actual spend in monetary units cies. This could be due to households (Figures 39 and 40). Fewer than two in makingnot transacting extremely with prudent reported decisions frequen 10 smallholders in Uganda feel they al- about what to spend and where to ways or most of the time spend more spend it, even if it means cutting back on than they make. Ugandan smallholder necessities. farmers also share that they often

FIGURE 42. Did you do the following activities AT LEAST ONCE in the past 30/90 days?

43% 39% 41% 30% 28% 27% 25% 23% 21% 22% 18% 17% 5% 6%

Buy airme top- Pay a school fee Deposit moneyReceive money Withdraw moneySend money to Pay ulity bills ups from family family members (electricity, solar members or or friends lantern, water, friends TV)

Past 30 days Past 90 days

Sample: Smallholder farmers, n52,771

27 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 43. How often does the following apply to you?

19% 38% 33% 41% 20%

22% 26% 27% 41% 26% 25% 25% 18% 7% 14% 5% I spend less money than I make I pay my bills on me My savings are larger than my I have an emergency fund to each month debts cover for unplanned expenses Always / Most of the me Somemes Rarely Never

Sample: Smallholder farmers, n52,771 cannot pay their bills on time. This could also means they do not have an emer- potentially support the theory that they gency fund for unplanned expenses make tough decisions about what to pay (Figure 43). and what not to pay each month in order to live within their monthly incomes Most smallholder households in Uganda (Figure 43), even though we also see do not have plans to manage unexpect- that their reported expenses typically ed expenses. Just over one-third have a are less than income (shown earlier). plan to cover expenses associated with a major medical emergency (Figure 44). Having more savings than debt is not a Much smaller numbers of farmers have guarantee, nor is having an emergency plans to respond to a death in the fam- fund. In fact, it is more likely that small- ily, events that affect their crops or holders never or rarely have these than livestock, or the lack of food. The lack have them. Life circumstances often of planning was mirrored by their sav- mean smallholder households need ex- ing habits over the past year. Few save, tra time to pay back loans or lines of - credit. Lack of a monthly surplus (either tution (Figure 45). Only 9 percent said income or agricultural production) theyand evenhave fewer saved save at aat bank a financial in the instipast

FIGURE 44. Does your family have a plan to manage these unexpected expenses, which might result from the following?

Major medical emergency, including illness, injury and childbirth 33% Loss of harvest or livestock due to weather condions or a disease 24% An extended period of me without your own food supply 23%

Death in the family 19%

Loss of a house due to ﬁre, ﬂood or another natural disaster 18%

Loss of property due to the or burglary 15%

Bankruptcy/loss of a job or a business 13%

“Yes” answers Sample: Smallholder farmers, n52,771

28 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 45. In the past 12 months, FIGURE 46. In the event of an have you saved money with any of the emergency, could you get extra money following groups? through relatives sending money or by selling assets? Friends and family 36% Don’t know 2% Savings and credit group 28%

Bank 9% No 34% SACCO 6%

Microﬁnance instuon 3% Yes 64% Cooperave 2%

“Yes” answers Sample: Smallholder farmers, n52,771 Sample: Smallholder farmers, n52,771

12 months; however, informal saving is much higher, and over one-third have of Ugandan smallholder farmers think theyother could options get for extra liquidity. money The from majority rela- on earlier observations, disparities be- tives sending money or by selling some tweensaved with income friends and and expenses family. limitReflecting their of their assets in the event of an emer- ability to put money aside. gency (Figure 46). Still, the possibili- ty of coming up with a relatively small Limited savings leaves smallholder amount of money—114,000 shillings households in a very risky position, es- (approximately US$34)22—in the next pecially because there are not many month causes some pause. Less than

(Figure 47). Half of smallholders said it FIGURE 47. Imagine that you have wasone-quarter either not said possible it was or very they possibledid not an emergency and you need to pay 114,000 shillings. How possible is it know if it was possible to come up with that you could come up with 114,000 this money in a month. For those who shillings within the next month? said it would be possible, they would most likely draw the money from family or friends or get the money from their Don't know 2% Very limited savings. possible 23% Unable to cope with negative Not possible events that affect them 48% Somewhat possible 27% eventsFour-fifths in the of year smallholder prior to this farmers survey in Uganda endured financial shocks and smallholder farmers experienced multi- Sample: Smallholder farmers, n52,771 (Figure 48). A significant number of

ple financial shocks over the past year;

up with a nominal amount, set at 1/20th GNI per capita in local currency. Source: . 22 This amount was derived from the World Bank Global Findex question series, which tests whether respondents could come http://bit.ly/1QqNaHl

29 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 48. In the past 12 months, or diseases are more prevalent in the have you experienced any events? Central region; weather is a major is-

No, have sue in the Western region (Figure 52). not Across all experienced events, the ma- experienced an event jority said they did not do anything, 20% or at least did not do anything special, to cope with the challenges (Figure 53). For two of the top three events— weather/diseases or unexpected price

Yes, experienced an event somethingfluctuations about in the it, market—lesssuch as taking than a 80% temporaryone-quarter job, said borrowing they actuallymoney, using did Sample: Smallholder farmers, n52,771 some of their savings, or selling livestock or assets.

two or more of these events in the past An intermittent water supply approximately two-fifths experienced affects growth was a medical emergency, followed by a - deathyear. The in the most family frequently (Figure reported 49). event Smallholder farmers feel weather pos- households’A significant farming portion activities of smallhold are lim- es the greatest risk to their households’ iteder farmers by the inamount Uganda of finds available that theirwa- agricultural activities (Figure 50). They also experience nonweather-related supply of water, and their lack of ac- risks. In the past three years, more than cesster. Two-fifthsto water affects have their an intermittentagricultural - - ricultural activities seriously affected ter have enough for their farm, but are byfour-fifths a weather-related of smallholders event, had and their almost ag productionunable to grow (Figure their 54). agricultural Another quar ac- - fected by pests or diseases (Figure 51). supply situation. Just over one-third three-quarters of smallholders were af saidtivities they quickly have enoughbecause water of their for water their Overall the regions of Uganda face the agricultural activities or that their in- same type of events, but there are some termittent water supply does not affect regional nuances. Problems with pests the farm.

FIGURE 49. In the past 12 months, have you experienced any of these events?

Medical emergencies 69% Death of a family member 26% Income lost due to the 16% Housing repair or construcon 14% Wedding or marriage 7% Loss of wage labor 5% Loss of job 3% Relocaon 2%

Sample: Smallholder farmers, n52,771 Multiple responses allowed

30 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 50. What poses the most significant risk to your agricultural activities?

Weather-related event 68%

Pests / diseases 15%

Market prices 8%

Input prices 2%

Health 2%

Land being taken away 1%

Power failure/shortage 1%

Other 2%

Don’t know 1%

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296

FIGURE 51. Have your agricultural activities been seriously affected by any of the following events in the past three years?

Weather-related event 81%

Pests / diseases 74% Unexpected price ﬂuctuaon in the market 38% Unexpected price ﬂuctuaon of inputs 22%

Health-related event 20%

Accident or the 14% Market downturn / crops or livestock not able to be sold 6% Breakdown of equipment 5%

Contracts not honored 1%

Don’t know 2%

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296

31 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 52. Have your agricultural activities been seriously affected by any of the following events in the past three years?

Central 67% 77% 33% (n=531) Eastern 81% 81% 41% (n=641) Northern 89% 72% 35% (n=479) Western 84% 66% 40% (n=645)

Weather-related event Pests / diseases Unexpected market price ﬂuctuaon

Sample: Smallholder farmers who participate in household’s agricultural activities in each region

FIGURE 53. How did you mainly cope when this happened?

59% 55% 45%

18%20%20% 14% 8% 4% 6% 5% 7% 4% 7% 6% 3% 2% 3% 2% 2% 0% 3% 2% 3%

Temporary jobTook a loan Borrowed Sold livestockSold assetUsed savingsDid not need to Did not do do anything anything special

Weather-related event Pests / diseases Unexpected market price ﬂuctuaon (n=1,810) (n=1,759) (n=812)

Sample: Smallholder farmers who say their agricultural activities have been seriously affected by each category

FIGURE 54. Which of the following best describes your water situation?

I always have water available, and it is enough I have intermient water I always have enough water for the needs supply, but this does not available, but if I had more I have intermient water of my aﬀect my agricultural water, I would be able to supply, which does aﬀect agricultural acvies. grow my agricultural my agricultural acvies. acvies. acvies faster. 15% 23% 22% 40%

Sample: Smallholder farmers who participate in household’s agricultural activities, n55,203

32 National Survey and Segmentation of Smallholder Households in Uganda

3. AGRICULTURAL RISKS AND MITIGATION AMONG SMALLHOLDER HOUSEHOLDS IN UGANDA Smallholder farmers try to mitigate 93 percent want to be able to set aside their agricultural risks money for seed, only 40 percent are actually able to do so. The gaps for saving for Agricultural risks are a reality for smallholder farmers in Uganda. Their fertilizer and even water are much larger. life experiences have taught them to pesticides, equipment, storage, irrigation, recognize their own vulnerabilities. Although they may not be keeping money aside for their agriculture needs, lower-than-expected yield or insuf- Ugandan smallholders are storing crops. Drought, flood and disease, along with and real concerns. In concert with their crops after the harvest (Figure 57). smallholderficient crop storagefarmer areaspirations both known for TheNearly most three-quarters commonly stored currently crop is maize store themselves and their farms, there is an followed closely by beans (Figure 58), importance placed on risk mitigation matching the most commonly grown as well as an aspiration to be better crops, and those that are perceived as most important. Storage focuses almost exclusively on food or staple crops and equippedMajorities to of mitigate smallholder risk. farmers see not cash crops. The storage location the importance of keeping money aside is normally the home (Figure 59). The for certain agricultural needs, most no- main reason for storing crops is so the tably seeds and fertilizer (Figure 55). family can consume them later, further emphasizing the dependence on their aside for security, fuel and irrigation. own farms for subsistence (Figure 60). Fewer find it imperative to keep money - - tention and ability to mitigate risk holder farmers do not store their crops (FigureWhere we56) . see Most conflict smallholder is between farmers in Approximatelyafter the harvest one-quarter(Figure 61), mostly of small be- in Uganda want to be able to set money cause no crops remain. They are sold, aside for their agricultural needs. In fact, traded, or consumed by the families recognized importance of setting aside money is often more than double that farmers do not store their crops because of actual practice. For instance, while they(Figure lack 62). access Less to storage than one-quarter facilities. of

FIGURE 55. How important is it to keep money aside for the following agricultural needs?

Seeds 77% 20% 3% Ferlizer 68% 19% 12% Pescides 67% 26% 6% Equipment 58% 29% 11% For future investment opportunies 56% 32% 10% Crop storage a er harvest 54% 34% 10% Hiring staﬀ / workers 49% 29% 21% Security 36% 30% 30% Irrigaon 34% 27% 33% Fuel 32% 25% 37% Very important Somewhat important Not important

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296

33 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 56. Do you want to keep money aside for any of the following agricultural needs? vs. Do you currently keep money aside for any of the following agricultural needs?

Seeds 93% 40% Pescides 85% 26% Future investments 82% 24% Equipment 81% 23% Crop storage a er harvest 80% 22% Ferlizer 79% 19% Hiring staﬀ/workers 75% 25% Security 59% 9% Irrigaon 53% 6% Fuel 52% 12% Want to keep money aside Currently keep money aside

Sample: Smallholder farmers who participate in household’s agricultural activities

FIGURE 57. Do you currently store any FIGURE 58. Which crops do you of your crops after the harvest? normally store?

Maize 61% Beans 55% Groundnuts 29% Cassava 20% Millet 20% Sorghum 17% Simsim 9% No Yes Sweet potatoes 6% 27% 73% Soya beans 6% Coﬀee 6% Rice 4% Irish potatoes 4% Bananas 4% Pigeon pea 2% Cowpea 2% Other 8%

Sample: Smallholder farmers who participate in house- Sample: Smallholder farmers who currently store any hold’s agricultural activities, n52,296 crops after harvest, n51,633

FIGURE 59. Where do you store your crops?

77%

12% 7% 2% 2%

In the home In a shop In a grainery / In bags Somewhere barn else

Sample: Smallholder farmers who currently store any crops after harvest, n51,633

34 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 60. Why do you store your FIGURE 61. Do you currently store any crops? of your crops after the harvest?

I store it so we can consume it later 76% I am waing for the price to get beer 48% I store for another major expense 25% Yes No I store so I can pay for school 73% 27% fees 22% So my family can have extra money a er harvest season 21% Storage is good way to minimize hazards or risks 20%

Sample: Smallholder farmers who currently store any crops after harvest, n51,633 Sample: Smallholder farmers who participate in house- Multiple responses allowed hold’s agricultural activities, n52,296

Investing in livestock also helps house- households’ business. They tend to rely primarily on themselves and their families of Ugandan smallholder farmers have for labor to support all of these activities madeholds mitigatethis type risk;of investment a significant. More number than and use their family and friends as sources of information. They turn to family for an investment, and a large portion of those - saidtwo-fifths they currentlyhave ever purchasedhave any (Figure livestock 63) as. nority do not use any help to manage the 23 landhelp orfirst livestock when they(Figure need 64) it.. Of A thosesmall who mi where less than 10 percent currently prac- do use labor, it is throughout all phases Thistice this is in form stark of contrast risk mitigation to Mozambique,. of the harvest (Figure 65). Much smaller numbers of farmers use labor for selling Agriculture revolves around family crops or for the care or sale of livestock. Smallholder farmers in Uganda view their Farmers turn to their family and households’ agricultural activities as their

friends most often and frequently for FIGURE 62. Why do you not currently FIGURE 63. Have you ever purchased store any of your crops? livestock as an investment?

There are no le over crops to store 68% Yes, currently There is no available storage 21% have place nearby I need to use the money a er the 18% 35% harvest Yes, do not Storage is too expensive 8% currently have 56% It is not a good idea to store crops 2% No Other 7% 9% Don't know 2%

Sample: Smallholder farmers who currently do not store any crops after harvest, n5663 Sample: Smallholder farmers who participate in house- Multiple responses allowed hold’s agricultural activities, n52,296

http://www.cgap.org/ . 23 CGAP National Surveys & Segmentation of Smallholder Households in Mozambique, 2015-2016; publications/national-survey-segmentation-smallholder-households-mozambique

35 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 64. For managing the land and FIGURE 65. What do you use the labor livestock, what types of labor do you use? for?

Family labor 75% Planng 93%

Hired labor for extended Weeding 92% period 20% Friends or neighbors labor, Land plowing and preparaon 88% on a reciprocity basis 19% Harvesng 85% Daily rate for agricultural 11% labor Selling crops 42% Other 1% Livestock care 30%

None 15% Livestock sale 15%

Sample: Smallholder farmers who participate in house- Sample: Smallholder farmers who use labor for managing hold’s agricultural activities, n55,203 land and livestock n54,437 Multiple responses allowed Multiple responses allowed information on agricultural activities, information sources, and many farming followed closely by messages delivered households might be working from the via the radio (Table 10). All other sourc- same information passed around the village, person to person. usees are. Men used and less women frequently, have similar with some ten- denciesgetting onlywhen single it comes digits to forthe frequentsources they turn to for agricultural informa- Friends and family are also the first tion. This suggests there are limited sources for financial advice. Nearly three-quarters of smallholders go to them first, and no other source rates above TABLE 10. How often do you use each of the following sources of information for agricultural activities? More than Daily Weekly Monthly monthly Never Friends or family members 23% 20% 18% 18% 20% Radio 27% 20% 10% 13% 28% Community members 5% 13% 17% 19% 43% Religious leaders 1% 22% 10% 12% 54% Input suppliers ,1% 3% 9% 17% 66% Cell phone/SMS 10% 7% 6% 7% 68% Government officials ,1% 1% 5% 19% 70% Rural development agents/NGOs ,1% 2% 7% 15% 73% Government extension workers ,1% ,1% 4% 17% 75% Merchants ,1% 2% 6% 9% 78% Intermediaries/middlemen ,1% 1% 4% 11% 80% School teachers 1% 3% 4% 8% 80% Newspapers/magazines 1% 3% 4% 6% 84% Television 1% 2% 2% 4% 89% Internet ,1% ,1% ,1% 1% 96%

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296

36 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 66. When it comes to financial or income-related advice, who do you regularly talk to?

Friends and family 72% Savings and credit group (e.g. chama, VSLA) 10% Local Councilor or LC 7% Lead farmer 6% Chief or Village leader 5% Other community group 4% Other community leader 4% Farmers’ associaon or cooperave 4% Extension agents 2% Middle men 2% Financial instuon like a bank or microﬁnance 1% Don’t know who to go to 1% Don’t have anyone to go to 3% Don’t ask for advice 13%

Sample: Smallholder farmers, n52,771 Multiple responses allowed

Transactions are entirely cash-based 10 percent (Figure 66). Only a small por- and without contracts tion of smallholder farmers in Uganda (both men and women) turn to any groups - or associations related to farming, saving holder farmers in Uganda are purchasing or credit; the majority of smallholders are agriculturalApproximately inputs three-quarters of any kind (fertilizer, of small not members of any of these. The highest seed, etc.), and buy them largely from re- tailers. Small portions of these farmers saving and credit group (Figure 67). The buy from wholesalers and middlemen. concentrationpercentage, one-fifth, on family are and members friends ofas a Purchasing inputs from processors and cooperatives is less common (Figure 68). the very limited exposure to other outside Transactions, across all sources, tend to sources,source of suggests financial there advice, could combined be a lack with of be in cash, paid at the point of purchase information channels for bringing in new, (Figure 69). Very few farmers even have an option to pay later (Figure 70), which mechanisms into smallholder farming can be a source of strain for their bud- communitiescurrent and relevant in Uganda news. about financial geting and planning.

FIGURE 67. Are you a member of any of the following groups or associations?

Saving and credit group 20% Women’s group or associaon 7% A planng, weeding, and harvesng group 7% SACCO 4% Farm implement group 2% Cooperave/Producers’ group 2% Trade union 0% Processors group 0% An exporng group or associaon 0% Other 4% None 55%

Sample: Smallholder farmers who participate in household’s agricultural activities, n52,296 Multiple responses allowed

37 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 68. Who do you normally purchase your agricultural and livestock inputs from?

64%

24% 14% 9% 2% 1% 3%

RetailerWholesalerMiddleman / Processor Cooperave OtherDo not buy Trading company inputs

Sample: Smallholder farmers who participate in household’s agricultural activities, n55,203 Multiple responses allowed

Most smallholder farmers in Uganda sell too few customers (Figure 73). Trans- directly to the public, usually at a local portation seems to be less of an issue for market or in the village (Figure 71), and this group. that behavior might be driven more by access than by choice. The majority of In addition to capturing where farmers smallholder farmers lack the means to bring their goods to sell (Figure 71), get their crops and livestock to other the survey also asked to whom they markets (Table 11). Only approximately sell their goods. There are also close to half of smallholder farmers in Uganda believe they get the best price at the lo- (57 percent) and nearly half sell directly cation where they sell their crops and tothree the inpublic five who(45 percent)will sell to(Figure a retailer 74). livestock. Nearly all sales happen outside of a formal agreement. Only 8 percent of Adding further complexity, just over half smallholder farmers have a contract get the current market price for their to sell their crops or livestock, leaving goods (Figure 72). The most common the remainder in loose value chains reason that smallholders report that (Figure 75). Transactions are almost they do not get the current market price exclusively conducted in cash. No oth- is that they are taken advantage of by er form of payment surpasses 1 percent customers, followed closely by having (Figure 76).

FIGURE 69. How do you usually pay FIGURE 70. Do your suppliers give you your suppliers of inputs? the option to pay them later or do you have to pay immediately?

Cash 95%

Payment in-kind 1%

Prepaid debit card 0% Pay Mobile banking 0% immediately Pay later 7% Pay cash into bank 0% 93%

Electronic funds transfer 0%

Sample: Smallholder farmers who purchase main agricultural and livestock inputs, n53,983 Sample: Smallholder farmers who pay suppliers for Multiple responses allowed inputs, n53,789

38 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 71. Where do you normally sell your crops and livestock?

In village 75%

Local market 47%

At a farm to neighbor or traveling 25% merchant

Regional market 9%

Other 1%

Sample: Smallholder farmers who grow and sell crops, n54,375 Multiple responses allowed

TABLE 11. Why do you sell your crops and livestock at this location? I do not have access to transport to other markets 52% I get the best price at this market 46% Poor road conditions to other markets 24% I am not aware of prices at other markets 20% Other reason 8% Don’t know 2%

Sample: Smallholder farmers who know where crops and livestock were sold, n54,363 Multiple responses allowed

FIGURE 72. When you sell your crops FIGURE 73. Why do you not get the and livestock, do you get the current current market price? market price? My customers take advantage of me 44% Too few customers 21% No access to transport to other markets 15% Yes I have to pay high commission rates 5% 51% No to middlemen 35% Poor crop quality 3%

Don't Corrupon 3% know I do not know why 4% 14%

Sample: Smallholder farmers who do not get current Sample: Smallholder farmers who grow and sell crops, market price for crops and livestock sold, n51,556 n54,375 Multiple responses allowed

39 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 74. Who do you sell your FIGURE 75. Do you have a contract to crops and livestock to? sell any of your crops or livestock?

Retailer 57% 3% Direct to the public 45% 8% Wholesaler 23% Middleman / Trading company 18% Yes Processor 6% No Cooperave 2% Don't know Government agency 0% Other 2% 89% Sample: Smallholder farmers who grow and sell crops, n54,375 Sample: Smallholder farmers who grow and sell crops, Multiple responses allowed n54,375

FIGURE 76. How do you usually get paid for what you sell?

Cash 100%

Prepaid debit card 1%

Payment in-kind 0%

Check 0%

Mobile banking 0%

Other 0%

Sample: Smallholder farmers who grow and sell crops, n54,375 Multiple responses allowed

40 National Survey and Segmentation of Smallholder Households in Uganda

4. MOBILE PHONE TOOLS Mobile phones: A critical tool that Limited knowledge can become a valuable channel for farmers Smallholder farmers mainly see mobile phones as a channel for communi- Mobile phones are very important for cating with friends or family. Utility for smallholder farmers in Uganda, and farmers recognize them as such. A contained to small portions of the pop- phone’s importance, however, is as a ulationbusiness (Figure and financial78), which transactions demonstrates is communications tool. Farmers do not a disconnect between the perceived yet see its greater application or utility importance of mobile phones to their for their household or agricultural activ- households and agricultural activities, ities. While mobile phones are familiar, and the use of advanced phone activi- building smallholders’ association with ties. It is imperative to build knowledge activities outside of calls and texts will be and the connection between phone and critical for developing the phone as more of a digital channel for a household. and farm. household and financial transactions Higher perceived importance High current use - er farmers in Uganda recognize mo- Uganda (27 percent) have never used bileRoughly phones three-quarters as “very important” of smallhold to their aOnly mobile one-quarter phone (Figure of smallholders 79), indicating in households, or their agricultural activities a higher familiarity in the market, com- (Figure 77). The remainder (approxi- 24 for example. - This access, however, is much lower bile phones are as important, or do not thanpared the with national Mozambique, average, where only seemately their one-quarter) value to the dohousehold not believe or farm mo. 15 percent of Ugandan adults have no While importance is high overall, it still access to a mobile phone.25 However, means that a portion of the marketplace those who have used a phone primar- does not have enough context for how ily use a basic phone with no internet capability (Figure 80). Smartphone use they would benefit from a mobile device. FIGURE 77. Regardless of what you have, how important is it to your household/ agricultural activities to have a mobile phone?

79% To household 16% (n=2,870) 4% 1%

72% To agricultural acvies 21% (n=2,771) 6% 1%

Very important Somewhat important Not important Don't know

Sample: Smallholder households, n52,870; Smallholder farmers, n52,771

http://www.cgap . 24 CGAP National Surveys & Segmentation of Smallholder Households in Mozambique, 2015–2016; 25 Source: InterMedia Uganda FII Tracker survey Wave 3 (N53,000, 151), July-August 2015 .org/publications/national-survey-segmentation-smallholder-households-mozambique

41 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 78. What are the benefits to having your own mobile phone or SIM card?

99%

27% 24% 4% 3%

Talking to friends and Conducng ﬁnancial Running your business Downloading / Watching / Browsing social media family transacons Listening to music, games, (Facebook, Twier, videos, ringtones Instagram, WhatsApp)

Sample: Smallholder farmers who have ever used a mobile phone, n52,104 Multiple responses allowed

is in the single digits, with only 3 per- transactions, showing at least a poten- cent of smallholders saying they have tial for advanced activities. used one. Phone ownership lags use, but only by Limited interest for nonusers 4 percent. While 73 percent of small- Half of those without a mobile phone are holder farmers in Uganda have used a decidedly interested in obtaining such phone, only 69 percent actually have a a device. An additional 26 percent are phone in their homes (Figure 81). There somewhat interested. Pure disinterest can be multiple handsets in the house- is contained (19 percent), leaving only hold suggesting that, with exposure, 5 percent uncertain of their desire. there is recognized utility in the device. Bringing a larger share of consumers How people use their mobile phones mirrors both perceived knowledge and value-building, as there is only limit- type of handset. Those with a mobile ed,to mobileand tepid phones interest will in using require a mobile some phone typically use it to make calls or phone among nonusers. While nonusers send texts (Figure 82). Less than 15 percent of smallholder farmers are us- smallholder farmers, only half of them ing applications, taking color pictures, wouldare limited be very to interested just over one-quarterin using a mo of- browsing the internet or using social bile phone (Figures 83 and 84). - cant, portion of smallholder farmers has The main reason for not having a mo- networking sites. A small, but signifi bile phone is cost. More than two-thirds

FIGUREused their 79. mobile Have you phones ever forused financial a FIGURE 80. What type of phone have mobile phone? you used?

Basic phone 71%

No Yes Feature phone 33% 27% 73%

Smartphone 3%

Sample: Smallholder farmers who have ever used a mobile phone, n52,104 Sample: Smallholder farmers, n52,771 Multiple responses allowed

42 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 81. How many people in your household own a mobile phone?

41%

31%

21%

4% 3%

None OneTwo ThreeFour or more

Sample: Smallholder farmers, n52,870

FIGURE 82. Apart from today, when was the last time you performed the following activities on the mobile phone you use?

71% Yesterday

50% In the past 7 days 40% In the past 30 days 24% 17%18% More than 30 days ago 13%12% 11% 11% 10% 7% 7% 1% 1% Never

Made/received callsSent/received text messages or Made a ﬁnancial transacon photos

Sample: Smallholder farmers who have ever used a mobile phone, n52,104

FIGURE 83. Have you ever used a FIGURE 84. How interested would you mobile phone? be in using a mobile phone?

Very interested 50%

Somewhat interested 26% Yes 73% No 27% Not interested 19%

Don't know 5%

Sample: Smallholder farmers who have never used a Sample: Smallholder farmers, n52,771 mobile phone, n5667

43 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 12. What is the main reason you do not have a mobile phone? I don’t have money to buy phone 67% I am not allowed to use a phone by my spouse or family 5% I don’t have money to pay for airtime 4% I don’t have a need to use a phone 3% There is no place to charge a phone 1% There is no network where I live/work 1% Using a phone is against my culture/religion ,1% No specific reason 3% Other 15%

Sample: Smallholder farmers who currently do not own a phone but have used a phone, n5657

of smallholder farmers feel they do not Some smallholders lack the have the means to purchase a phone or necessary identification to buy airtime (Table 12). There is inter- open an account est in purchasing one in the future; over Many smallholders do not have the they are very likely to purchase one in one-quarter of smallholder farmers said- ing at least their aspirations, if not their (DFS)required account, national creating identification a potential— (ID) actualthe next actions, 12 months to purchase (Figure 85),a phone reflect if butto opensurmountable—barrier a digital financial to servicesaccount - finances permit. ment-issuedownership. TheID card, most which prolific only form three- of identification in all cases was a govern FIGURE 85. How likely are you to purchase a mobile phone in the next (Figure 86). This is despite the ongoing 12 months? campaign,fifths of smallholders launched in April in Uganda 2014, to have in- 26 Very likely 27% creasevoter’s uptakecard—barely of official IDsreaches in Uganda. over Somewhat likely 34% halfThe nextof smallholder likely form offarmers identification—a in Uganda Not likely 33% (Figure 86). Across all types of identi-

Don't know 6% fication tested, men and women were Sample: Smallholder farmers who currently do not own a phone but have used a phone, n5657 alsoroughly less equallyprevalent likely in rural to possess areas. them. Most of these forms of identification are

26 http://www.ugandahighcommissionpretoria.com/National-IDs.html.

44 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 86. Do you have any of the following types of an official identification?

Government-issued ID 61% Voter's card 56% Birth cerﬁcate 27% Village / Local ID 24% School-issued ID 10% Employee ID 3% Driver’s license 2% Raon card 2% Internaonal Passport 1% East African Passport 1% Military ID 1%

Sample: Smallholder farmers, n55,517 Multiple responses allowed

45 National Survey and Segmentation of Smallholder Households in Uganda

5. FINANCIAL INCLUSION AMONG SMALLHOLDER HOUSEHOLDS IN UGANDA Financial inclusion: Value of formal Ugandans are now active account holders financial institutions is appreciated (have used their registered accounts in the but not used by smallholders past 90 days) versus 30 percent in 2013.27

Uganda is building towards meaningful Smallholder farmers in Uganda lag behind the nation in their exposure - ershipfinancial in inclusion, recent years showing due toslow, increases steady - ingrowth mobile in formalmoney financial use. The account nationally own cludedto and. useTwenty-six of some percent financial of Ugandan services, representative Financial Inclusion In- smallholderwhich makes farmers them lesscurrently financially have in a sights study of Uganda shows that in 2015, almost four in 10 (39 percent) their own name, characterizing them - full-service formal, financial product in cially included, having their own full ser- the nationally representative Financial Ugandans across the country are finan Inclusionas “financially Insights included,” survey as of defined Ugandan in This is up from one-third (33 percent) in adults. Overall, smallholders in Uganda 2013vice, registered. In 2015: formal financial account. lag the national population by 12 points

■■ registered mobile money account inThe terms limited of being availability financially of banks included. in some Thirty-five percent had their own of the rural areas of Uganda factor heav- ■■ Eleven percent had their own regis- ily into smallholder farmers’ limited ac- tered bank account cess to banks. Less than one-third said ■■ Six percent had their own registered they have ever been inside of a bank (Figure 87). The majority of smallhold- institution er farmers, regardless of whether they account with a nonbank financial have been inside a bank, consider the Nationwide, active account ownership also increased. Thirty four percent of all having a bank account (Figure 88). ability to save money a key benefit of FIGURE 87. Have you ever been inside FIGURE 88. What are the benefits to a bank? having a bank account?

Ability to save money 69%

Yes Saving money in a secure locaon 42% 30% Ability to send or receive money to/from family or friends 19% Ability to send or receive payments 12%

Ability to do more business 10% No Avoid lengthy wait mes for bill 70% payments 10% Don’t know 15%

Sample: Smallholder farmers, n52,771 Sample: Smallholder farmers, n52,771 Multiple responses allowed

27 InterMedia Uganda FII Tracker surveys Wave 1 (N53,000, 151), September-December 2013; Wave 2 (N53,001, 151), June-July 2014; Wave 3 (N53,000, 151), July-August 2015.

46 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 89. Do you personally have a FIGURE 90. What is the main reason bank account that is registered in your you do not have a bank account? name? I do not have money/ I do not have enough money to make transacons 62% Yes I never thought about using a bank 10% 10% I do not know how to open one 9% There are no banks close to where I live 7% I do not know what it is 6% I do not need one, I do not make any transacons 4% Registraon fee is too high/ Fees for using a bank account are too high 1% No Banks do not oﬀer the services I need 1% 90%

Sample: Smallholder farmers who don’t have a bank Sample: Smallholder farmers, n52,771 account, n52,415

I live”), and education (“I do not know of adoption of accounts. Roughly two- how to open an account”) is a barrier for This perception will influence any kind less than 10 percent, each, of smallhold- would grant them the ability to save ers (Figure 90). Out of 62 percent who fifths appreciated that a bank account say they do not have a bank account for they can use a bank account to send or lack of money, 54 percent are female and receivein a secure money location from andfamily one-fifth or friends said. 76 percent are below poverty line. Only 15 percent of smallholders did Smallholders with bank accounts tend account. to use their accounts monthly or infre- not know of any benefits to having an Ten percent of smallholder farmers have their accounts in the seven days prior a bank account registered in their name toquently. taking Fewer part in than the twosurvey in 10(Figure had used 91). (Figure 89). An additional 2 percent use There are two primary ways that small- a bank account that belongs to someone holder farmers use bank accounts— else if they need to. Of those smallhold- either through an ATM or over the er farmers who have a bank account counter at a bank branch (Figure 92). registered in their name, roughly half When asked which was their preferred have their accounts at Centenary Bank method, just over half of smallholder farmers said they prefer using an ATM at Stanbic Bank Uganda. Full-service and another 45 percent prefer going banksand nearly can offer one-quarter a range of have services, accounts in- to a bank branch and banking over the cluding savings, money transfers, insur- counter (Figure 93). ance, investments, and even sometimes loans. Nearly all smallholder farmers have or use accounts at full-service who have ever used a full-service bank institutions. doThree-quarters not use their of accounts smallholder for business farmers purposes (Figure 94). Only small per- For those smallholders who do not have centages of Ugandan smallholders use a bank account, the main reason was lack their accounts for business, with the of means (“I do not have money”). Inter- highest percentages using them for re- estingly, awareness (“I do not know what ceiving payments from customers or it is”), access (“no banks close to where making investments.

47 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 91. Apart from today, when was the last time you made a deposit or withdrawal using a bank account or used a bank account for any other financial activity?

44%

24%

13% 12%

3% 3% 1%

YesterdayPast 7 days Past 30 days Past 90 days More than 90 NeverDon’t know days ago

Sample: Smallholder farmers who have ever used a full-service bank for any financial activity

FIGURE 92. When you use a bank FIGURE 93. Of the different ways you account for any financial activity, do use a bank for financial activities, which you use any of the following? is your preferred way?

Over the counter in a branch of 81% the bank ATM 52%

ATM 64% Over the counter in a branch of 45% the bank Over the counter at a retail store 4% Over the counter at a retail store 1% A door-to-door agent or person 1% A door-to-door agent or person associated with bank 0% associated with bank Through a mobile wallet, transfer 1% money from account to phone Mobile app 0%

Mobile app Through a mobile wallet, transfer 1% 0% money from account to phone Bank’s website 0% Bank’s website 0%

Sample: Smallholder farmers who have ever used a full- service bank for any financial activity, n5357 Sample: Smallholder farmers who have ever used a full- Multiple responses allowed service bank for any financial activity, n5357

FIGURE 94. Do you use a bank account for the following payments or purchases?

Receive payments from customers 6%

Make investments 6%

Receive payments from suppliers 5%

Pay business associated expenses 5%

Pay for agricultural inputs 4%

Pay employees 3%

Pay suppliers 1%

Do not use bank account for business transacons 74%

Sample: Smallholder farmers who have ever used a full-service bank for any financial activity, n5357

48 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 95. Have you ever heard of FIGURE 96. Are there benefits to something called mobile money? having a mobile money account?

No 6% No 13%

Yes Yes 87% 94%

Sample: Smallholder farmers who are aware of mobile Sample: Smallholder farmers, n52,771 money concept, n52,411)

Financial inclusion: Mobile money Smallholder farmers in Uganda are awareness is very high, does not aware of utilizing mobile money services convert to use - posit and withdrawal at 78 percent and Almost nine-tenths of smallholder farm- person-to-personfor basic financial money activities, transfers with deat ers say they have heard of mobile money 49 percent. Smallholders are less aware (Figure 95), and almost all smallholders of advanced uses of mobile money, with saving or storing at 34 percent and purchasing airtime at 31 percent (Figure 98). citedsee benefits most often to having included a mobile the ability money to sendaccount or (Figurereceive 96).money The to/from benefits family they Overall awareness of mobile money pro- or friends and the ability to save money viders is high for only two providers: (Figure 97). This widespread awareness and relevance does not translate named MTN unprompted and just un- derMTN half and named Airtel. Airtel Nearly (Figure three-quarters 99). The than one-third of smallholders think largest number of smallholder farm- savinginto more money specific in a productsecure location value. Lessis a ers became aware of these providers - through the radio and a lesser number count. Conducting business via a mo- from friends or family members. All of bilebenefit money of having account a mobiledoes not money seem acto the other providers were mentioned by less than 10 percent of all smallholder the lowest-scoring applications for this farmers, even those who were aware of register as a benefit, as it was one of the concept of mobile money.

FIGUREfinancial 97. mechanism. What are the benefits to having a mobile money account?

72% 63%

30% 22% 15% 11%

Ability to send or Ability to save Saving money in a Ability to send or Avoid lengthy wait Ability to do more receive money money secure locaon receive payments mes for bill business to/from family or payments friends

Sample: Smallholder farmers who believe there are benefits to having a mobile money account, n52,411 Multiple responses allowed

49 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 98. To the best of your knowledge, for what types of financial activities can you use mobile money?

78% 49% 34% 31% 15% 13% 7%

Deposit and/or Person-to-person Save or store Buy airme Make business Bill payDon’t know withdrawal money transfers money for a long- transacons term purpose

Sample: Smallholder farmers who are aware of mobile money concept, n52,272 Multiple responses allowed

Just under one-third of smallholder money for saving purposes (“I wanted farmers in Uganda have used mobile to start saving with a mobile money account”) or because of a word-of-mouth (Figure 100). Access to mobile money recommendation (“A friend or family variesmoney across in the pastthe different for a financial demographic activity groups, with the highest number of small minority of users. The main rea- users located in the Central region and sonmember for never recommended having used mobile it”) reflect money a in urban areas (Figure 101). Male smallholder farmers are more likely to have under half of nonusers said they “never used mobile money than are females. haveis a lackenough of finances money (Figureto make 103). transac Just- tions with this service.” The main reasons for starting to use mobile money is to receive money There is a slight drop off from users of from another person followed by the the mobile money service to actual ac- need to send money to another person counts; only 21 percent of smallholder (Figure 102). Starting to use mobile farmers in Uganda report having a

FIGURE 99. Please tell me the names of any mobile money providers that you are aware of?

MTN Mobile 72% Money 83% 44% Airtel Money 50% 6% M-Sente 6% 3% Orange 4% 2% Overall awareness Africell Money 3% (n=2,771) 2% Smart Money Aware of mobile money concept 2% (n=2,411) 1% Ezee Money 2% 1% M-PESA 1% 1% MCash 1% 0% Micropay 0%

By overall awareness and awareness of mobile money concept Sample: Smallholder farmers Multiple responses allowed

50 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 100. Have you ever used a FIGURE 101. Have you ever used a mobile money provider for any financial mobile money provider for any financial activity? activity?

83% 76% 71% 72% 73% 66% 52% 59% 48% 41% Yes 34% No 29% 29% 28% 27% 24% 71% 17%

Central Urban Male Eastern Western Rural Female Northern Yes No

By region, urban/rural, and gender Sample: Smallholder farmers, n52,771 Sample: Smallholder farmers, n52,771

FIGURE 102. What is the main reason you started using mobile money?

I had to receive money from another person 49% I had to send money to another person 35% I wanted to start saving money with an m-money account 4% A friend or family member recommended it 2% I had to receive money from an organizaon/government agency 2% Somebody/a person requested I opened an account 2% I wanted a safe place to store my money 1% I saw other people using it and wanted to try by myself 1% I saw posters/billboards/radio/TV adversing that convinced me 1% An agent or sales person convinced me 1% I got a promoonal amount of money to spend if I start using 0% I got a discount on airme 0% I had to send money to an organizaon/government agency 0%

Sample: Smallholder farmers, n5888

FIGURE 103. What is the main reason you have never used mobile money services?

I never have money to make transacons with this service 43%

I do not know how to open one 13%

I do not need one, I do not make any transacons 11%

I do not have a phone 10%

There is no point-of-service/agent close to where I live 5%

I do not know what it is 5%

Using such account is diﬃcult 2%

I do not have a state ID or other required documents 2%

I do not have a mobile money account 1%

It is too expensive 1%

I do not trust that my money is safe on a mobile money account 1%

Sample: Smallholder farmers, n5602

51 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 104. Do you have a registered smallholder; or simply lack of knowl- account with a mobile money provider? edge about options.

- nancial institutions could make way for Yes Seemingly, these barriers to formal fi 21% No orthe savings nonbank and financial credit cooperativechannels, such orga as- 79% nizationsmicrofinance (SACCO), institutions, to be very cooperatives, popular alternatives in rural areas.

Smallholder farmers in Uganda, how- Sample: Smallholder farmers, n52,771 ever, are not turning to other nonbank - registered account with a mobile money est use number for these types of in- provider (Figure 104). Similar to those stitutionsfinancial institutionis at 7 percent options. for The SACCOs, high who have used mobile money, the high- and account ownership is even less est number of account holders are con- (Figure 106). centrated in the Central region and in urban areas (Figure 105). The lowest are used at a higher rate than nonbank concentrations for mobile money ac- Informal financial service providers- count holders are female smallholder est numbers saying they have used a farmers and those that live in the North- VSLA—villagefinancial institutions, savings withand loan the asso high- ern region of the country. ciation (Figure 107). Over one-tenth of Financial inclusion: Smallholder smallholders have used shop keepers or farmers are not using nonbank or rotating savings and credit associations informal financial institutions (ROSCA). The smallholder farmers who use these providers do so on a some- It can be rationalized that low bank what regular basis; about half said they participation stems from distances to have used VSLAs and ROSCAs in the past banks, due to farmers located in rural shop keepers in that time (Figure 108). week, and over four-fifths visited their areas; requirements that do not suit the FIGURE 105. Do you have a registered account with a mobile money provider?

88% 83% 79% 80% 81% 74% 63% 66%

37% 34% 26% 21% 20% 19% 17% 12%

CentralUrban Male EasternWestern RuralFemaleNorthern Yes No

By region, urban/rural, and gender Sample: Smallholder farmers, n52,771

52 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 106. Have you ever used any of the following? Do you have an account/ membership in your name with any of the following?

7% SACCO 6%

Have used 3% Microﬁnance instuon 2% Have account

1% Cooperave 1%

Sample: Smallholder farmers, n52,771

FIGURE 107. Have you ever used any of the following?

VSLA - village savings and loan associaon 26%

Shop keepers 12%

ROSCA / Chama 11%

Other informal saving and credit group 5%

Money lenders 3%

Savings collectors 2% Money guard / Someone in workplace or neighborhood that collects and keeps savings deposits 1% Digital card, recharge card not aached to a bank or MFI account 0%

Sample: Smallholder farmers, n52,771

FIGURE 108. Apart from today when was the last time you used these services or service providers for any financial activity?

Past 7 days

54% 51% Past 30 days 81%

More than 30 days ago 26% 26% 10% 11% 3% 9% Stopped using altogether 8% 5% 11% VSLA - village savings and Shop keepers ROSCA / Chama loan associaon (n=275) (n=320) (n=687)

Sample: Smallholder farmers, n52,771

53 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 109. Why do you not have a membership with any of these groups?

59%

19% 14% 13% 11% 6% 5% 8%

You don’t have You don’t trust You don’t know People steal your You don’t need Groups require You have an Another reason any money them about them money any service from toomuchme in accountinabank them meengs or other formal instuon

Sample: Smallholder farmers who do not have any membership with an informal financial service provider, n51,509 Multiple responses allowed

The main reason smallholder farmers common among farmers, with approx- do not have membership with any in- -

ofimately smallholder one-third farmers and one-quarter,do not have any re halfformal of smallholder financial service farmers institution do not have is spectively, having them. Only two-fifths- anymostly money financial to have (Figure a membership; 109). More about than ning (Figure 110), though they value it. type of financial product for future plan groups. Among the 4 percent who have insur- one-fifth say they do not trust these ance, medical and life insurance are most Some traction for financial common. Measured against the full sam- planning products ple, this only constitutes approximately 1 percent of smallholder farmers. De- Use of living wills, insurance, or retire- spite this, a majority of smallholders be- ment plans is very contained, with no lieve their household needs insurance. more than 5 percent of farmers using The top three types of insurance cited as any one of these. Savings plans and in- the most needed were medical, agricul- vestments, however, are slightly more tural, and life (Figure 111).

FIGURE 110. Do you have any of the following?

A savings plan 31%

An investment 22%

A living will; I know what will happen to my money if I die unexpectedly 5%

An insurance plan 4%

A rerement plan that will help me live comfortably a er I stop working 2%

None of the above 40%

Sample: Smallholder farmers, n52,771 Multiple responses allowed

54 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 111. Which of the following types of insurance do you feel your household needs the most?

37%

21% 16% 17%

4% 3% 0% 2%

MedicalAgriculture Life House / Unemployment CarOther Don't know property / income protecon

Sample: Smallholder farmers, n52,771

Financial inclusion: High amount of an opportunity for banks and mobile trust in banks and mobile money money providers. highlights an opportunity For instance, roughly two-thirds of small- Ugandan smallholder farmers have lim- holders trust banks, bank agents, mobile money providers or their agents, at least providers, but they nevertheless have somewhat (Figure 112). This moderate- aited high exposure perceived to somerelevance financial. With service that, ly high level of trust is a solid foundation they also show relatively high trust that can build smallholders towards infor these entities, which can present clusion if providers can capitalize on it.

FIGURE 112. How much do you trust each of the following as financial sources?

Banks 42% 24% 10% 9% 7% 8%

Bank agents 27% 31% 13% 9% 8% 12%

Mobile money agents 31% 36% 12% 6% 4% 11%

Mobile money providers 35% 33% 11% 6% 5% 11%

Friends, family who borrow from / 22% 37% 16% 10% 10% 5% save money Microﬁnance instuons 20% 30% 16% 9% 9% 16%

Savings groups 28% 33% 14% 9% 8% 8%

Fully trust Somewhat trust Neither trust nor distrust Somewhat distrust Fully distrust Don't know

Sample: Smallholder farmers, n52,771

55 National Survey and Segmentation of Smallholder Households in Uganda

6. TOOLS AND FINANCIAL INCLUSION: SEGMENTATION— UGANDA’S FIVE UNIQUE SMALLHOLDER FARMING HOUSEHOLD SEGMENTS The Segmentation Technique in the sample are similar to each other Often a collection of demographic, psy- andquestions, how they examining differ from how one respondents another. chographic, behavioral and attitudinal Truly effective segmentation analyses groups within an overall population, are rooted in dimensions that lead to a moredimensions so than can any characterizesingle factor or unique vari- common, desired and shared goal for able. The CGAP National Survey and Seg- the population overall. This allows a mentation of Smallholder Households segmentation analysis to be more ger- in Uganda anticipated the complexity mane and better targeted, therefore, of smallholder households, expecting more useful for interested parties. In the case of the smallholder households the broader population.28 To that end, it in Uganda, the common, shared goal is soughtthere would to explore be unique those personas key dimensions within building strategies for bringing about that underlie different groups of smallmore useful, reliable, trusted, consum- holder households using a segmentation analysis. or informal) both connected to agricul- tureer-focused and that financial meet the services wide range (formal of Segmentation is a form of statistical other household needs. Therefore, this multivariate analysis that groups people 29 - based on their psychographics, atti- ments that correlate with greater formal tudes, expectations or behaviors with re- segmentation is rooted in defining ele spect to their own household dynamics. The groups, also referred to as clusters, financialLooking inclusion.beyond the initial analysis, that emerge from the analysis ultimately this segmentation can be repeated in allow us to deepen our understanding of follow-up or tangential studies, where - cial inclusion. Classifying smallholder householdshow various bycharacteristics key attitudinal drive and finan be- the discerningsame groups indicators within the that target define audi the- havioral characteristics provided a bet- enceunique. For segments instance, are an organization included to createbring- ter understanding of the population and use these segments to do the following: inclusion. ing a financial mechanism to market can the challenges on the path to financial ■■ Identify which segment poses the The segmentation process uncovered most potential for the organization various underlying structures that de- and its intentions. lineated groups of people. This cluster- ■■ Customize type of mechanism based groups that exist within the population upon the needs of a desired segment. sampleing technique examined looked. It did for not homogenous create these ■■ Fine-tune application and go-to mar- groups through analysis of the responses ket strategy based on market readi- givengroups. by Rather, each therespondent technique to identified various ness of the segment.

28 Personas as profiles that create reliable and realistic representations of key audience segments for reference. other attitudinal factors. 29 Psychographics refer to behaviors, interests, activities and acquisitions of a population, together with demographics and

56 National Survey and Segmentation of Smallholder Households in Uganda

■■ Optimize market positioning of the by the surveys. The model showed that - listed variables (Figure 113) correlated ment of the population. the most with the tendency to have a for- mechanism to capture a specific seg ■■ Level-set expectations for uptake and use based on the size of the mal financial account. - desired segment. - countNone ofownership the farm (mobileor land-specific money, quesbank, ■■ Track impact of mechanism with- NBFI)tions correlated with enough with relative formal financialstrength ac to in the most relevant and intended be considered part of the model. segment. - Phases of the smallholder ing that agriculture is central to small- household30 segmentation holderAt first, households this seemed. Further perplexing, exploration know suggested that the relative homogeneity Predicting corollary values of farming activities in Uganda, limited value chain and digital ecosystem, as analysis involved a machine learning al- well as the limited, contained ownership Thegorithm first called phase Random of the Forest segmentation31 that as- of formal accounts was in fact manifest- sessed the individual factors that most ing itself in the modeling. For instance, the number of crops or tendency to sell ownership (mobile money, bank, NBFI) versus consume them are not the factors (Annexcorrelate 2) with. The formalsix most financial predictable account and - - cial account. In an ecosystem where pay- that drive smallholders to have a finan count ownership are as follows: ments were digital, or loans were more discerning measures of financial ac formal, you might see some more direct 1. Educational attainment of the head correlations. Here, correlations manifest of household. themselves through socioeconomic elements including education, PPI, access 2. Socio-economic status or PPI of head to funds, phone ownership, and other of household. experiences or attitudes. 3. Access to emergency funds. Forming Segments 4. Mobile phone ownership. The second phase of the segmentation 5. Attitude toward the future. analysis was to explore the degree to 6. Encountering unexpected life and which these factors together explained farming events. the variation within the population, and formed meaningful cleavages with- These measures emerged as the most in it, carving out distinct personas. discerning after extensive tests and mod- Individually, these measures are the eling which took into consideration more - than 30 demographic, psychographic and sion, and useful in helping determine farmographic (size of land, type of crops, thestrongest likelihood predictors of becoming of financial part of inclu the value chains, inputs used, cash crops, consumption crops, etc.) variables collected segmentation model, these factors cause financial fold. Compiled together in a

30 The segmentation analysis is based off of a three-part survey, which gathered information from all aspects of the smallholder farmer—the household, all household members who contribute to the income of the household, and a randomly selected household member. The term “smallholder household” is used throughout this report to refer to the sampled population, which draws information from the head of household or a randomly selected household member. 31 See Annex 2 and http://www.statsoft.com/Textbook/Random-Forest for documentation on the Random Forest Algorithm.

57 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 113. Uganda Smallholder Farmers Educaon Socio-Economic Status (PPI) Access to emergency funds $2.50+, 33% 23% Very possible 69% 10% up to secondary 27% Somewhat primary or more Below 48% Not possible $2.50, 67% 2% Unsure Mobile Phone Atude: The Future Will Experienced an Ownership Take Care of Itself unexpected event

Agree: None 69% 27% 20% have at least one mobile Disagree: 68% At least one phone in 80% their household

(Sample: All Smallholder Farmers) meaningful divisions that enable greater smallholder farming populations out- understanding of the population and can side of Uganda, though the description facilitate targeted strategies for moving and the incidence of each reported here- the group to the end goal.

Using the most predictive variables in is unique to Uganda. 32 exer- clusters or segments are as follows: cise, the clustering analysis produced By segmentation variables only, the five identified in the Random Forest 1. Farming for sustenance: The “farming for sustenance” segment households:a five-segment solution, delineating - five unique segments of smallholder dan farming household. The seg- 1. “Farming for sustenance” mentrepresents indexes the low quintessential on the Progress Ugan out of Poverty Index, has been farm- 2. “Battling the elements” ing for many years, and is the highest in wanting their children to continue farming. This segment has the low- 3.4. “Diversified“Options for andgrowth” pragmatic” truly does live off of what the farm 5. “Strategic agricultural entrepreneur- produces,est household either income consuming, of all five, selling and ship” or trading the fruits of their agricultural labor. This is a highly vulner- Since the sample was randomly select- able group, and perhaps stands to ed and represents the population of - smallholder farmers and households ricultural mechanisms that can opti- across Uganda, we can reasonably as- mizegain thetheir most daily from labor financial. and ag groups in the population as a whole. We 2. Battling the elements: The “bat- alsosert theexpect five thatsegments similar represent groups existnatural in tling the elements” segment is also a

32 See Annex 2 and http://www.statsoft.com/Textbook/Random-Forest for documentation on the Random Forest Algorithm.

58 National Survey and Segmentation of Smallholder Households in Uganda

AQ2 FIGURE 114. Uganda Smallholder Household Segments

Strategic Farming for Baling the Diversiﬁed and Opons for agricultural sustenance elements pragmac growth entrepreneurship

Persevered The realies would Every day, Stable, opmisc Acvely engaged in through harsh lead them to quintessenal for the future and agriculture, and agricultural diversify out of farming household has opons for the growing their challenges and something they that struggles to future within and agricultural remain opmisc enjoy if given the support its needs outside of farming acvies toward farming choice

Stands to gain the Financial Embody the Could pivot into or Model or “use most from mechanisms have realism and inner out of farming case” for inspiring ﬁnancial and enabled some of conﬂict that can depending on growth in other agricultural their perseverance characterize opportunies segments mechanisms households

vulnerable group, but as a group, does earn more, have more income not face the income limitations of streams and have more connectivi- those “farming for sustenance.” A greater portion of this group, com- ways, they have an aspirational pro- pared with the “farming for sus- ty to financial mechanisms. In some tenance” group, generates income “battling the elements.” They have from agriculture, and a greater por- sufferedfile like “farming unexpected for sustenance” life events toand a tion of these households has multi- similar level as other segments, and ple income sources. This segment is have had resources to overcome still challenged by somewhat lim- what they do experience. The con- ited education and the incidence of unexpected life or farm-related despite enjoying farming, taking events, such as weather challenges prideflict that in it, arises and looking in this forgroup opportuni is that,- or illnesses in the family. Experience ties to grow their farms, many would with negative farm events (e.g., pests diversify out of agriculture if given and diseases) is the highest for this the opportunity, potentially out of group. Challenges have not damp- frustration with the realities of farm- ened their future aspirations or dis- ing. They are thoughtful about what suaded them from working hard. This they do, but can also be impulsive, group has persevered through those looking for opportunities to improve - their situation. This is an important cial tools, making them a group that group as they represent smallholder mightchallenges, better sometimes understand using the value finan of having some form of a safety net. The within and outside of agriculture to biggest difference between this group besthouseholds sustain their that household have diversified needs. versus the “farming for sustenance” 4. Options for growth: The “options limited education and experiences for growth” group earns a higher in- moresegment negative is the agriculturalfirst group has events a more. come, has more resources for when the unexpected occurs, and is even 3. Diversified and pragmatic: The optimistic about their future. Their - future, though, could take them in one of two directions, within or “diversified and pragmatic” seg- outside of agriculture. This diversi- holderment reflects farming the households realism and. Theseinner householdsconflict that growcan characterize more, sell smallmore, a household has options to take care fication appears deliberate, so that

59 National Survey and Segmentation of Smallholder Households in Uganda

of itself if their agricultural yield be- resources to get them through tough times. What characterizes them more the segment relies heavily on agricul- turalcomes income, too difficult it is also to maintain. the most While likely The smallholders in this segment put to be engaged in more stable income muchdefinitively, thought though, into what is their they mindset. do and sources outside of agriculture. They has big aspirations that include a fu- - ture in agriculture. Farming is what ing, continuity in agriculture, and satisfactionare equally as with passionate farming about as they farm are and where their legacies live. They are embracing of opportunities outside notthey as want likely to to do, want what out, satisfies or be willing them, of the agricultural sector. The young- to take work outside of agriculture. est of all groups, this segment could This is a group that can be a model pivot in either direction depending or a use-case for carrying meaningful in part on how they are cultivated by messages or examples for growth to policymakers, development organi- other segments of the population. - ization of these segments when we ex- 5. Strategiczations and agriculturalfinancial institutions. entrepre- ploreThere more is greater deeply definition how they and behave, character what neurship: The “strategic agricultural they believe, and where their interests lie. entrepreneurship” segment includes households that appear to be active- - ly engaged in building their agricul- iorally characterize smallholder house- tural work with some indications holdsAs a whole, across these Uganda five. Thesegments “farming behav for of success or at least progress. The sustenance” group is the most predom- group is more enabled than others, inant in the country, comprising 54 per- having higher income, greater access cent of farming households. They differ - slightly from the next largest group, cial mechanisms at their disposal. “battling the elements,” comprising toThis emergency group of funds,smallholders and more has finan been 21 percent (Figure 115). The remain- impacted by the realities and challenges of farming, and have been able and pragmatic,” “strategic agricultur- to rely upon their savings or other aling entrepreneurship,” 24 percent include and the the “diversified “options

FIGURE 115. Uganda Smallholder Household Segments

Diversiﬁed and pragmac, Opons for growth, 4% 16%

Baling the Strategic agricultural elements, entrepreneurship, 21% 4%

Farming for sustenance, 54%

(Shown: All smallholder farmers)

60 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 13. Uganda Smallholder Household Segments by Clustering Criteria Farming Battling Diversified Options Strategic for the and for agricultural Segment5 sustenance elements pragmatic growth entrepreneurship n5 1,324 574 144 569 160 Educational attainment of household head Never attended 25% 15% 19% 5% 12% school Preprimary 52% 57% 63% 10% 39% Primary 17% 24% 14% 43% 39% Secondary 5% 4% 3% 33% 8% Higher education 1% 0% 1% 8% 1% Socioeconomic status Above the 14% 20% 100% 62% 100% poverty line Below the 86% 80% 0% 38% 0% poverty line Access to emergency funds: Can come up with 114,000 shillings within the next month Very possible 12% 29% 0% 42% 73% Somewhat 22% 32% 60% 31% 17% possible Not possible 63% 37% 38% 26% 9% Don’t know 3% 2% 1% 1% 0% Mobile phone ownership No 50% 1% 2% 1% 0% Yes 50% 99% 98% 99% 100% Attitude: The future will take care of itself Agree 32% 23% 17% 21% 16% Disagree 62% 72% 73% 78% 79% Don’t know 6% 4% 10% 1% 5% In the past 12 months, experienced any unexpected events (including, but not limited to death, illness, accidents, etc.). No, I didn’t 22% 15% 14% 19% 15% Yes, I did 78% 85% 86% 81% 85%

(Shown: All smallholder farmers) for growth” segments, who represent a ownership, attitude toward the future small, but important, component of the and experience with unexpected events. farming population. - Table 13 shows each segment and how it ics of each segment, bringing character - andThe depth profiles to beloweach of detail them the.33 Perhaps dynam ables: education, socioeconomic status, the best illustration of the differences accessfares on to each emergency of the cluster-defining funds, mobile phone vari in the segments, however, is the linear

33 Caution: Small segment sizes for the “diversified and pragmatic” and the “strategic agricultural entrepreneurship” groups can limit analysis. Proceed with caution in extrapolating findings. 61 National Survey and Segmentation of Smallholder Households in Uganda

Characterizing attributes: the “farming for sustenance” (and larg- comparisons to other segments estprogression group) is the of themost five impoverished groups, where and in need, and the “options for growth” are The “farming for sustenance” group tends at the far other end showing models of to be older (only 36 percent are under progress within the population. 39 years old) but is as tenured in farming as most other segments (58 percent, 101 Segment 1: “Farming for with their farming achievements (58 per- sustenance”: Dependent on the years in farming). They are less satisfied farm for day-to-day survival in agriculture (91 percent), and intend The “farming for sustenance” segment tocent keep “satisfied”), working in yet it (88 still percent) enjoy working. Near- is the predominant segment among small- ly half (47 percent) say they would not holder households in Uganda. Comprising want to do any other kind of work. 54 percent of farming households, it is a sizable segment, containing the majority, Demographics: Nearly all smallholder households in this households in the country. segment live in poverty, are and, therefore, the “status quo” for farming concentrated in the Northern and Segmentation analysis typically clus- Eastern regions of the country, and ters a population in smaller groups be- largely headed by older farmers. neath the 50 percent mark. In the case of Uganda, there is a preponderance of vul- Relative to other segments, the “farming nerability due to a host of factors that infor sustenance” segment skews older, clude education, resources and income that cannot be ignored, hidden, or even groups. Only 14 percent of the “farming further segmented into anything mean- forwith sustenance” an unequal segmentdistribution is between across age 15 - and 29, leaving the balance spread across cial inclusion. That alone tells us that 30 and over. Households are concentrat- acknowledgingingful when it comes the income to predicting vulnerability finan ed in the Eastern and Northern regions with the greatest concentration in the East at just over one-third (36 percent), mechanismsand other unique to scale aspects in ofthe this country popu. followed by 30 percent in the North. Makinglation will it beeven critical more to critical bringing is financial the fact that no other single or group of segments Nearly all, 87 percent, live under the come close to the size of this segment. poverty line.

SEGMENT SYNOPSIS

Representing the quintessential Ugandan farming household, the “farming for sustenance” segment indexes very low on the Progress out of Poverty Index, shows a high number of years in farming, and generally wants their children to continue farming.

This segment truly does live off of what the farm produces, either consuming, selling or trading the fruits of their agricultural labor, without much else to sustain their households.

This is a highly vulnerable group, and perhaps stands to gain the most from financial and agricultural mechanisms that can facilitate their daily labor.

62 National Survey and Segmentation of Smallholder Households in Uganda

AQ2 FIGURE 116. Uganda Smallholder Household Segment Mindset

4% Uganda 54% 21% 4% 16%

Strategic Mindsets Farming for Baling the Diversiﬁed Opons for agricultural (% agree) sustenance elements and pragmac growth entrepreneurship Work hard to be 90%95% 96% 99% 98% the best Look for opportuniesto 85% 88% 87%88% 88% improve situaon Have many 76% 81% 87%85% 87% aspiraons Do things a er 86% 90% 92%91% 90% much thought Impulsive 27% 26% 37%28% 24% Life determined by powerful 42%43% 36% 34% 37% people

(Shown: All smallholder farmers)

Farming: Experience, income and are newer to agriculture, working in it crops

“Farming for sustenance” households underThese fivehouseholds years. mostly intend to con- are tenured in their craft. Fifty-eight per- tinue working in agriculture (88 percent). cent have been working in agriculture They generally enjoy it (91 percent), and for more than 10 years. An additional many would like to expand their capa- 16 percent have been part of agricul- bilities (85 percent). That said, it is fair ture for six to 10 years. Only 25 percent to point out that full-time employment

AQ2 FIGURE 117. Smallholder Farmers in Uganda Financial Inclusion* by Segment

67% 63%

46% Total: 26% 36%

Farming for Baling the Diversiﬁed and Opons for growth Strategic agricultural sustenance elements pragmac entrepreneurship

Uganda: Financially included

(Shown: All smallholder farmers) * Financial Inclusion defined as having a full-service bank, mobile money or nonbank financial institution account with access in one’s own name.

63 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 118. View of success in agriculture vs. willingness to continue working in it

Farming for sustenance 58% 88%

Baling the elements 59% 86%

Diversiﬁed & pragmac 57% 82%

Opons for growth 59% 87%

Strategic ag entrepreneurship 65% 96%

Sample: All smallholder households who participate in agricultural activities by segment could also be attractive to smallholder three hectares, and the rest have over households in this segment (66 percent). three hectares of land.34 That fewer than six in 10 (58 percent) On average, the smallholder households with what their agricultural work has in the “farming for sustenance” segment achievedfarmers in(Figure this segment 118) suggests are satisfied they are growing seven crops (6.93 precisely) are critical of themselves, and perhaps each year on their land. They tend to sell wanted better outcomes than their cir- on average four of the crops they grow cumstances could support. (3.57 precisely).

Ninety-one percent of “farming for sus- Vulnerable to outside elements tenance” households generate income from agriculture and, overall, these Their vulnerability becomes even more households tend to have fewer sourc- apparent when comparing the percent- es of income (30 percent, one source of age generating income from agriculture income; 34 percent, two sources of in- against the percentage whose agricultural events have been seriously affected generates income from occasional jobs, by an outside element. Nearly all small- andcome). 9 percentClose to one-fifthgenerates of income the segment from holders in this group (98 percent) were - impacted by weather, pests and disease, cent receive salary and wages from a - regularretail or job manufacturing.. Only five per ment failure and/or their own health issuesaccidents,. Seventy-four market percent fluctuations, were affected equip Collective reporting from all household by more than one issue and 85 percent members active in agriculture shows were affected by weather, while 73 per- that two-thirds of the households in cent faced pests and disease. “farming for sustenance” have two hectares of land or less. Up to an additional Among those who were seriously af- 13 percent have between two and fected by any of the above events, over

34 The land size measurement comes from the household survey where multiple members of the agricultural household offer up their recollection of various dynamics so as to capture full dynamics instead of relying upon just one member’s knowledge of the household. An aggregate estimate of this measure was then created and appended to the segmentation, which is

These data are weighted accordingly. Use data with caution surrounding extrapolation and inferences. These should be usedbased only on participantas added descriptive responses measures to the individual. questionnaire (asked of just one randomly selected household member).

64 National Survey and Segmentation of Smallholder Households in Uganda

money account. Overall, 11 percent can farming households did nothing. For access mobile money either through thosethree-fifths who did (63 take percent) action when of affected coping their own or through someone else’s with unexpected events: account. Seventy-six percent of “farming for sustenance” had heard of mobile ■■ Only 3 percent coped by taking a money prior to the survey, compared to temporary job at least 90 percent awareness among other segments. ■■ 5 percent sold an asset Three percent of “farming for suste- ■■ 2 percent took a loan and 10 percent nance” smallholders have an NBFI ac- borrowed count but just 4 percent have used an ■■ 11 percent sold livestock or crops NBFI. Three percent of the smallholder households in the “farming for suste- ■■ 14 percent used savings nance” segment reported having used a bank and just 2 percent have an account. Financial attitudes Fifteen percent had ever been inside a bank. The segmentation model is built on -

full-servicepredictors ofbank, financial mobile inclusion, money or which NBFI (42While percent) more prevalentof the “farming than formal for sus fi- accountis defined in their here name as .those It follows, having then, a tenance”nancial services,segment stillhas access only two-fifthsto an in- that ordering segments from more vulnerable “farming for sustenance” and savings or loan association or ROSCA, “battling the elements” groups to “stra- moneylenderformal financial or mechanismmoney guard such. Twen as a- tegic agricultural entrepreneurship” ty-two percent of those in this segment shows a somewhat linear relationship have used a village savings and loan association and 10 percent have used a ROSCA. Three percent have used a Extremelywith financial limited inclusion. access to financial moneylender, and 1 percent have used services a money guard or someone who collects savings deposits. Overall, 26 percent of Ugandan small- - Lower perceived importance of cluded, meaning they have a full-service financial practices versus other bank,holder mobile households money, are or financiallyNBFI account in groups in their name. The “farming for sustenance” segment comes in lowest, with - - nisms, there is also a gap in perceived ed. Since this segment is the largest, Separate from having financial mecha encapsulatingonly 6 percent 54being percent financially of the includsmall- as savings. Large majorities of this seg- holder households in the country, it is, mentimportance feel it isof very financial important behaviors, to save such for therefore, a driving force behind the future purchases (82 percent) or school fees (74 percent), and more than two- smallholder farmers in Uganda. overall financial inclusion number for unexpected (69 percent) or regular pur- Mobile money accounts are the most chasesthirds find(64 percent) it important. to save for the the “farming for sustenance” segment. There also is a strong sense of im- Thatpopular said, formal only 3financial percent servicehave a mobileamong portance of investing in the farm

65 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 119. View saving money through different mediums as “very important”

Financial instuon 57%

On a mobile phone 44%

With an informal group 48%

At home 47%

Sample: “Farming for sustenance” households, n51,324

(80 percent very important), though we expect to see when leveraging a this is muted relative to other seg- - ments. Members of this segment do ant group because these households are not perceive as much importance in facingfive-segment many of solution. the same This limiting is an importcircum- saving through financial mechanisms, stances as those in the “farming for sus- formal or informal, or even in the home tenance” group, but are more able to call (Figure 119). - Segment 2: “Battling the elements”: cialon their steps support in their networks lives, despite for financial having Challenged, with limited resources lessassistance. education They and are facing taking the better brutal finan re- but persevere alities of farming.

The “battling the elements” group is Characterizing attributes: the second largest segment in Uganda, comparisons to other segments comprising 21 percent of smallholder farming households. The size of this Unlike the “farming for sustenance” seg- segment is more typical based on what ment, this group tends to be younger

SEGMENT SYNOPSIS

The “battling the elements” segment is also a vulnerable group, but as a group, does not face the limitations of a support network like the “farming for sustenance” segment. A greater portion generate income from agriculture, and a greater portion of these households has multiple income sources. This segment is still challenged by limited education, and the incidence of unexpected life or farm-related events.

Experience with unexpected life events is somewhat greater for this group than the others. Challenges have not dampened their future aspirations or dissuaded them from working hard. This group has persevered through those challenges, sometimes using financial tools, which might make them the group that better understands the value of having some form of a safety net.

The biggest differences between this group and the “farming for sustenance” group, is that the “battling the elements” group experiences more unexpected life events but can call on a stronger financial support network.

66 National Survey and Segmentation of Smallholder Households in Uganda

(18 percent, ,30) and more tenured in Farming: Experience, income and farming (66 percent, 101 years). Many crops achievements (59 percent), and generally “Battling the elements” households are enjoyare satisfied (88 percent) with and their intend agricultural to contin- the most tenured in their craft. Two- ue working in agriculture (86 percent). thirds of those in this segment have There is some appeal to work outside of been working in agriculture for more agriculture. Only 43 percent would not than 10 years. An additional 18 percent want to do any other kind of work, and have been part of agriculture for six to 31 percent would not take another job 10 years. Only 15 percent are newer to - ment lives under the poverty line. This years. isif offered.a smaller Eighty-five portion than percent in the of “farmingthis seg agriculture, working in it five or fewer for sustenance” segment, but substan- Enjoyment of farming tial nonetheless. These households mostly intend to continue working in agriculture (86 per- Demographics: A majority of cent). They generally enjoy it (88 per- smallholder households in this cent), and many would like to expand segment live in poverty, are their capabilities (87 percent). Close concentrated in the eastern and western regions of the country, farming achievements (59 percent) and represent a younger group of (Figureto three-fifths 120). That are said, satisfied it is worth with theirnot- farmers. ing that full-time employment also could be attractive to some households (65 Versus other segments, the “battling the percent). elements” smallholders are one of the youngest. Roughly half (48 percent) are Over nine-tenths (94 percent) of “battling the elements” households generate in- (38 percent) live in the Western region come from agriculture, and overall these under the age of 40. Close to two-fifths households tend to have more sources (30 percent) live in the Eastern region. of income (24 percent, one source of Nineteenof the country, percent and come a nearly from equalthe North part. income; 35 percent, two sources of in- The vast majority, 85 percent, live under come; 29 percent, three sources of in- the poverty line. come; and 11 percent, between four and

FIGURE 120. View of success in agriculture vs. willingness to continue working in it

Baling the elements 59% 86%

Diversiﬁed & pragmac 57% 82%

Opons for growth 59% 87%

Strategic ag entrepreneurship 65% 96%

Farming for sustenance 58% 88%

Sample: All smallholder households who participate in agricultural activities by segment

67 National Survey and Segmentation of Smallholder Households in Uganda

seven sources). This group, however, Among those that were seriously affected skews slightly higher in income sources by any of the above events, roughly three- than the “farming for sustenance” group. As with the “farming for sustenance” did nothing. When facing unexpected group, smaller percentages earn wag- events:fifths (59 percent) of affected households es from occasional jobs (21 percent), ■■ 11 percent sold livestock or crops a business in retail or manufacturing. ■■ 16 percent used savings Onlyand close 7 percent to one-fifth get wages (17 or percent) salary from run a regular job. ■■ 4 percent coped by taking a temporary job Roughly 66 percent of the smallholder farming households in “battling the ■■ elements” have two hectares of land service provider and 9 percent bor- or less. Up to an additional 17 percent rowed6 percent from took others a loan from a financial have between two and three hectares, and the rest have over three hectares of ■■ 5 percent sold an asset land.35 Financial attitudes On average, the “battling the elements” households are growing seven crops Compared with all of the other groups, (7.11 precisely) each year on their land. the “battling the elements” segment They tend to sell, on average, four crops comes in with the second lowest per- they grow (3.61 precisely). - cluded, meaning they have a full-service, centage of those who are financially in Vulnerable to weather bank, mobile money, or NBFI account in - Their vulnerability becomes even more cially included (versus 26 percent of apparent when comparing the percent- Ugandatheir name. smallholder Only 36 percenthouseholds are finanover- age generating income from agriculture all). This group is six times as likely to against the percentages whose agricultural events have been seriously affected place compared with the “farming for by an outside element (including weather, sustenance”have formal group financial. mechanisms in pests, illness, loss, accidents). Overall, nearly all of this group (97 percent) were Some formal financial accounts impacted by weather, pests and disease, - Here too, mobile money accounts are ment failure and/or their own health issues,accidents, and 75 market percent fluctuations, faced two or equip more mechanisms than NBFIs or banks. Twenty-ninemore prevalent percent among of smallholders the financial in were affected by weather. Nearly as many this group have their own accounts, and (76issues. percent) Over three-quarters faced pests (78and percent)disease 93 percent of the segment had heard issues. of mobile money prior to the survey.

35 The land size measurement comes from the household survey where multiple members of the agricultural household offer up their recollection of various dynamics so as to capture full dynamics instead of relying on just one member’s knowledge of the household. An aggregate estimate of this measure was then created and appended to the segmentation, which is

These data are weighted accordingly. Use data with caution surrounding extrapolation and inferences. These should be basedused only on participantas added descriptive responses measures to the individual. questionnaire (asked of just one randomly selected household member).

68 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 121. View saving money through different mediums as very important

Financial instuon 63%

On a mobile phone 55%

With an informal group 49%

At home 49%

Sample: “Battling the elements” households, n5574

Nine percent of these individuals have “farming for sustenance” group, but hav- an account with an NBFI, and 9 percent ing savings within the home is still im- also have a bank account. Just over one- portant to half of this group (Figure 121). third have been inside a bank. Segment 3: “Diversified and Pragmatic”: Realistic, grounded and surpasses that of formal accounts, with plan accordingly for the realities of 54Access percent to informalhaving accessed financial some accounts infor- agricultural life36 loan associations are the most common The “diversified and pragmatic” seg- informalmal financial service service. (30 percent Village savingshave used), and ment, which includes just 4 percent of followed by a ROSCA (19 percent have Uganda’s smallholder households, is used). Only 2 percent have used money moving away from vulnerability and guards or someone who collects savings onto a path of stability. Perhaps what is deposits. most important about this group is its relatively small size, suggesting a lack - of use cases and models in the market- nisms, we also see a gap in perceived im- place for coming out of vulnerability. Its Separate from having financial mecha size is also important in level-setting “battling the elements” group. Like the “farmingportance for of financialsustenance” behaviors segment, for just the agricultural mechanisms meant for aexpectations less-entrenched as to whathousehold financial might and the “battling the elements” segment feel achieve. In time, this group can grow to itover is veryfour-fifths important (82 percent) to save offor those future in be more substantial. purchases, or school fees (84 percent), Characterizing attributes: the unexpected (73 percent) or regular comparisons to other segments purchasesand fewer (67find percent)it important. Many to (82save per for- Perhaps the most distinguishing ten- money in the farm. cent) find it very important to invest group is how so many of them look for There is a greater perceived importance opportunitiesdency of the “diversified to improve and their pragmatic” situations (87 percent) and their relatively both formal and informal, versus the negative outlook on the agricultural in saving through financial mechanisms,

36 Caution: Small segment size limits analysis. Proceed with caution in extrapolating findings

69 National Survey and Segmentation of Smallholder Households in Uganda

SEGMENT SYNOPSIS

The “diversified and pragmatic” segment reflects the realism and inner conflict that can characterize smallholder farming households. These households grow more, sell more, earn more, have more income streams and have a broader port- folio of financial mechanisms.

They are the most likely to have experienced unexpected events in the past 12 months, but least likely to have had their agricultural activities affected by multiple negative events (e.g., pests, weather) in the past 3 years. They do not believe it is very possible to acquire emergency funds and are the most likely to take a full-time job.

The conflict that arises in this group is that despite enjoying farming, taking pride in it, and looking for opportunities to grow it, many would diversify out of agriculture if given the opportunity.

They are empowered, but believe that someone else or circumstances, more generally, might have more power than they do to shape their destiny. They think through decisions, but also know that reality can get in the way of the best-laid plans. This is an important group, as it represents smallholder households that have diversified within and outside of agriculture to best sustain their household needs.

sector (82 percent want to stay in agri- live in the central and western regions culture, 50 percent want other kinds of of the country, leaving close to one- work and just 82 percent enjoy working in agriculture). 5 percent in the Northern regions. This entirefifth (18 group percent) (100 percent) in the Eastern, lives above and They also work hard to be the best the poverty line. (96 percent), and can be impulsive (37 percent), suggesting they tend to be Farming: Experience, income, and risk-takers. Despite their predisposition crops to being risk-takers and opportunity seekers, a sizable portion of those in this group do feel their lives are determined are mostly experienced in farming. More by other, powerful people (36 percent), than“Diversified half have and been pragmatic” farming households for 10 or more years (53 percent), and an addi- from circumstances beyond their control. tional 19 percent have been in it for six perhaps reflecting they have emerged to 10 years. Versus the “battling the el- Demographics: All households ements” group, we see fewer people in live above the poverty line, are that tenured, more than 10 years, cat- concentrated in the Central and egory. Twenty-six percent have been Western regions of the country, and are largely headed by older farmers. farmingThese households for five or mostlyfewer years. intend to continue working in agriculture (82 per- tends to be slightly older (56 percent, cent), showing similar intentions as 40The years “diversified1) than and the pragmatic” prior groups group. other segments. They generally enjoy Just over three-fourths (77 percent) it (82 percent), and many would like

70 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 122. View of success in agriculture vs. willingness to continue working in it

Diversiﬁed & pragmac 57% 82%

Opons for growth 59% 87%

Strategic ag entrepreneurship 65% 96%

Farming for sustenance 58% 88%

Baling the elements 59% 86%

Sample: All smallholder households who participate in agricultural activities by segment to expand their capabilities (85 per- ■■ Wages from occasional jobs (22 percent). That said, it is fair to point out cent) that full-time employment could also be ■■ Running a business that provides attractive to these smallholder house- services (18 percent) pragmatism. Farming is hard work, and For all household members active in ag- susceptibleholds (67 to percent), elements reflecting that make their it riculture, 62 percent of the households business. This segment is the least sat- hectares of land or less. Up to an addi- unpredictable, or difficult to grow the tionalin “diversified 23 percent and have pragmatic” between have two twoand has achieved, at just under six in 10(57 three hectares, and the rest have over percent),isfied with suggesting what their they agricultural are critical work of three hectares of land.37 themselves, and perhaps wanted better outcomes that their circumstances - could not support (Figure 122). matic” are growing six crops (6.18 precisely)On average, each theyear “diversified on their land and. pragThey tend to sell, on average, three crops they More sources of income grow (3.12 precisely). Nearly nine in 10 (89 percent) of “di- Affected by outside elements generate income from agriculture, and, This group was most likely to have overall,versified these and households pragmatic” tend households to be experienced unexpected externalities the least diverse in growing and selling (98 percent), but least likely to have crops (85 percent) and rearing live- had their agricultural activities affect- stock (43 percent). Other sources of in- pests) in the prior three years. Just over households can include: ed by two or more issues (e.g., floods, come for the “diversified and pragmatic” two-thirds of this segment (69 per- ■■ Money from family and friends cent) was affected by weather alone. (28 percent) Almost three-fourths (76 percent) of

37 The land size measurement comes from the household survey where multiple members of the agricultural household offer up their recollection of various dynamics so as to capture full dynamics instead of relying on just one member’s knowledge of the household. An aggregate estimate of this measure was then created and appended to the segmentation, which is

71 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 123. View saving money through different mediums as very important

Financial instuon 58%

On a mobile phone 51%

With an informal group 47%

At home 54%

Sample: “Diverse and pragmatic” households, n5144

these smallholders faced pests and dis- segment. Thirty-nine percent have a ease issues. The majority (58 percent) mobile money account, and 41 percent have used mobile money either through with these events and did nothing in their own or through someone else’s particulardid not have when a specific they occurred response. to cope and pragmatic” smallholder households Financial attitudes wereaccount. aware Ninety of at least percent one ofmobile “diversified money provider prior to the survey. The segmentation model itself is built Nine percent of smallholders in this segment have NBFI accounts, and 12 per- bank,off of mobile predictors money of financialor NBFI account inclusion, in cent have bank accounts. Close to four theirdefined name as . Close those to having half (46 a full-servicepercent) of in 10 (38 percent) had ever been in a bank. doublethe “diversified that of the and general pragmatic” population segment of - farmersis financially. included, which is close to informalThe “diversified services and does pragmatic” not surpass small that While a small portion of the popula- ofholders formal are accounts, the first withfor which nearly access four toin tion, this segment is the one that of- 10 (38 percent) having access to some fers hope that farming households can - put their livelihoods on a path toward ings and loan associations (VSLAs) are greater stability that includes mech- informal financial service. Village sav- anisms for household management. nancial services, used by 26 percent of With close to half of this segment not thisthe mostsegment common. Sixteen form percent of informal had used fi financially included, it suggests that ROSCAs or Chamas, and 15 percent had targeted efforts might be well placed, used a shopkeeper. and the collective reach of those efforts might extend far beyond the peo- - ple within the segment. cial mechanisms, there is less acknowl- Separate from having access to finan A “pragmatic” approach to finances behaviors, such as saving. This seg- mentedgment is one of the of theimportance least likely of financialto view Mobile money accounts are the most saving through any mechanism—other than cash at home—as very important. Though a majority believes saving popular formal financial mechanisms among the “diversified and pragmatic” 72 National Survey and Segmentation of Smallholder Households in Uganda

That they are optimistic does not regular expenses, the future) is very mean there is no room for growth. important,for specific they purposes are least (e.g., likely school to value fees, Their optimism conveys they have im- saving for regular expenses or investing proved their current situations, largely in future educational opportunities. because of their net incomes. Further- more, it might not be farming income Similar to all other groups, those in the alone that helps stabilize their households, as agriculture is part of the household’s diverse revenue streams. the“diversified best, or, on and par pragmatic” with informal segment sav- ingsconsider options saving. However, at financial it is in institutions the “diver- Characterizing attributes: comparisons to other segments greatest emphasis is on saving at home (Figuresified and 123) pragmatic”. segment that the “Options for growth” households tend to be newer to farming and are looking Segment 4: “Options for Growth”: to expand their agricultural activities. Stable, optimistic and building They already look to farming as a legacy, various paths for the future and something in which they take pride. Despite their interest in other sectors Smallholder households in the “options of work, they are also the most likely for growth” segment comprise 16 per- to want their children to default into cent of the smallholder population. The agriculture without considering other biggest characterizing elements of those opportunities. inclusion, youth, and recent entry into The “options for growth” smallholder agriculture,in this group which are their distinguishes level of financial them households are also motivated, hard- from every other segment. They have ac- percent share they have many aspira- growth” and feel more empowered than tions,working and and nearly opportunistic. all (99 percent) Eighty-five work othercess to groups financial (just tools, 34 morepercent “options feel their for hard to be the best, while 88 percent lives are determined by other, powerful look for opportunities to improve their people). situations. Very few are impulsive

SEGMENT SYNOPSIS

The “options for growth” group has greater access to financial tools and external support, few uneducated individuals, and feels least oppressed by powerful figures. But their youth, optimism and interest in opportunities outside of agriculture could also mean their future takes them in one of two directions, within or outside of agriculture.

The segment relies heavily on agricultural income, but is also the most likely to be engaged in more stable income sources outside of agriculture, such as running one’s own business or working a regular, full-time job.

The youngest of all groups, this segment could pivot in either direction depending in part on how they are cultivated by policymakers, development organizations and financial institutions.

73 National Survey and Segmentation of Smallholder Households in Uganda

(28 percent), and a small portion are These smallholder households mostly inclined to think the future will take care intend to continue working in agri- of itself (21 percent). culture (87 percent), showing similar intentions as other segments. Demographics: Less than half of They derive great enjoyment from it smallholder households in this (90 percent), and many would like to segment live in poverty and are expand their capabilities (92 percent). spread across the central, eastern In fact, they are even more likely to and western regions of the country, want to expand their capabilities than and tend to be younger farmers. other segments.

The “options for growth” group tends Wanting to expand their agricultural to be young, with just over half un- activities der 40 (50 percent). This segment is fairly spread out geographically, with That said, it is worth noting that full-time almost 90 percent living across three employment could also be very attrac- regions (Central: 24 percent, Western: tive. Many say they want to grow their 27 percent, Eastern: 36 percent). farming (92 percent), but, at the same time, most say they would welcome full- Over half of this group (58 percent) time employment (66 percent). This lives above the poverty line. It may seem suggests they may, at some point, deter- counterintuitive that some portion of mine that the best path for their future is a segment called “options for growth” outside of agriculture. falls below the poverty line, however, individual upward mobility and drive Just under six in 10 (59 percent) small- segmentation model was built) can with what their agricultural work has crossfor financial the poverty services line, resulting (off of which in lower the achievedholders in(Figure this 124), segment exacerbating are satisfied this income groups behaving more like high- contradiction. This is also a potential er income groups. call to action in that, if this group cannot be successful in farming and discov- Farming: Experience, income and ers other options, then they may change crops do they need to stay in farming?” given “Options for growth” individuals are theirdirection. limited The success question with becomes, it. “What newer to farming. One-third (33 percent) Up to 59 percent of the smallholder years, a higher percentage by at least households in the “options for growth” 8have percent, been compared farming with for fiveother or groups fewer. group have two hectares of land or less. Forty-six percent have been farming Up to an additional 15 percent have for 10 or more years, the smallest per- between two and three hectares, and the rest have over three hectares of land.38 20 percent being engaged in agriculture This group tends to have larger amounts fromcentage six amongto 10 years the .five segments, with of land than some other segments.

38 The land size measurement comes from the household survey where multiple members of the agricultural household offer up their recollection of various dynamics so as to capture full dynamics instead of relying on just one member’s knowledge of the household. An aggregate estimate of this measure was then created and appended to the segmentation, which is

74 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 124. View of success in agriculture vs. willingness to continue working in it

Opons for growth 59% 87%

Strategic ag entrepreneurship 65% 96%

Farming for sustenance 58% 88%

Baling the elements 59% 86%

Diversiﬁed & pragmac 57% 82%

Sample: All smallholder households who participate in agricultural activities by segment

Nearly 90 percent (88 percent) of “op- Among those who were seriously affect- tions for growth” households gen- ed by any of the above events, 44 per- erate income from agriculture, and, overall, these households tend to use in response. In fact, there is no one both livestock (56 percent) and crops solutioncent of households that conjured did overnothing 32 percentspecific. (85 percent) for income. The “options For instance, when this group experi- for growth” households tend to have enced an unexpected event: the most diverse sources of income be- ■■ 32 percent coped by using savings obtain income from family and friends ■■ 14 percent sold livestock or crops (32yond percent), agriculture, regular at jobs least (27 one-quarter percent), occasional jobs (25 percent), or from ■■ 11 percent borrowed money running their own retail/manufacturing informally business (25 percent). ■■ 7 percent sold assets Sell a high number of crops ■■ 4 percent obtained a temporary job On average, the “options for growth” smallholders are growing seven different Financial attitudes crops each year on their land, the second highest number (6.86 precisely) among The segmentation model itself is built the groups. They tend to sell, on average, four crops they grow (3.74 precisely). bank,off of predictorsmobile money of financial or NBFI inclusion, account Ninety-seven percent of those in the inwhich their is definedname. asIt havingfollows, a full-servicethen, that “options for growth” segment have been ordering segments from “farming for seriously affected by an outside element, sustenance” to the more optimized including weather, pests and disease, groups shows a somewhat linear rela- - - ment failure and/or their own health all, 26 percent of Uganda smallholder issuesaccidents,. Nearly market eight fluctuations, in 10 (79 percent) equip tionship with financial inclusion. Over were affected by weather alone. A simi- meaning that they have a full-service lar amount (77 percent) faced pests and bank,households mobile aremoney, financially NBFI accounts included, in disease issues. their name.

75 National Survey and Segmentation of Smallholder Households in Uganda

Much higher financial inclusion Roughly half (52 percent) have access to an account at an informal institu- Just over two-thirds (67 percent) of tion, most commonly a VSLA (31 per- - cent) followed by a ROSCA or Chama nancially included, which is 21 percent (12 percent). the “options for growth” segment is fi pragmatic” group. While a small portion There is also some use of the following: ofhigher the population,than that of thisthe segment“diversified is alsoand one that offers hope that smallholder ■■ Shopkeepers (10 percent) households can put their livelihoods on a path toward greater stability as well as ■■ Informal saving and credit groups growth. other than ROSCAs or Chamas (8 percent) Mobile money accounts are the most ■■ Moneylenders (6 percent) those in the “options for growth” group. Fifty-sixpopular formalpercent financial have a meansmobile among mon- - ey account, and 64 percent can access nisms, there is acknowledgment of the mobile money either through their own Separate from having financial mecha or through someone else’s account. as saving. Nearly all of the smallhold- Ninety-nine percent of “options for ersimportance in this group of financial (89 percent) behaviors, feel suchit is growth” had heard of mobile money pri- important to save for future purchases, or to the survey. save for school fees (85 percent). Fewer, and many also find it very important to Moderate bank account use for the unexpected (77 percent), or reg- Nineteen percent of smallholders in ularbut stillpurchases most, find(70 itpercent) important. There to savealso the “options for growth” segment have is a strong belief in the importance of an NBFI account, and 31percent have investing in the farm (86 percent very a bank account—the highest among all important). For the “options for growth” segments. Almost two-thirds (64 percent) had been in a bank— four times is considered the most favorable option that of those in the “farming for suste- bothsegment, within saving the segmentat financial and institutions across all nance” segment. five segments (Figure 125).

FIGURE 125. View saving money through different mediums as “very important”

Financial instuon 73%

On a mobile phone 62%

With an informal group 43%

At home 50%

Sample: “Options for growth” households, n5569

76 National Survey and Segmentation of Smallholder Households in Uganda

Segment 5: “Strategic Agricultural the fewest technological barriers. The Entrepreneurship”: Actively group presents itself as having en- engaged, empowered and growing countered—and emerged from—more their agricultural activities39 unexpected events in recent years than have most other segments. Their The “strategic agricultural entre- experience appears to have served preneurship” segment includes just them well given their other desirable 4 percent of Uganda’s smallholder psychographics. households, and comes in as the sec- Like the “battling the elements” group, segment, comprised of those who have these smallholder households tend to emergedond smallest from of life’s all fiveevents segments. empowered This be tenured farmers, who are looking and enabled, is one of the smallest. Like to expand their agricultural activities. - However, there is collectively less fo- come optimized” segments, this tells cus on agriculture as the path for future usthe there“diversified are limited and pragmatic” examples and in “inthe generations. agricultural community of optimization and success. The “strategic agricultural entrepreneurship” smallholder households are motivated, thoughtful in their work, Characterizing attributes: and opportunistic. They have many as- comparisons to other segments pirations (87 percent), take action after The “strategic agricultural entrepre- much thought (90 percent), and most neurship” group is wealthier and more are constantly looking for ways to im- educated, and more prepared in the prove their situations (88 percent). event of an emergency (having ac- They, like most other groups, are risk cess to emergency funds) and faces averse (24 percent are impulsive).

SEGMENT SYNOPSIS

The “strategic agricultural entrepreneurship” segment includes households who appear to be actively engaged in building their agricultural work, with some indications of success or at least progress. The group is more enabled than others, has higher income, more education, greater access to emergency funds, and more financial mechanisms at their disposal. They have been impacted by the realities of farming, and have been able to rely upon their savings or other resources to get through tough times.

What characterizes those in this segment more definitively though is their mindset. They put much thought into what they do. They have big aspirations that include a future in farming. Farming is what they want to do, what satisfies them, where their legacy lives. They are not as likely to want out of agriculture, and are the most satisfied with their achievements in the sector, but they are also the most pessimistic about the future.

This is a group that can be a model or a use-case for carrying meaningful messages (or examples) for growth in other segments of the population.

39 Caution: Small segment size limits analysis. Proceed with caution in extrapolating findings.

77 National Survey and Segmentation of Smallholder Households in Uganda

Demographics: None of the Continue working in agriculture smallholder households in this segment live below the poverty line. These smallholder households mostly They are mostly concentrated in the intend to continue working in agricul- central and western regions of the ture (96 percent), showing similar in- country, and are the second oldest tentions as other segments. They almost group of farmers. all enjoy it (94 percent), and many would like to expand their capabilities (92 per- The “strategic agricultural entrepre- cent). In fact, they are even more likely neurship” group tends to be older, to want to expand their capabilities than with 62 percent over 40; an addition- other segments. al 19 percent are between 15 and 29, and 19 percent are between 30 and That said, it is fair to point out that 39. This segment is lightly concentrat- full-time employment could also be ed in the North (3 percent). The larg- attractive to many in this segment est groups are in the Central region (67 percent). Seven in 10 (65 percent) (38 percent) and the Western region smallholders in this segment are satis- (36 percent), and the remaining 22 percent are in the Eastern region. The en- has achieved (Figure 126), the most tire segment, 100 percent, is above the fied with what their agricultural work- poverty line. cant contradiction, but also a potential callamong to actionall segments,. If this showinggroup cannot a signifi be Farming: Experience, income and successful in farming, and do wind up crops with other options, they may change di- “Strategic agricultural entrepreneur- do they need to stay in farming?” given ship” individuals are relatively tenured theirrection. success The at question it. becomes, “What farmers. One-fourth (23 percent) have Nearly 50 percent of the households ac- an additional 12 percent have been tive in agriculture in the “strategic ag- farmingbeen farming between for fivesix andor less 10 years,years . andAl- ricultural entrepreneurship” segment most two-thirds (64 percent) have been have two hectares of land or less. Up to farming for 10 years or more. an additional 13 percent have between

FIGURE 126. View of success in agriculture vs. willingness to continue working in it

Strategic ag entrepreneurship 65% 96%

Farming for sustenance 58% 88%

Baling the elements 59% 86%

Diversiﬁed & pragmac 57% 82%

Opons for growth 59% 87%

Sample: All smallholder households who participate in agricultural activities by segment

78 National Survey and Segmentation of Smallholder Households in Uganda

two and three hectares, and the rest Financial attitudes have over three hectares of land.40 This group tends to have the largest amount Overall, 26 percent of Ugandan small- of land among all the groups. holder households across the country

have a full-service bank, mobile money, More sources of income orare NBFI financially account included, in their name meaning. The theyma- Just over nine-tenths (91 percent) of jority (63 percent) of the “strategic ag- “strategic agricultural entrepreneur- ricultural entrepreneurship” segment is ship” households generate income from agriculture, and overall, these small- in this group (57 percent) have mobile holder households tend to be more like- financially included, and most of those ly to raise livestock (67 percent) and segments. money accounts—the highest of all five grow and sell crops (85 percent) than Sixty-six percent have used mobile the other groups, suggesting that farm- money, and nearly all (99 percent) have ing is a collection of income sources, heard of mobile money. Bank account not the sole income source. In addition and NBFI account ownership are at to agricultural income, more than one- 20 percent and 18 percent, respectively. eighth of “strategic agricultural entre- Just over half (52 percent) of these preneurship” smallholders generate smallholders have been in a bank. income from running a retail or manufacturing business (26 percent), family Just over half (54 percent) have access and friends (24 percent), occasional to an informal account: jobs (16 percent), running one’s own services business (15 percent), or a reg- ■■ 32 percent have used a VSLA ular job (13 percent). ■■ 16 percent have used a ROSCA or On average, those in the “strategic ag- Chama ricultural entrepreneurship” group are ■■ 13 percent have used a shop keeper growing the third most crops each year on their land at seven (6.82 precisely). ■■ 13 percent have used an informal They tend to sell on average four crops saving and credit group other than they grow (3.91 precisely). ROSCAs

Ninety-seven percent of the “strategic Importance of saving agricultural entrepreneurship” smallholders have been seriously affected by - an outside element including weather, nisms, we see greater acknowledgment pests and disease, accidents, market Separate from having financial mecha- iors, such as saving. Most (90 percent) their own health issues. Just 40 percent feelof the it is importance very important of financial to save behavfor fu- didfluctuations, nothing to equipment cope with these failure events and/or. ture purchases, unexpected expenses

40 The land size measurement comes from the household survey where multiple members of the agricultural household offer up their recollection of various dynamics so as to capture full dynamics instead of relying upon just one member’s knowledge of the household. An aggregate estimate of this measure was then created and appended to the segmentation, which is

79 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 127. View saving money through different mediums as very important

Financial instuon 72%

On a mobile phone 66%

With an informal group 42%

At home 52%

Sample: “Strategic agricultural entrepreneurship” households, n5160

(66 percent), and school fees (75 per- - cent). There is also more emphasis on tural tools are important for them. Their investing in the farm (84 percent). And, focalcertain point financial is often mechanisms the household or agricul and it follows that saving with formal in- household needs (vs. their agricultural stitutions outweighs informal options activities), and connecting to this target (Figure 127). consumer and their household needs could present opportunities for building Market implications for all segments that needed relevance and importance. In a population like smallholder farming What it takes to build inroads into these households in Uganda, where households segments may feel laborious, especially share more attitudinal, behavioral, and given the low income of these individ- circumstantial commonalities than they uals. These two groups, however, also do differences, this segmentation model comprise “critical mass,” as they are the offers a dynamically nuanced perspective two largest segments within the population. The value may be in the collective can be appreciated and even leveraged size of the market, instead of the yield of forso thatpositive uniqueness market interventions within a population. each person in the market. While it is safe to say that a large portion of the population as a whole proceeds “options for growth” groups share with their livelihoods without the use of characteristicsThe “diversified with and the pragmatic” others, andbut - distinguish themselves by their current cilitative tools for improving their house- point in farming–a juncture that could holdformal stability financial and mechanisms agricultural and/or yield, it fa is prove critical. These groups have more really the “farming for sustenance” and options, and may give more serious “battling the elements” groups that are consideration to leaving agriculture if a the furthest away from those mecha- - nisms and tools. These groups also lack self. They take pride in their agricultur- context for why those mechanisms and almore work, profitable and enjoy alternative it, but the presented pragmatic, it tools can be so essential, what they can realistic, and even futuristic side of them bring to their lives, or why they should suggests they might need (and want) to adopt them if available. pursue other paths.

It follows then these are households In Uganda, these two segments com- that will need a greater degree of bined total 20 percent of the small- conditioning, or information about why holder farming population, and have a

80 National Survey and Segmentation of Smallholder Households in Uganda

- - cial services entities looking to make ing in agriculture, growing their farms, productivepromising, inroads attractive into profile the population for finan. buildingdiversified, a legacy are most and committed future in farming, to stay and by many indications, currently do- products, along with life experienc- ing it well. Their value is not in their size. esThey (e .g have., hardship, experience crop loss) with that financial have Their value is in their potential role as - “use cases,” illustrating for other groups upward mobility in a sector that has a nottaught convince them thethem value of the of financialimportance prod of shared passion for their craft. As well, ucts. A financial services provider need have them. Instead, a provider needs to practices can provide an opportunity showfinancial the products,virtues of ora product that they and should how totheir test existing more advanced, financial deeper-reaching knowledge and it enables farmers’ aspirations more so interventions before taking them to the than existing tools. mass market. Even in a more homogenous market, entrepreneurs,” presents the best op- multiple approaches and strategies portunityThe fifth group,for deepening “strategic a household’s agricultural productive—and fruitful—commitment agricultural mechanisms for meaning- to agriculture. However, this is the fulcombine uptake to and best use position within financiala population, and smallest segment, constituting only giving providers the opportunity to 4 percent of the smallholder population. better calculate out their approach and These are households who, while still potential return.

81 National Survey and Segmentation of Smallholder Households in Uganda

7. DESIRES AND ASPIRATIONS: SMALLHOLDER HOUSEHOLDS STRONGLY SEE THE IMPORTANCE OF SAVING AND INVESTING IN FINANCIAL INSTITUTIONS 41 there is a present. Saving with an informal group relatively high level of perceived rele- received the lowest level of importance. vanceContrary among to Mozambique, smallholder households Smallholder farmers in Uganda place most notably a mobile money account, a relatively high importance on their bankin Uganda account of (nonsavings), some financial and products, savings household saving habits. The majority feel it is very important to save for fu- smallholder households said either it is ture purchases, school fees, an unex- “notaccount. important” Between or two-fifths they “do andnot know”half of pected event, and for regular purchases the importance of credit or insurance (Figure 131). The greatest number be- - lieves that saving for a future purchase lar when you ask smallholder farmers is the most important, followed closely about(Figure the 128). perceived The findingsrelevance are of these simi by saving for school fees. When asked what they need to do the most, a plural- activities (Figure 129). ity chose school fees (Figure 132), fol- financial products to their agricultural lowed closely by a future purchase. tools for either their households or their agriculturalThis perceived activities relevance carries of financialthrough on storing and saving money. They to the importance of saving. When placeSmallholder a strong farmers emphasis have unique on storing views asked where they should save, a ma- money somewhere they trust and need- jority of smallholder farmers believe it ing to access their money immediately (Figure 133). They also feel they need institution and even on a mobile phone (Figureis very important130). There to issave a great at a opportufinancial- purpose, and they like to save money in nity here as the inherent importance for caseto store of anmoney emergency somewhere. These for desires a specific do Uganda smallholder farmers is already not always translate into practice.

FIGURE 128. Regardless of what you have, how important is it to your household to have the following?

Mobile money account 64% 24% 8% 3%

Bank account (non-savings) 60% 23% 13% 4%

Savings account 56% 28% 10% 6%

Insurance 35% 25% 24% 16%

Credit 22% 28% 33% 17%

Very important Somewhat important Not important Don't know

Sample: Smallholder households, n52,870

http://www.cgap.org/ . 41 CGAP National Surveys & Segmentation of Smallholder Households in Mozambique, 2015-2016; publications/national-survey-segmentation-smallholder-households-mozambique

82 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 129. How important is it to your agricultural activities to have the following?

Mobile money account 63% 26% 9% 2%

Savings account 57% 28% 10% 4%

Bank account (non-savings) 56% 25% 17% 2%

Insurance 37% 26% 23% 14%

Credit or a loan 24% 30% 33% 14%

Very important Somewhat important Not important Don't know

Sample: Smallholder farmers, n52,771

FIGURE 130. How important is it for your household to save at each of the following?

2% 4% 0% 4% 12% 15% 18% 16% 24% 30% 33% 34%

62% 50% 49% 47%

Save money at a Save money on a Save money at home Save money with an ﬁnancial instuon mobile phone informal group Very important Somewhat important Not important Don't know

Sample: Smallholder farmers, n52,771

FIGURE 131. How important is it for your household to save for each of the following?

Save money for a future purchase 84% 15% 1%

Save money for school fees 78% 15% 6%

Save money for an unexpected 71% 25% 3% event

Save money for regular purchases 66% 29% 5%

0% 20% 40% 60% 80% 100% Very important Somewhat important Not important

Sample: Smallholder farmers, n52,771

83 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 132. Which of the following do you feel your household needs to trail family and friends, (39 percent, save for the most? 25Banks percent and respectively) microfinance. institutions Combined with their habits, this shows that not only do Ugandan smallholders

Save money for a - future purchase tempting to borrow, they would prefer Save money for 29% togo do to so friends in the and future family (Figures first when137 and at school fees 43% 138). When asked where they would at- Save money tempt to borrow from for agricultural Save money for an activities, almost nine-tenths of farmers for regular unexpected said they would go to friends and family purchases event 10% 18% bank (Figure 138). Only 4 percent said theyfirst, hadand ever just attemptedunder half to would borrow go from to a Sample: Smallholder farmers, n52,771 a bank, underscoring the lack of expo- - Smallholder farmers in Uganda place cial institutions. the highest level of importance on insure and experience with formal finan vesting in their farms, and a strong ma- Smallholder farmers may have interest in jority feels it is very important to invest borrowing, but the overwhelming major- in a future educational opportunity ity of farmers do not currently have any (Figure 134). When it comes to what outstanding loans (Figure 140). The top they need to invest in the most, small- reasons for borrowing money would be holder farmers are mostly split between focused on their emergency expenses or an educational opportunity and their businesses, whether it is to start/expand farms (Figure 135). their businesses or use the money for other agricultural activities (Figure 141). Desires and Aspirations: Smallholder households prefer to Desires and Aspirations: There is borrow from friends and family high interest in plans for credit or savings inputs and school fees More smallholder farmers in Uganda recognize the importance of borrowing from family and friends than they do for- orients smallholder farmers in Uganda While financial practices and top interests

mal financial institutions (Figure 136). toward informal financial mechanisms, a FIGURE 133. Do you agree or disagree with the following statements?

I like to store my money somewhere I trust 91% 7%2% I need to be able to access my money immediately 88% 9% 4% I like to store money somewhere for a speciﬁc purpose 85% 12% 3% I like to save my money in case of an emergency 83% 14% 3% I like to save my money in an account because it is safer 68% 25% 7% Storing my money somewhere is easier than saving in an account 61% 31% 8% When my money is in an account, it is constantly working for me 53% 30% 17%

Agree Disagree Don't know

Sample: Smallholder farmers, n52,771

84 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 134. How important is it for your household to invest in each of the following?

Invest money in a farm 81% 16% 2%

Invest money in a future 72% 20% 6% educaonal opportunity

Invest money in a home/home 53% 38% 8% improvement

Very important Somewhat important Not important

Sample: Smallholder farmers, n52,771

FIGURE 135. Which of the following important for this group, presenting an do you feel your household needs to opportunitynumber of financial to build products meaning resonate and rele as- do the most? -

Don’t vance for more formal financial mecha know and credit or payment plans for school 1% feesnisms. and Smallholder inputs (Figure farmers 142) find to savings, be im- Invest money in portant to their agricultural activities. Invest a future Comparatively, prepaid cards and mobile money educaonal in a farm money accounts have less recognized im- opportunity 45% portance. Some of the lower importance 43% comes from lack of awareness (roughly Invest money in a home/home improvement important). Very few smallholder farm- 11% ersone-quarter have any said of these they “don’tproducts know” current if it is- ly, with the highest percentages at 5 and

Sample: Smallholder farmers, n52,771 7 percent for payment and savings plans for inputs, respectively (Figure 143).

FIGURE 136. For your agricultural activities, how important to you is it to borrow from each of the following?

Friends and family 48% 40% 11% 2%

Bank 39% 29% 27% 4% Microﬁnance 25% 33% 36% 6% instuon SACCO 24% 31% 37% 8%

Cooperave 20% 30% 41% 9% Informal money 16% 29% 46% 9% lender

Very important Somewhat important Not important Don't know

Sample: Smallholder farmers, n52,771

85 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 137. In the past 12 months, FIGURE 138. If the need arose, would have you attempted to borrow from you attempt to borrow from any of the any of the following? following?

Friends and family 53% Friends and family 87%

Informal money lender 6% Bank 44%

SACCO 5% SACCO 36%

Bank 4% Microﬁnance instuon 33%

Microﬁnance instuon 2% Informal money lender 33%

Cooperave 2% Cooperave 28%

“Yes” answers “Yes” answers Sample: Smallholder farmers, n52,771 Sample: Smallholder farmers, n52,771

FIGURE 139. What factors would you FIGURE 140. Do you currently have consider when you want to borrow any loans? money?

Quickest access to money 56% Best interest rates 55% Yes Most convenient to get to 41% 19% Best repayment terms 33% Easiest to use 30% Recommended by a friend 23% Have borrowed from them before 23% Loan size 21% Met minimum requirements 16% Trust in a ﬁnancial instuon 16% No Was desperate / no other opons 13% 81% Other 1% Don't know 5%

Sample: Smallholder farmers, n52,771 Multiple responses allowed Sample: Smallholder farmers, n52,771

FIGURE 141. What would be the main reasons for borrowing money?

44% 39% 36% 36% 26% 16% 12% 11% 3% 4%

For To start For other To pay for To buy To cover To improve To make big Other Don't know emergency a new agricultural school fees inputs daily the cash ﬂow purchases expenses business or acvies expenses situaon such as land expand my of my or modern business business equipment

Sample: Smallholder farmers, n52,771 Multiple responses allowed

86 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 142. How important is each of the following products to your agricultural activities?

A savings plan for inputs 35% 30% 24% 11% A credit plan for school fees 35% 26% 28% 11% A payment plan for inputs 33% 29% 27% 11% A goal savings plan or contractual savings plan for school fees 31% 27% 29% 12% A mobile money account that came with 17% 17% 44% 22% a smartphone A pre-paid card to make payments 15% 19% 44% 23% A pre-paid card for receiving income 14% 21% 42% 23%

Very important Somewhat important Not important Don't know

Sample: Smallholder farmers, n52,771

For those who do not currently have Smallholders do not place high impor- these products, the highest demand is tance on loans that come with any kind for payment and savings plans for inputs, showcasing how important these farmers said this is very important to are to smallholders’ agricultural activ- theirof accounts; agricultural less activities than one-quarter (Figure 144) of. ities. School fees present an important Only single digits of smallholder farm- opportunity, as nearly half of smallholders currently have any of these loans, er farmers want a product that gives yet large numbers say they want them them a credit or layaway plan to address (Figure 145). school fees. This comports with what we know about the smallholder’s economic Desires and Aspirations: Mobile cycle. Income is cyclical with the agricul- products conjure interest tural cycle, and payments can be due regardless of whether or not any crops are Many smallholder farmers can see the producing and/or generating income at importance in leveraging their mobile that time. phone as a tool for agricultural activities.

FIGURE 143. Do you currently have any of the following products for your agricultural activities? Do you want to have any of the following products for your agricultural activities?

55% 57% 50% 53%

32% 28% 28%

7% 2% 2% 2% 4% 5% 5% Pre-paid card Pre-paid card Mobile money Goal savings Credit plan Payment plan Savings plan to make for receiving account that or contractual for school for inputs for inputs payments income came with a savings plan fees smartphone for school fees Currently have Want

Sample: Smallholder farmers, n52,771

87 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 144. How important is each of the following products to your agricultural activities

A loan that came with a bank account 24% 25% 37% 14% A loan that came with an insurance plan 23% 21% 38% 18% A loan that is accessed through a mobile money account 22% 23% 39% 16% A loan that is accessed directly through a bank account 21% 28% 38% 13% A loan that came with a mobile money account 21% 22% 39% 17% A loan that is accessed through a mobile money account and 19% 24% 40% 16% linked to a bank account

Very important Somewhat important Not important Don't know

Sample: Smallholder farmers, n52,771

Half feel it would be very important to In Uganda, while small percentages have use a mobile device to access weather the ability to access information, larg- information, and nearly as many said the er percentages are able to charge their phones at a central location or access (49 percent) and market pricing for - theirsame cropsabout (46accessing percent) financial (Figure services 146). The only ability that did not rate as financial services (Figure 147). A sig high was the ability to track shipments abilitynificant to do portion—between all these actions on two-thirds a mobile of inputs and crops on a mobile phone. phoneand three-quarters—said. they want the

FIGURE 145. Do you currently have any of the following products for your agricultural activities? Do you want to have any of the following products for your agricultural activities?

41% 41% 41% 39% 40% 37%

2% 2% 2% 3% 3% 3%

Loan that is Loan that is Loan that came Loan that came Loan that is Loan that came accessed through accessed through with an with a mobile accessed directly with a bank a mobile money a mobile money insurance plan money account through a bank account account and linked account account to a bank account Currently have Want

Sample: Smallholder farmers, n52,771

88 National Survey and Segmentation of Smallholder Households in Uganda

FIGURE 146. How important is each of the following abilities to your household’s agricultural activities?

Ability to access weather informaon on a mobile phone 50% 25% 18% 7%

Ability to access ﬁnancial services on a mobile phone 49% 27% 18% 7%

Ability to charge my phone at a central locaon 48% 26% 19% 7%

Ability to access farming informaon on a mobile phone 47% 29% 17% 7% Ability to access market pricing informaon on a mobile phone 46% 30% 18% 6% Ability to buy and sell on a mobile phone 46% 27% 19% 7% Ability to track shipments of inputs and crops on a mobile phone 40% 28% 23% 8%

Very important Somewhat important Not important Don't know

Sample: Smallholder farmers, n52,771

FIGURE 147. Do you currently have any of the following abilities for your agricultural activities? Do you want to have any of the following abilities for your agricultural activities?

72% 76% 76% 74% 70% 71% 67%

17% 12% 8% 4% 4% 4% 3%

Ability to charge my Ability to access Ability to buy and sell Ability to access Ability to access Ability to access Ability to track phone at a central ﬁnancial services on a on a mobile phone market pricing farming informaon weather informaon shipments of inputs locaon mobile phone informaon on a on a mobile phone on a mobile phone and crops on a mobile phone mobile phone

Want Currently have

Sample: Smallholder farmers, n52,771

89 National Survey and Segmentation of Smallholder Households in Uganda

ANNEX 1: METHODOLOGY AND DESIGN

Sample design. The smallholder house- the distribution of households by region hold survey in Uganda is a nationally rep- and by urban and rural strata (Table 14). resentative survey with a target sample size of 3,000 smallholder households. B. Sample allocation and selection The sample was designed to provide reliable survey estimates at the national In order to take nonresponses into ac- level and for the following four adminis- count, the target sample size was in- trative regions: Central, Eastern, North- creased to 3,158 households assuming a ern, and Western regions. The Central household nonresponse rate of 5 percent region includes central metro (i.e., four observed in similar national households. municipalities surrounding Kampala), - the parishes in Kampala with poultry ac- ed to the four regions proportionally to tivity but it excludes Kampala city, which theirThe totalnumber sample of households size was . firstWithin allocat each is entirely urban. region, the resulting sample was then distributed to urban and rural areas proportionally to their size (Table 15). A. Sampling Frame Given that EAs were the primary sam- The sampling frame for the small- pling units and 15 households were se- holder household survey is the list of lected in each EA, a total number of 216 enumeration areas (EAs) created for EAs were selected (Table 16). the 2014 Uganda National Population and Housing Census. Uganda is divid- The sample for the smallholder survey is a ed into 112 districts with each district comprised of counties/municipalities. was achieved by separating each region Each county/municipality consists of intostratified urban multistage and rural sample. areas. Stratification The urban/ subcounties/town councils with each of them being further divided into parish- population census. Therefore, eight strata es/wards and villages/cells. wererural createdclassification and the is sample based onwas the selected 2014 independently in each stratum. Prior to For the 2014 population census, each the sample selection, the sampling frame village and cell was further divided into was sorted by the nine agricultural zones EAs. Information on the number of ag- called Zardi (Zonal Agriculture Research ricultural households at the EA level Development Institute). was not available in time for the smallholder survey. As a result, the sample allocation of the survey was based on primary sampling units with probability In the first stage, EAs were selected as TABLE 14. Distribution of Agricultural Households by Region, Urban and Rural Strata (2014 Census) Region Rural Urban Total Central* 1,284,685 587,901 1,872,586 Eastern 1,436,156 306,514 1,742,670 Northern 1,152,847 171,743 1,324,590 Western 1,575,224 332,606 1,907,830 Uganda 5,448,912 1,398,764 6,847,676

*Central includes Central metro but excludes Kampala

90 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 15. Sample allocation Region Rural Urban Total Central 592 271 864 adds to 863 Eastern 662 141 804 803 Northern 532 79 611 Western 726 153 880 879 Uganda 2,513 (2,512) 645 (644) 3,158 (3,159)

proportional to size, the size being the The household listing was done on number of households in the EAs. A household listing operation was carried out in develop a script in Dooblo for the listing all selected EAs to identify smallholder smartphones and this required IPSOS to validated before it was used for the list- used in the survey, and to provide a frame ingforms. operation The script. was field tested and forhouseholds the selection according of smallholder to the definition households to be included in the sample. In the D. Sampling weights second stage, 15 smallholder households - The sample for the smallholder house- ability. Due to rounding, this yielded a to- hold survey is not self-weighting, there- talwere of 3,240selected smallholder in each EA households with equal .prob fore, sampling weights were calculated.

C. Household listing design weight based on the probability ofThe selection first component for each of stage the weights of selection is the. The household listing operation was The second component is the response conducted in all selected EAs between rate at both household and individual 15 July and 7 August 2015. For this pur- levels. pose, Intermedia developed a manual describing the listing and mapping pro- The design weights for households cedures. This manual was used for the were adjusted for nonresponse at the training of 30 listing teams held in Kam- household level to produce adjusted pala on 11–13 July 2015. Each listing household weights. Sampling weights team consisted of one supervisor, one lister, and one mapper recruited from were derived from adjusted household Ipsos’ pool of enumerators. The train- weightsfor the multipleby applying respondent to them data nonre file- ing involved both classroom sessions as sponse rates at the individual level. For

well as field practice. the single respondent data file, the same TABLE 16. Distribution of the number of EAs by region, urban and rural Region Rural Urban Total Central 40 18 58 Eastern 44 10 54 Northern 36 8 44 Western 50 10 60 Uganda 170 46 216

91 National Survey and Segmentation of Smallholder Households in Uganda

process was applied after taking into ac- on household assets and dwelling char- count the subsampling done within the acteristics was also collected in order to household. derive the socioeconomic/poverty status of households. A Multiple-Respondent Finally, household and individual sampling weights were normalized adult members in each selected house- separately at the national level so the holdquestionnaire to collect wasinformation administered on their to ag all- the total sample size. The normalized and mobile money use. In addition, samplingweighted weights number were of casesattached equalled to the inricultural each selected activities, household, financial behaviorsonly one different data files and used during household member was selected using analysis. the Kish grid and was administered the

E. Sampling error Single-Respondent questionnaire. The sample design for the smallholder nine languages—Acholi, Ateso, Langi, household survey is a complex sample Luganda,The questionnaire Lugbara, wasLugishu, translated Lutooro, into - Ngakaaramojong, and Runyankole—and - then pretested and validated in all lan- tiondesign. For featuring key survey clustering, estimates, stratificasampling errorstion and taking unequal into probabilities account the of design selec features were produced using either the inguages. all languages Before theto make start sure of fieldwork, that the SPSS Complex Sample module or STATA all three questionnaires were pretested- based on the Taylor series approxima- derstood by the respondents. At the end tion method. questions were clear and could be un

Questionnaire Implementation. To of the pretest, debriefing sessions were capture the complexity of smallholder theheld observations with the pretest from field the staffpretest and. Folthe- - questionnaires were modified based on a script was developed to support data askedhouseholds, of all relevant the questionnaire individuals consist in the collectionlowing the onfinalization mobile phones of questionnaires,. The script household,ed of three notparts, just with one certain household questions mem- was tested and validated before its use in ber (Table 17). In each selected house- in the user guide accompanying the data administered to the head of the house- setthe forfield. this The household questionnaires survey. are found hold, the a Householdspouse, or any questionnaire knowledgeable was adult household member to collect infor- Main Training, Fieldwork, Data Pro- mation about household characteristics. cessing. - Basic information such as age, gender, ner conducted the recruitment of education attainment, schooling status, interviewers InterMedia’s and supervisors local field for part the relationship with the household head was collected on all household mem- their language skills. Following the remain fieldwork, taking into account collected information on whether each - householdbers. The household member contributesquestionnaire to alsothe ingcruitment was conducted of field staff in Kampala by InterMedia’s on 3 to household income or participates in the 8local August field 2015, partner, and included a centralized instruction train household’s agricultural activities. This - information was later used to identify cedures, a detailed review of the survey all household members eligible for the on interview techniques and field pro- tween participants in the classroom, and questionnaires, mock interviews be other two questionnaires. Information

92 National Survey and Segmentation of Smallholder Households in Uganda

TABLE 17. Smallholder survey in Uganda: Questionnaire sections, respondents, and content Questionnaire Household Sample section respondent(s) size Content 1. Household Head of the house- n52,870 • Basic information on all Survey hold, the spouse, household members (e.g., or a knowledgeable age, gender, education at- adult tainment, schooling status) • Information about household assets and dwelling characteristics to derive poverty status 2. Multiple- All household mem- n55,517 • Demographics (e.g., land Respondent bers over 15 years size, crop and livestock, Survey old who contribut- decision-making, associations ed to the household and markets, financial income or partici- behaviors) pated in its agricul- • Agricultural activities (e.g., tural activities selling, trading, consuming crops, livestock, suppliers) • Household economics (e.g., employment, income sources, expenses, shocks, borrowing, saving habits, investments) 3. Single- One randomly n52,771 • Agricultural activities (e.g., Respondent selected adult in market relationships, storage, Survey the household risk mitigation) • Household economics (e.g., expense prioritization, insurance, financial outlook) • Mobile phones (e.g., use, access, ownership, desire and importance) • Formal and informal financial tools (e.g., ownership, use, access, importance, attitudes toward financial service providers)

staff members from InterMedia’s local the areas outside the sampled EAs. Five - a field practice with real respondents in (“QC team”), directly hired by InterMe- thefield QC partner team hired coordinated by InterMedia and super. The dia,independent also attended field the quality training control and staffpar- QCvised team fieldwork stayed activitieswith the insurvey addition teams to

and monitor them. Data collection took Interviewingticipated in the teams field practice.carried out data placeduring from fieldwork 16 August to closely to 7 September supervise collection for the survey on mobile 2015. During data collection, InterMe- phones. Each team consisted of one dia received weekly partial data from supervisor and five interviewers. Two the field, which was analyzed for quality

93 National Survey and Segmentation of Smallholder Households in Uganda

control and used to provide timely feed- yielding a household response rate of 92.6 percent, which is well within accept- checked for inconsistencies and errors able industry parameters for household byback InterMedia to field staff. and The correctionsfinal data file were was surveys of this kind. In the interviewed made as necessary and where possible. households 5,893 eligible household

Response Rates. - interviews. Completed interviews were ble below (Table 18) show household and conductedmembers werefor 5,517 identified yielding for a individual response individual response The rates questionnaire for the Uganda ta rate of 93.6 percent for the Multiple- smallholder household survey. A total of 3,193 households were selected for the 2,870 selected for the Single-Respondent sample, of which 3,101 were found to be Respondent questionnaire. Among the occupied during data collection. Of these, interviewed corresponding to a response 2,870 were successfully interviewed, ratequestionnaire, of 96.6 percent 2,771. were successfully

TABLE 18. Response rates for Household, Multiple-Respondent, and Single-Respondent questionnaires Response Rates by Region and Urban-Rural Central Eastern Northern Western Urban Rural Total Household questionnaire Household selected 865 806 635 887 672 2,521 3,193 Households 817 791 618 875 639 2,462 3,101 occupied Household 714 744 581 831 556 2,314 2,870 interviewed Household 87.4% 94.1% 94.0% 95.0% 87.0% 94.0% 92.6% response rate Multiple-Respondent questionnaire Number eligible 1,417 1,691 1,189 1,596 1,116 4,777 5,893 Number of eligible 1,225 1,605 1,124 1,563 1,014 4,503 5,517 interviewed Response rate 86.5% 94.9% 94.5% 97.9% 90.9% 94.3% 93.6% Single-Respondent questionnaire Number eligible 714 744 581 831 556 2,314 2,870 Number of eligible 662 724 563 823 525 2,246 2,771 interviewed Response rate 92.7% 97.3% 96.9% 99.0% 94.4% 97.1% 96.6%

94 National Survey and Segmentation of Smallholder Households in Uganda

ANNEX 2: RANDOM FOREST ALGORITHM A Random Forest consists of a collection average vote for any other class present or ensemble of simple tree predictors, in the dependent variable. This measure each capable of producing a response provides us not only with a convenient when presented with a set of predictor way of making predictions, but also with values.42 - this response takes the form of a class sure with those predictions. membership, For which classification associates, problems, or clas- a way of associating a confidence mea - For regression problems, Random For- ues with one of the categories present in ests are formed by growing simple trees, thesifies, dependent a set of independent variable. predictorAlternatively, val each capable of producing a numerical for regression problems, the tree re- response value. Here, too, the predictor sponse is an estimate of the dependent set is randomly selected from the same variable given the predictors. The Ran- distribution and for all trees. Given the dom Forest algorithm was developed by - Leo Breiman and Adele Cutler. dom Forest is given by: above, the mean-square error for a Ran 2 A Random Forest consists of an arbi- mean error 5 (observed 2 tree response) trary number of simple trees, which are The predictions of the Random Forest are taken to be the average of the predictions of the trees: simpleused to treesdetermine vote thefor finalthe most outcome. popular For 1 K classclassification. In the problems,regression the problem, ensemble their of Random Forest Predictions 5  K th responses are averaged to obtain an es- tree response K K21 timate of the dependent variable. Using where the index k runs over the individ- improvement in prediction accuracy ual trees in the forest. (itree.e., ensemblesbetter ability can to lead predict to significantnew data cases). incorporate missing data in the predic- torTypically, variables Random. When Forestsmissing candata flexiblyare en- Technical Details countered for a particular observation The response of each tree depends on (case) during model building, the pre- a set of predictor values chosen inde- diction made for that case is based on pendently (with replacement) and with the last preceding (nonterminal) node the same distribution for all trees in the in the respective tree. So, for example, forest, which is a subset of the predictor values of the original data set. The opti- of trees, a predictor variable is select- mal size of the subset of predictor vari- edif, atat a theparticular root (or point other in thenonterminal) sequence node for which some cases have no valid ables is given by log2 M11, where M is the number of inputs. data, then the prediction for those cases is simply based on the overall mean at the root (or other nonterminal) node. simple trees and a set of random predic- Hence, there is no need to eliminate cas- torFor variables, classification the Randomproblems, Forest given method a set of es from the analysis if they have missing data for some of the predictors, nor is the extent to which the average number it necessary to compute surrogate split ofdefines votes a for margin the correct function class that exceeds measures the statistics.

42 See documentation on Random Forest Algorithm at http://www.statsoft.com/Textbook/Random-Forest.

QUERIES:

AQ2: Please provide citation for this figure.