Cash and voucher assistance targeting for education outcomes: How to select beneficiaries to advance equity and maximize results

Targeting Guidance 1 Acknowledgements

This document was produced by Gabrielle Smith, UNICEF would like to thank Global Education Livelihoods, Social Protection and Humanitarian Cluster Cash Task Team members who provided Programming Consultant. The project was comments on this report. Inter-agency collabora- supervised by Lisa Bender, Programme Division, tion was an invaluable contribution to the prod- Education, UNICEF, in partnership with UNICEF’s uct, as were the many contributions from UNICEF Office of Emergency Programmes Humanitarian country offices and national stakeholders around Cash Transfer team. the globe.

December 2020

UNICEF works in over 190 countries and territories to save children’s lives, to defend their rights, and to help them fulfil their potential, from early childhood through adolescence. And we never give up.

2 Targeting Guidance

Glossary and acronyms1

Administrative targeting: Beneficiaries are Institutional targeting: Targeting in which selected from an existing administrative list (e.g. vulnerable children or households are identified school enrolment list, list of social protection and referred by an institution ( e.g. school, scheme beneficiaries, beneficiary list from another protection service provider). CVA programme, list of registered displaced households, households listed in a national social Minimum expenditure basket (MEB): A minimum registry). expenditure basket requires the identification and quantification of basic needs, items and Cash and voucher assistance (CVA): Programmes services that can be monetized and are accessible in which cash transfers or vouchers are directly in adequate quality through local markets and provided to beneficiaries that can be used for services. Items and services included in an MEB goods or services. are those that households in a given context are likely to prioritize on a regular or seasonal basis. Cash Learning Partnership (CaLP): A global An MEB is inherently multisectoral and based network of humanitarian actors engaged in on the average cost of the items composing the policy, practice and research in cash and voucher basket. assistance. Modality: The form of assistance, such as cash Community-based targeting (CBT): Eligible transfer, vouchers, in-kind, service delivery or a beneficiaries are identified by the community. combination of these.

Conditional cash transfer (CCT): Cash assistance Multipurpose Cash (MPC): Cash transfers (periodic that is based on beneficiaries undertaking a or one-off) corresponding to the amount of money specific action (e.g. enrolling in and/or attending required to cover, fully or partially, a household’s school). basic and/or recovery needs. The term refers to transfers designed to address multiple basic needs Delivery mechanism: The means of delivering of the household, including the needs of children, cash or voucher transfers (e.g. mobile money which might include education needs. transfer, cash in hand, cheque, ATM card, etc.) to beneficiaries. Proxy means testing (PMT): Statistical analysis to identify household characteristics that Education in Emergencies (EiE): Programmes correlate with poverty (defined as expenditure/ that work to provide uninterrupted education for consumption), then screening household children affected by humanitarian crises, especially characteristics against a formula which weights, girls, children with disabilities, internally displaced or scores, these according to the strength of the children, refugees and migrants. relationship, generating a poverty score for each household. Geographical targeting: Prioritizing assistance to particular locations and/or to schools in these Scorecard: Combining a range of indicators that locations. are each assigned a score. Data on these indicators

1 Definitions from UNICEF (2018) HCT Programmatic Guidance, incorporating definitions from the Cash Learning Partnership (www.cashlearning.org/). Definition for teacher incentives is from Snilstveit (2015).

4 Targeting Guidance

Glossary and acronyms1

can then be collected through household surveys identify those who are most in need of assistance, to develop a cumulative score, which determines according to the programme objectives. eligibility. Targeting error: Targeting can have inclusion Sector-specific transfer: A transfer that is and exclusion errors. Inclusion errors occur designed to achieve sector-specific objectives by: when people who should not be programme 1) restricting what the transfer can be spent on beneficiaries receive benefits. This is also known (e.g. using vouchers for scholastic materials or as leakage (i.e. programme benefits leak to those transport); 2) clearly communicating to recipients who are not eligible). Exclusion errors occur when the programme objective and intended use of people who should be enrolled in a programme the cash (e.g. to cover education expenditures or are not. This is also known as under-coverage (i.e. to offset income from child labour; this is called the programme does not reach those that it should ‘labelling’); 3) or attaching conditions of certain under the agreed targeting criteria). actions required of recipients (e.g. attendance in school). Targeting mechanism or targeting method: The processes through which households and Self-targeting: This targeting mechanism relies individuals that fit a programme’s targeting criteria on those within the affected population to actively are identified. There are a range of different come forward to join a programme or to apply targeting mechanisms that can be used. to join. The implementing agency does not pro- actively identify households or individuals who fit Teacher incentives: An intervention targeting the criteria. teachers directly, providing payments to improve working conditions in schools and motivate Social protection: Actions carried out by the state attendance. or privately to address risk, vulnerability and chronic poverty. Social protection can include Vouchers: A paper token or e-voucher that can be safety nets, social assistance/social transfers, exchanged for a set quantity or value of goods or labour market policies, social insurance and social services. Vouchers are a restricted transfer. services.

Targeting: The processes by which individuals, households or groups are identified and selected for assistance based on needs. It is a way to focus scarce resources on those within the population that would most benefit from support.

Targeting criteria: A list of individual or household level characteristics that is used to define who within the affected population is eligible to receive assistance. These characteristics should

Targeting Guidance 5

Contents

2 Acknowledgements 4 Glossary and acronyms 6 Contents 8 About this Guidance

10 STEP 1 Select targeting criteria 12 Selecting criteria for education outcomes: tips for success

16 STEP 2 Select targeting mechanisms

20 STEP 3 Implement geographical targeting 20 Geographic targeting for education outcomes: tips for success

22 STEP 4 Implement targeting mechanisms 25 Effective targeting mechanisms for education outcomes: tips for success

32 References 34 Resources for targeting 35 Resources for needs assessments

36 Tool 1 Selection of targeting criteria 39 Tool 2 Selection of targeting mechanisms 45 Tool 3 Geographic targeting 47 Tool 4 Risk analysis on targeting

6 Targeting Guidance ©UN0309462

Targeting Guidance 7 About this Guidance

Education stakeholders increasingly recognize outcomes for education in emergencies (EiE) the role that cash and voucher assistance (CVA) responses, especially regarding the retention of can play in meeting education objectives in girls in education.3 emergencies. To achieve better results for children in an efficient and effective way, and in line with This is in recognition of the wider barriers to global commitments, UNICEF is scaling up use accessing and remaining in school that girls, of cash and voucher assistance across all sectors particularly adolescent girls, face4 and of the in the organization, including education. UNICEF transformative power that education can have on has developed this targeting Guidance to support the lives of girls, young women and, by extension, scaling up of CVA2 in a way that maximizes the wider population.5

©UN0248439

2 Various CVA modalities can be used to achieve education outcomes for children; these are detailed in UNICEF’s Business Case on Using CVA for Education Outcomes in Emergencies. 3 This is aligned with the objectives of UNICEF’s education strategy 2019 to improve equitable access to learning opportunities regardless of gender and UNICEF’s Gender Action Plan to achieve gender parity in education and improve equitable outcomes for girls. 4 European Commission (2018), UNICEF (2016). 5 UNICEF (2015), UNESCO (2013).

8 Targeting Guidance What is targeting? In humanitarian assistance and balance the need for timely responses with programmes, targeting is the process by which the need for targeting those most in need. It aims individuals, households and groups are identified to help minimize errors, while bearing in mind and selected for assistance based on needs. It is the realities and constraints of programming. a way to focus scarce resources on those within a Each section offers concise, practical guidance population that would most benefit from support. for education practitioners, illustrating salient Targeting is one of the most critical elements in points through programme examples, when the design of a CVA programme as it influences relevant. The Guidance contains simple tools, tips the effectiveness of the programme in meeting and checklists of vital considerations to support its objectives. CVA is targeted to the level of the decision-making. individual (in this case, the child) or the household This guidance should be used in conjunction – in comparison to other types of EiE interventions with your organization’s policies, guidance, in which assistance is targeted at the level of the procedures and tools on CVA programming and school. Accountability to Affected Populations.7

Who is the Guidance for? The primary users of What is not included? Because other more general this Guidance are education actors seeking to guidance on targeting for CVA programming engage in, design and implement CVA to achieve exists, this Tool focuses on actions, considerations outcomes in EiE responses.6 and analyses that are important specifically The tool’s secondary users are cash and social for education practitioners and with education protection specialists leading on multipurpose outcomes in mind. It does not provide guidance cash (MPC) or supporting education programme on how to conduct needs assessments to obtain staff in designing CVA for education outcomes. data informing selection of targeting criteria, but offers signposts indicating some best practices What does the Guidance do? The Guidance related to these wider aspects. Another useful aims to equip education actors with advice and reference is the Cash Learning Partnership’s tools to undertake effective targeting of CVA for Programme Quality Toolbox. achieving outcomes in EiE responses. Targeting done well will contribute to creating high quality, consistent and accountable cash-based responses for education outcomes. This guidance provides practical advice and enables practitioners to think through the implications of targeting decisions

SECTIONS OF THIS GUIDANCE

i. STEP 1: Select targeting criteria

ii. STEP 2: Select targeting mechanisms

iii. STEP 3: Implement geographical targeting

iv. STEP 4: Implement targeting mechanisms

6 CVA can contribute to a range of EiE outcomes (UNICEF’s Business Case on Using CVA for Education Outcomes in Emergencies). 7 Also refer to Accountability to Affected Populations: a handbook for UNICEF and partners, www.nutritioncluster.net/resource/ AAP_handbook_for_UNICEF_and_partners.

Targeting Guidance 9 STEP 1

Select targeting criteria

Households and children will not be equally affect- to negative effects of the crisis and who are ed by shocks and crises. Aid agencies usually aim most in need of assistance, according to the to maximize the best use of scarce resources programme objectives. When the objective of by prioritizing the most vulnerable for inclusion CVA is to achieve education outcomes, criteria in CVA programmes. The first step in targeting, should identify those facing highest risk of edu- therefore, is to choose targeting criteria that cational exclusion. allow the programme to define who to prioritize Since CVA primarily overcomes economic barri- for assistance and why. ers to accessing education, the characteristics Targeting criteria are a list of individual or house- for targeting criteria should be indicative of eco- hold characteristics, or indicators, used to define nomic insecurity. In addition to direct economic who within an affected population is eligible to indicators, such as household income, a range receive the assistance. These characteristics of other household characteristics can be used should identify those who are most vulnerable (see Box 1).

©UNI223831

10 Targeting Guidance BOX 1: Selecting targeting criteria for CVA

Characteristics

A household’s or individual’s vulnerability can be defined in various ways, according to their different characteristics, as listed below.

·· Socioeconomic status: livelihood-related factors, such as income and assets at their disposal and capacity to use these. ·· Situational status: displacement status, such as whether they are a refugee, IDP or resident. ·· Demographic categories: defined population groups and demographic characteristics, such as age and gender. ·· Protection needs: protection-related characteristics.

Benefits and limitations of the characteristics

While characteristics can be used as targeting criteria, practitioners should be aware that each has benefits and limitations, which will vary depending on context, as described below.

·· Their robustness and accuracy (whether they are a good indicator for identifying the most vulnerable and economically insecure). ·· Their relevance and acceptability to communities and authorities (whether they are well understood or risk exacerbating existing tensions and social divides). ·· Their ease of measurement (whether it is a visible indicator or self-reported, whether new data will need to be collected or not, which targeting mechanisms will be used and the time and resources required)

Tool 1: Selecting targeting criteria provides guidance on the benefits and limitations of different types of criteria for achieving education outcomes and lists typical indicators that can be used.

©UN0248443

Targeting Guidance 11 Selecting criteria for education outcomes: tips for success

Make use of a range of data sources. While the for Humanitarian Response Plans, needs selection of targeting criteria takes place at the assessments, etc. Education actors can also programme level, the data sources used to inform use findings from vulnerability analyses this selection do not have to be generated at the completed by cash actors in other sectors. level of the programme. Rather, selection can be Criteria selected to inform targeting of other based on data generated at sector and inter-sector CVA programmes, although perhaps not levels, as described below. directly related to education, nevertheless may identify households struggling to cover i. Engage in joint multi-sectoral vulnerability their basic financial needs and, by extension, assessments. Ideally, selection of criteria to struggling to cover costs of accessing be used in a CVA programme should be based education. These sources should be verified, on an analysis of robust household level and ideally triangulated, to reduce the risk of vulnerability data, however assessments can biased or non-representative findings. be complex and costly to implement. Rather than implementing assessments directly at the level of individual programmes, education practitioners can engage strategically, at In Nigeria, a previous study by UNESCO sector level, to participate in multi-sectoral and UNICEF on out-of-school children was socioeconomic vulnerability assessments. used to inform the selection of targeting This can enable more efficient and powerful criteria for UNICEF’s Girls Education analyses, while leveraging the expertise of Programme. clusters and actors with specialized skill sets. When engaging in multi-sectoral vulnerability (UNICEF Nigeria) assessments, ensure that education expenses are factored into Minimum Expenditure Basket calculations and the setting of MPC transfer values. In its Emergency Cash Transfer programme in Iraq, UNICEF undertook regression analysis of existing household data to identify demographic characteristics that In Iraq, the education cluster, along with were correlated with monetary poverty and several other clusters, commissioned risks of dropping out. In contrast, for its REACH to lead a detailed household needs cash for education programme in Mosul, and vulnerability assessment in displaced UNICEF took the vulnerability criteria used and host communities. To inform selec- by UNHCR for targeting multi-purpose tion of targeting criteria, the assessment cash transfers and led a qualitative analysis identified key characteristics of households with the Cash Working Group and the NGO commonly associated with having low and/ cash consortium to decide which of these or temporary forms of income or being indicators were most useful for identifying economically inactive. households for education interventions.

(REACH and UNICEF Iraq, 2016) (UNICEF Iraq)

ii. Make use of available secondary data In Somalia, World Vision’s integrated sources, including education data and Education, WASH and Food Security pro- other sources. When there are constraints gramme targeted households affected by to data collection (lack of time, resources drought with school-aged children. Food or access), education actors can make use insecurity was used as a proxy indicator for of existing secondary data that highlight experiencing deprivations across sectors. a risk of education exclusion. Education clusters and other coordination bodies are (Cristescu (2018); WVI Somalia) increasingly managing secondary data tools

12 Targeting Guidance iii. Use professional judgement. Where access For the ECHO-funded cash for education to household level data, and/or expertise for programme for refugees in Uganda, the de- statistical analysis of relationships between cision to target new arrivals was not based variables is lacking, criteria can be selected on a mapping of economic vulnerability in based on professional judgement and prior the refugee population, rather it was based experience of key informants and qualitative on ECHO’s prior global experience that this analysis of the context. This should be group tends to be among the most eco- informed by consultation with multiple nomically insecure. stakeholders to reach consensus and reduce risk of bias. (ECHO Uganda)

©UN075718

Targeting Guidance 13 Use a mix of targeting criteria. Relying too much For cash for education programmes, harmonization on one type of criteria can result in inclusion and of targeting criteria and coordination of geograph- exclusion errors. Cash for education programmes ical targeting with other cash for education pro- commonly combine a variety of targeting criteria grammes is important and needs to be improved to to fill gaps and exclude fewer vulnerable cases reduce duplication and fill gaps. This can be sup- from assistance. ported by developing guidance at cluster level.

In Egypt, Plan International’s cash for edu- In the drought response in Somalia, ‘house- cation programme for Syrian refugees and hold size’ and ‘number of children’ are children in host communities used a range often included as criteria in MPC target- of targeting criteria to inform selection. ing, ensuring the programme is reaching These were: households with school aged children. However, since the overarching priority in ·· households classified as extremely or the response is to address malnutrition, highly vulnerable according to UNHCR’s child-specific targeting criteria tend to assessment of their ability to cover the prioritize younger children (especially those minimum expenditure basket; under two years) rather than school-aged ·· households headed by single women; children. ·· households with unaccompanied, separated or orphaned children; Cash Working Group Somalia ·· households with out-of-school children of school-going age and/or children engaged in child labour; ·· households where members have a When targeting teacher incentives in South disability or a chronic disease; and Sudan, UNICEF and ECHO harmonized ·· households considered economically- targeting criteria across their programmes deprived, according to the number of to ensure achievement of a consistent employed members, existence of debts, pupil - teacher ratio according to national size of rent and type of housing. guidelines.

(Plan International) UNICEF South Sudan

Coordinate for best results. Joint targeting design Incorporate protection in targeting. It is decisions among actors and clusters involved in or important to consult protection colleagues leading multi-purpose cash (MPC) programming, for the contextualization of protection-related including education actors, is essential. The vulnerabilities and any proposed protection- objective of an MPC programme is to meet the related targeting criteria. basic needs of the most economically insecure households in the population. This broad objective does not necessarily align the programme with households that have children of school going age. Ensuring the inclusion of such households in the programme is a prerequisite to achieving education outcomes in an MPC. It requires that an intentional link is made in the targeting design to maximize consideration of child-specific vulnerability in the selection. Useful criteria to include are: ·· households with children of school age (includ- ing adolescents); ·· age and gender of children; ·· number of children; ·· household size/dependency ratio; and ·· specific vulnerable groups (e.g. children in shared accommodation).

14 Targeting Guidance ©UN0248437

Targeting Guidance 15 STEP 2

Select targeting mechanisms

The next stage of targeting is to define the target- from community-based organizations that provide ing mechanisms, or methods, which can be de- services for vulnerable groups, so called ‘institu- fined as “the process through which households tional targeting,’ are also being leveraged as a way and individuals that fit the programme’s targeting to identify isolated, hard to reach and extremely criteria are identified.” vulnerable cases.

Education actors can make use of a range of No mechanism is perfect: each has benefits and targeting mechanisms. Some, such as communi- limitations and all will generate errors of inclu- ty-based targeting and self-targeting, have been sion and exclusion.8 Box 2 lists possible targeting commonplace in wider CVA programmes. As mechanisms and considerations for their use for humanitarian crises evolve – such as becoming education. increasingly urbanized and protracted in nature – so must emerge new ways of working on CVA and Tool 2: Selection of targeting mechanisms new ways to target. For example, Proxy Means provides more detail on the respective Testing, commonplace in long-term development benefits and limitations to bear in mind for programmes, has been used to target humanitari- EiE programming, as well as examples of their an assistance in several protracted crises. application by education practitioners.

The trend to link humanitarian CVA to national so- cial protection schemes is increasing the adoption of so called ‘administrative targeting,’ in which existing beneficiary lists and social registers are used to identify households for assistance. Efforts to improve harmonization of CVA programmes and generate efficiencies mean it is becoming common to make use of existing registries and beneficiary lists of other aid agencies. In the case of cash for education, another readily available administrative list is school enrolment registers. Finally, in programmes in urban areas, referrals

8 Inclusion errors occur when people who should not be programme beneficiaries receive benefits. This is also known as leakage (i.e. programme benefits leak to those who are not eligible). Exclusion errors occur when people who should be enrolled in a programme are not. This is also known as under-coverage (i.e. the programme does not reach those that it should under the agreed targeting criteria).

16 Targeting Guidance BOX 2: Possible targeting mechanisms and considerations for selection

TARGETING MECHANISM CONSIDERATIONS FOR USE BY EDUCATION ACTORS

Geographical targeting ·· Where vulnerability varies by location, this focuses resources Prioritizing assistance to on areas/schools where the financial barriers to accessing particular locations and/ education are greatest (for example, those with lowest or to schools in these attendance rates). locations. ·· Excludes vulnerable children living elsewhere. This can be addressed through area-based coordination with other education actors providing CVA. ·· Uses some data already available to education actors (school attendance). ·· The geographical targeting of MPC can be defined by indicators that are not education-specific, which may reduce alignment of the programme with those areas with the greatest barriers to education access. ·· The main mechanism for targeting teacher incentives.

Administrative targeting ·· Makes use of readily available data, some of which is already Beneficiaries are accessed by education actors (e.g. enrolment lists). selected from an existing ·· Need to consider whether the data fields match your education administrative list (e.g. targeting criteria (e.g. school enrolment list alone does not school enrolment list, inform about those at risk of dropping out). Can be addressed list of social protection by following up with household interviews to apply any scheme beneficiaries, missing criteria. beneficiary list from ·· Excludes those not on the list which maybe some of the most another CVA programme, vulnerable (e.g. school enrolment lists do not include OOSC). list of registered displaced households, households listed in a national social registry).

Community-based ·· Builds on and can help strengthen existing education-related targeting Eligible committees. beneficiaries are identified ·· Parent Teacher Association (PTA) and school management by the community. committee members have detailed knowledge of which children are vulnerable. ·· Community structures can reinforce existing inequalities, excluding marginalized groups (refugees; lower castes; ethnic minorities). Can be addressed through education committees that include diverse groups and children.

Institutional targeting ·· Benefits from the institution’s knowledge of the population Vulnerable children group, and their social capital within the community. or households are ·· School staff (and protection service providers) must have identified and referred sufficient capacity and must be committed to investing the by an institution – e.g. time required. school, protection service provider.

Targeting Guidance 17 TARGETING MECHANISM CONSIDERATIONS FOR USE BY EDUCATION ACTORS

Proxy means testing (PMT) ·· Education actors work with other CVA actors and employ Statistical analysis to joint targeting approaches, which can spread costs and make identify household char- investments more worthwhile. acteristics that correlate ·· Education actors make use of results of a pre-existing PMT with poverty (defined as administered by other cash actors which can be a quick and expenditure/ consump- efficient way to generate a list of economically insecure tion). Then screening households, for further screening. household characteristics ·· Requires specialist expertise in statistical analysis – expertise against a formula which outside of education actors. weights, or scores, these according to the strength of the relationship, gen- erating a poverty score for each household.

Scorecard ·· A more pragmatic, applicable and lower cost solution than Combining a range of the PMT, aligned with the capacities and expertise of indicators that are each education teams. assigned a score. Data on ·· Risk of exclusion of vulnerable groups most in need of these indicators then col- education support. Can be addressed by combining with lected through a house- institutional referrals (protection service providers). hold survey to develop a cumulative score, which determines eligibility.

Self-targeting ·· Used on longer term social protection style interventions, also The implementing agen- useful in rapid onset emergencies for targeting cash for work cy does not pro-actively programmes linked to school rehabilitation. identify households or ·· On social protection schemes there can be risk of poor individuals who fit the uptake/exclusion errors if there is low awareness of the social criteria. The mechanism protection scheme in the target population or barriers to relies on those within the making an application. Can be addressed with investments in affected population to communication and outreach support. actively come forward to join a programme, or to apply to join.

Education practitioners should consider the mechanism, the data and capacity requirements, trade-offs and how these will manifest in the speed, feasibility, accuracy and transparency. given context, in terms of, for example, the cost The checklist in Box 3 can be used to guide and the resources required to implement the decision-making.

18 Targeting Guidance BOX 3: Checklist for informing selection of targeting mechanisms

FACTORS TO CONSIDER IN YOUR DECISION

Targeting What are the targeting criteria you plan to use? Are these easy to criteria collect data on through the mechanism?

What is the probability that the mechanism will exclude those who fit Accuracy the targeting criteria or include those who do not fit the criteria?

How quickly can households or children be identified through the Timeliness mechanism? Will it be timely enough to enable children to access education according to the school calendar?

How complex is the mechanism to implement? What is required Resources in terms of data, staff numbers, expertise and logistical support to required identify beneficiaries using the mechanism? Is it ‘worth the effort’ considering the programme duration and scale?

Will the mechanism put certain targeted households or Risk of harm children at risk?

Are there security concerns that can restrict mobility of targeted Security and groups or access issues for implementers and how will these affect access implementation of the mechanism?

How often will it be necessary to replicate targeting exercises and how Replicability easy will this be through the mechanism?

How easy is it to explain the mechanism to communities and will they Accountability perceive targeting to be fair?

Can the mechanism work well for targeting your chosen assistance: Modality MPC versus sector-specific?

Inclusion of What is the potential for the mechanism to ensure identification of out- most vulnerable of-school children and other highly vulnerable groups?

Using this mechanism, how easy will it be to coordinate targeting of Coordination this CVA with targeting of CVA being provided by other cash actors?

Given the scale of need and the limitations of each targeting mechanism, education actors should use areas, data from UNHCR’s reception more than one targeting mechanism in combina- centres are used to identify households tion so as to reduce errors, ensure gaps are filled with specific vulnerabilities for further and maximize best use of resources.9 screening (households with children, dis- abilities, female headed households, etc.); ·· institutional targeting – NGOs managing protection referrals also identify poten- In Uganda, NGOs implementing an MPC tially eligible cases; and programme for newly-arrived refugees apply ·· score card – households are visited and several targeting mechanisms. These are: surveyed to understand more about their ·· geographical targeting – focuses the socioeconomic and protection-related programme on specific settlement zones vulnerability. where new arrivals live; ·· administrative targeting – within these (Danish Refugee Council, Uganda)

9 Highlighted, for example, in Mohiddin and Smith (2016), Patel et al. (2016) and UNHCR (2017).

Targeting Guidance 19 STEP 3

Implement geographical targeting

In almost all contexts, geographical targeting is an In the case of teacher incentives, geographical essential first ‘layer’ of targeting for education ac- targeting is generally the only layer of targeting tors. Implementing geographical targeting defines that is implemented as evidence shows it is best to the zone of the intervention to focus on particular blanket target incentives to all staff working in an schools or particular districts and communities.10 institution.11 This will help the programme focus resources on areas (and thus children) with greatest need and Education actors can use a range of indicators coordinate with other interventions implemented as to inform selection of the geographical zone of part of the sector-wide response to avoid duplica- intervention. Tool 3: Geographic targeting lists tion and gaps. Geographical targeting is generally the types of indicators that can be considered undertaken at the sector level and, increasingly, and why, along with examples of potential data inter-sectorally. It precedes implementation of other sources. targeting mechanisms that identify specific house- holds and children.

Geographic targeting for education outcomes: tips for success

Use indicators that highlight locations where house- Consider indicators of education service capacity. holds with children face high economic barriers to Education outcomes from CVA are contingent on accessing school. Of course, the selection of these education services being available and having ca- must be contextualized. For example, often locations pacity to accommodate the level of demand gener- with high levels of poverty incidence will be useful ated by the cash. These indicators are already used to identify those children facing financial barriers to by education actors for geographical targeting of education, however in some contexts where children supply-side education interventions. However, the travel long distances to school, distance from school way that these are used to inform targeting of CVA may be a more important indictor than poverty. interventions is different. Education supply side

10 UNHCR (2015). www.calpnetwork.org/wp-content/uploads/2020/01/operational-guidance-and-toolkit-for-multipurpose-cash-grants-web.pdf. 11 A systematic review by the International Initiative for Impact Evaluation (3ie) found that incentives awarded to individual teachers may act as strong motivation for individual teacher performance. However, they may also create competition among teachers that can discourage teamwork and peer-learning (Snilstveit et al., 2015).

20 Targeting Guidance interventions will logically prioritize those areas non-financial barriers such as a lack of availability where services need strengthening. It is therefore or capacity of education services or attitudinal important for sector-specific cash-for-education barriers to education.12 programmes to be integrated into education programmes when possible, so that geographical Ideally use several indicators. Using several indi- targeting is planned in a coherent way. cators will help build a more accurate analysis of overall vulnerability. For example, since poor atten- Inform decision-making by practical consider- dance can be caused by a range of factors, not only ations. For instance, avoid overlapping with inter- economic insecurity, triangulating data on enrolment ventions of other actors; work in areas in which and attendance with data on service capacity and agencies have a presence and opportunity to build poverty incidence can build confidence as to whether on previous relationships with schools and com- attendance indicates an economic barrier to access- munities; and consider the wider strategic priori- ing education or if there are other barriers at play. ties of agencies, government and donors. After identifying locations, identify schools. With Geographical vulnerability indicators should be a cash-for-education programme, once districts or carefully contextualized to ensure they are ac- settlements of interest have been identified, specific curately interpreted for effective decisions on schools can be identified and the zone of interven- targeting CVA. For example, high out-of-school tion can hinge on these. For instance, concentration children numbers/low rate of attendance in a can be placed on schools with the poorest educa- district could be due to higher financial barriers in tion indicators or those that have the capacity to these locations, but could also be caused by other absorb out-of-school children.

Geographical targeting of UNICEF’s In Somalia, the Intersos cash for educa- Emergency Cash Transfer programme in tion programme for IDPs selected two Iraq started with schools participating in neighbourhoods to work in, based on the UNICEF’s School Based Management pro- IDP population size and food insecurity gramme. indicators. They partnered with schools in these neighbourhoods that could absorb (Cristescu, 2018) more students. The catchment for the pro- gramme in the selected settlements was arranged around these schools.

In Mali, the entire region of Timbuktu faced (Intersos Somalia) huge needs for cash for education. Plan International prioritized limited resources to those districts where it already had a presence and avoided districts where other As part of a drought response in Soma- education actors were working. lia, clusters came together to undertake a multi-cluster needs assessment. Several (Plan International) indicators were used to map multi-sectoral vulnerability and identify priority loca- tions for the response. Indicators included school enrolment and attendance, food In Nigeria, for a girls’ education pro- insecurity, coping strategies and sickness gramme, UNICEF selected schools with the in the household. Geographical targeting highest incidence of out-of-school girls. of MPC focused on those districts and The catchment areas of these were used to localities with highest food insecurity and define the zone of intervention for the cash highest IDP populations. transfer. (Cash Working Group Somalia and Somalia (UNICEF Nigeria) Education Cluster)

12 For guidance on the causes of children dropping out of school and barriers to education, including types of financial barriers to education such as hidden costs, see the complementary UNICEF EiE CVA Response Analysis Tool.

Targeting Guidance 21 STEP 4

Implement targeting mechanisms

For cash and voucher assistance programmes Each approach can make use of a range of target- aimed at achieving outcomes for EiE, education ing mechanisms. Box 4 outlines the main features actors can follow either a school-centred approach of these approaches and considerations for imple- or a community-centred approach, or a combina- mentation. tion of both, to implement targeting.

BOX 4: School-centred and community-centred approaches to targeting for CVA

COMMUNITY-CENTRED SCHOOL-CENTRED APPROACH APPROACH

What the The school is the starting point for The community is the starting approach identifying and screening potentially point for identifying and screening looks like eligible cases. potentially eligible cases.

Can mean using school adminis- Uses targeting mechanisms trative data, using the school to that reach out into communities communicate with caregivers about and household visits to inform targeting, involving school staff in caregivers and identify potentially the selection process and/or using eligible cases. the school as a location to congre- gate households for screening.

Which •• Good at identifying children already •• Can be used to identify both out-of- children will enrolled in and at risk of dropping out school children and those at risk of this approach of school. dropping out. be useful for •• Less useful at identifying out-of- •• Also useful for identifying vulnera- identifying? school children (unless they have ble protection cases and less visible recently dropped out and thus are still groups (child labour, etc). on the enrolment register).

22 Targeting Guidance COMMUNITY-CENTRED SCHOOL-CENTRED APPROACH APPROACH

Used for •• Useful for targeting sector-specific •• Can be used for targeting MPC and targeting ‘cash-for-education.’ sector-specific cash programmes. which CVA •• To date, the school has not been modalities?12 the starting point for targeting MPC programmes for basic needs. •• Also used for targeting teacher incentives.

Advantages •• Aligns with how education •• Enables targeting linked to social practitioners are used to working protection systems. (i.e. through schools). •• Enables outreach into communities •• Can build synergies with the to find the most vulnerable. ‘safe schools’ approach, where •• Can engage other community-based implemented. organizations and stakeholders that •• Puts school at the centre, in line are familiar with family contexts. with the school hub concept. •• Can benefit from effective •• Educators are familiar with family engagement of protection contexts. colleagues and partners. •• Reliance on available data, •• May be a better option when knowledge and relationships support needs to be tightly targeted. can speed up and reduce cost of targeting.

Other consid- •• Children within a household may •• Need to train enumerators erations attend different schools. very well. •• Need to think carefully about how •• Outreach and house to house broad-based any targeting should targeting can be labour intensive be, as other types of assistance that and time consuming. This can limit are provided through the school the size of area that is possible to (feeding, education kits) tend to cover cover. all children. CVA is slightly different, •• When targeting MPC, community- as this assistance is not being centred approaches should be ’distributed’ at the school, however aligned with school catchment there could still be a risk of creating areas, where possible, to maximize tensions. potential for achieving education •• Involving school staff in targeting outcomes. could contribute to risks such •• A lack of links with specific schools as targeting bias, corruption, can make it more challenging to safeguarding risks or taking time away develop integrated programmes, from teaching activities. unless there is good coordination •• Can lead to exclusion error if not with district education authorities. accompanied with other mechanisms •• Challenging in urban/high that capture out-of-school children. population density areas where •• Need to have confidence in the populations in need are more accuracy of any data provided. difficult to find and where it is not •• Will exclude children in non-formal possible to visit all households in a education (unless these centres are district. This requires engagement also engaged as a partner). of knowledgeable community interlocuters and a focus on specific neighbourhoods.

13 See Glossary for details on these different modalities (sector-specific transfer, MPC, teacher incentives).

Targeting Guidance 23 COMMUNITY-CENTRED SCHOOL-CENTRED APPROACH APPROACH

Targeting •• Institutional targeting (teachers •• Institutional targeting (protection mechanisms identify vulnerable children). services identify and refer that fit this •• Administrative targeting (school vulnerable cases). approach provides enrolment list). •• Administrative targeting (e.g. social •• Self-targeting (school protection scheme list, other CVA communicates about the programme beneficiary list, national programme to pupils/caregivers, social register, register of refugees). and/or is the hub where applicants •• Self-targeting (caregivers actively congregate to lodge application/ ‘apply’ for assistance – such as with undergo screening). a social protection programme). •• Community-based targeting •• Community based targeting (e.g. (school management or community through camp committees, religious education committees involved in leaders, local authorities). selection). •• Proxy Means Test/scorecard (house- to-house visits and survey).

In practice, school-centred and community- provided information on which of these students centred targeting approaches can be combined. lived furthest from school and which had poor Norwegian Refugee Council’s cash for education attendance. Then the caregivers of these children programme in Iraq conducted its first phase of were visited at household level by the Muktar, targeting through schools. UNHCR provided who reported on their household situation. a list of students enrolled and the schools

©UN070697

24 Targeting Guidance Effective targeting mechanisms for education outcomes: tips for success

This section highlights some key considerations to Implications for effective targeting of girls. Some bear in mind for implementing effective targeting barriers to education, such as limits to classroom of CVA to achieve EiE outcomes for children. size, lack of teaching materials, concerns about quality of education, or protection risks due to As explained in the introduction to this guidance, a presence of armed groups, will affect all children. core objective for UNICEF is to maximize retention Other barriers can apply more to girls than boys of girls in education. This is in recognition of (while in some contexts, of course, the reverse is the wider barriers to accessing and remaining true). It is therefore important to apply a gender in school that girls, particularly adolescent girls, lens to barrier analysis. Typical barriers facing face and the transformative power that education girls, and ways to address these alongside CVA, can have on the lives of girls, young women and are shown in Box 5. by extension the wider population. The emerging evidence base is showing that CVA can be an Several barriers listed in Box 5 require a medium- effective tool that education actors can use to to long-term engagement to overcome. Barriers achieve EiE outcomes for girls and to overcome may also be possible to address in rapid onset gender disparities in education access.14 Several of and other emergencies contexts where the the sub-sections here provide additional, specific emergency acts as an entry point to longer-term considerations to bear in mind in targeting to programming to address underlying structural maximize EiE outcomes for girls.15 in practice a issues (disaster recovery and ‘ building back contextual gendered approach must always be better’ or programming ‘across the nexus’) and taken. In some contexts, boys can be equally, or where donors commit to longer-term transitional more, vulnerable to specific educational barriers, approaches. with child labour becoming a prominent concern.

Ensure harmonized targeting of CVA with interventions that address wider barriers to UNHCR Pakistan’s CVA programme for girl children accessing education. In many countries, refugees in secondary school had to be there will be multiple drivers underlying why sensitive to cultural constraints placed on children affected by emergencies are out of girls. This was addressed by targeting spe- school, beside economic insecurity. These will not cific supply-side interventions in schools be the same for all children and careful analysis within the CVA zone of intervention: and disaggregation is required to identify how drivers and barriers vary for critical demographic ·· separate girls-only classrooms with groups. Any CVA targeting households and female teachers; children for EiE outcomes must be designed with ·· separate latrine facilities for girls; and these wider demand- and supply-side barriers in ·· ensuring secondary schools located mind to be confident that children/households will within two kilometres of refugee villages also benefit from activities designed to overcome to allows girls to walk to school. the wider barriers. For example, this could be ·· These supply-side investments, through integrating CVA into wider education and combined with the financial reward for education-protection programmes or integration school attendance, encouraged families into wider WASH or nutrition-related activities, to allow their daughters to stay in school. when context demands this. In the case of MPC, it can be achieved through coordination with the (UNHCR, 2017) education and child protection clusters.

14 Evidence from a growing number of experiences of using CVA for EiE outcomes shows that cash is effective at covering the costs associated with attending school, can effectively support continued attendance, reduces risks of dropout for both girls and boys and contributes to a decrease in child labour for girls and boys. Emerging evidence suggests that when designed as part of an integrated and longer-term education programme, CVA can also help to reduce wider sociocultural barriers to girls’ education in crisis contexts. (UNICEF’s Business Case for the use of cash for education in emergencies offers more details.) 15 It is noted that besides girls, other groups of children that can also face greater barriers to accessing education include children with disabilities, displaced children/refugees and those from ethnic minorities.

Targeting Guidance 25 BOX 5: Aligning targeting of wider education and protection interventions to achieve EiE outcomes for girls

POSSIBLE TARGETED EXAMPLES FROM BARRIERS TO CONSIDER INTERVENTIONS EXPERIENCE

Are socio- Are there ·· Behaviour change In Nigeria, UNICEF’s Girls’ cultural negative communication targeting Education Programme factors attitudes to caregivers of CVA included a community a barrier education of recipients. sensitization campaign to girls’ girls in the ·· Engaging influential alongside CVA, engaging education? community? local opinion leaders in influential community Do cultural the zone of intervention leaders to change attitudes practices such as to champion girls’ of caregivers towards girls’ early marriage/ education. education. female-genital ·· Advocacy to education mutilation lead to service providers. drop out? ·· Provide support services Does stigma to to schools to support early pregnan- pregnant girls and cy lead to girls mothers in education. being pulled out of/excluded from school?

Is the school What is the ratio ·· Ensure sanitary kits UNICEF Iraq’s Emergency environment of male to female are prioritized by Cash Transfer targeted girl friendly? teachers? school management schools involved in the Are there committees of targeted School Based Management separate latrines schools/provide in kind programme. These for girls? distribution. schools developed school Are there ·· Recruit and mentor female improvement plans using facilities for teachers for schools in the grants to implement activities ensuring target area. to improve facilities. menstrual ·· Establish adolescent hygiene? spaces and girls’ clubs in target schools.

Are there Are there risks ·· Provide sex education for In the Rohingya crisis in specific of sexual- and girls and boys in targeted Myanmar, displaced adoles- protection gender-based schools. cent girls faced greater bar- risks facing violence en route ·· Protection programming riers to accessing education girls? to or at school? in schools (including due to cultural restrictions on teacher training, their movement. Plan Inter- code of conducts and national’s needs assessment safeguarding). identified the need for sup- ·· Support home-based porting home-based learning learning. for this group. (Joint Education and Child Protection Sub-Sector Needs Assessment, 2017)

26 Targeting Guidance Ensure strong communication with communities validate these with communities. On the other hand, on targeting. Experience demonstrates the impor- communities tend to be more accepting of targeting tance of strong engagement with and sensitization based on clear categorical demographic criteria, of communities to ensure Accountability to Affect- which are visible and are commonly in line with ed Populations and avoid the risk that targeting communities’ own notions of vulnerability and need. contributes to social tensions.16 Evidence from CVA globally shows that communities are more Implications for effective targeting of girls. accepting of targeting when they understand the Gender is a clear and visible criterion that criteria being used for selection and consider communities may accept. However, where these to be fair. poverty is widespread and caregivers struggle to send boys to school, this may also be hard Targeting based on intangible/less visible criteria, to understand and lead to envy and frustration. especially when collated using complex formulae, If girls are to be targeted exclusively, strong generally contributes to a lack of understanding, community sensitization is needed about the frustration and challenges.17 Where economic reasons why this is essential. vulnerability criteria are being used, it is important that investment is made to explain these and

Evaluation of Plan International’s cash for ed- In Uganda, targeting of an MPC pro- ucation programme in Egypt found 77 percent gramme for refugees included consultation of beneficiaries understood the beneficiary with community structures to validate the selection process. However, 23 percent did not vulnerability criteria to be used and ensure understand the criteria used, or the process for they aligned with the communities’ own selection, and felt that targeting was unfair. understanding of vulnerability.

(Alam, 2019) (Danish Refugee Council Uganda, 2019)

©UN0198747

16 Global Education Cluster (2018); Smith (2019). 17 These issues are documented in Smith and Mohiddin (2016), Patel et al. (2016) and Smith (2019) and are main findings of Ground Truth Solution’s research into beneficiary perspectives of humanitarian aid. See CALP (forthcoming) for a summary.

Targeting Guidance 27 Use complaints and feedback mechanism to Complaints and feedback mechanisms should inform targeting design. Processes that seek ideally include more than one communication feedback from communities are essential for channel to maximize their accessibility. Rather CVA programmes to help improve Accountabil- than set up community channels directly, some ity to Affected Populations, build evidence on education programmes have made use of a the efficiency and effectiveness of the targeting Complaints and feedback mechanism ‘hotline’ design, identify and address targeting errors and established for MPC programmes for basic needs. protection risks and negative effects of the target- To identify protection risks facing children, there ing process. This can be achieved through under- should be a feedback channel that is accessible taking post-distribution monitoring activities at for children and caregivers. In cash for education household level and through setting up complaint programmes, practitioners should build on exist- and feedback mechanisms. ing school-based mechanisms to ensure there is an accessible complaint and feedback channel for children inside the school.

In Turkey, to implement the Condition- Implications for effective targeting of girls. al Cash Transfer for Education for ref- School-based complaint and feedback mecha- ugee initiative, UNICEF made use of a nisms should include a communication channel call centre that had been set up under a staffed outside of the school for children, and multi-purpose cash programme run by the especially girls, to be able to safely report any World Food Programme and the Turkish incidence of school-related violence, including Red Crescent. sexual- and gender-based violence.

(Smith, 2017b)

©UN028771

28 Targeting Guidance For best results, coordinate across sectors Analyse and mitigate risks associated with tar- and agencies. Targeting CVA for education geting. As with targeting of any assistance, there outcomes requires coordination between can be risks in the targeting of CVA for education education actors and a range of other outcomes. Most of these risks are not specific to humanitarian agencies and clusters. If more cash modalities and they can be effectively mit- than one cash for education programme is igated when anticipated and addressed through being implemented in an area, coordination of risk analysis. Targeting should be based on a clear geographical targeting between programmes analysis of the likely risk and their impacts. Tool 4: is important to improve accountability to the Guiding Risk Analysis identifies the main risks of affected populations, reduce duplication and which to be aware, potential mitigating measures fill gaps. This can be supported by developing and examples from experience. Protection risks guidance at cluster level. for children, especially girls, are highlighted.

To target for its cash for education pro- In Turkey, the Conditional Cash Transfer gramme in Mosul, Iraq, UNICEF used for Education initiative for Turkish children UNHCR’s MPC programme beneficiary provides larger transfers for older children list. This was a good approach as it avoid- and for girls, acknowledging that in Turkey ed duplication of effort, reduced UNICEF’s adolescents, and especially girls, face great- costs and ensured that communities had er barriers to education. a consistent understanding of eligibility for assistance across the humanitarian (UNICEF Turkey, 2019) response. *** In contrast, for its cash for education pro- gramme for refugees in Lebanon, UNICEF In Uganda, under the ECHO-funded cash for made the decision not to use the same education programme for refugees, NGOs system that UNHCR used for targeting monitor outcomes disaggregated by gender MPC assistance due to concerns about the and age. This data will inform future discus- accuracy of the proxy means test method sions on whether to target older children in identifying specific child-sensitive vul- with larger transfers to overcome the op- nerabilities and the exclusion of non-reg- portunity costs that households incur when istered refugees. sending adolescents to school.

(Smith, 2017b) (Save the Children Uganda)

Education actors are increasingly coordinating Implications for effective targeting of girls. targeting of cash for education programmes with In contexts in which indicators for education the targeting of large scale MPC programmes enrolment, attendance and retention are lower for basic needs to avoid risks of overlaps and for girls than boys, it is common to focus CVA gaps. When an MPC programme provides programme objectives on reducing this gender good coverage of the population in need, and if disparity. However, this does not necessarily mean education actors are confident that their targeting that a CVA programme should exclusively target has successfully included the most vulnerable girls. Targeting only girls can lead to unintended households with children, education actors can negative impacts and protection risks for boys. make use of the MPC programme’s beneficiary To avoid these risks, it may be better to include lists and provide cash for education to those both boys and girls in the programme – especially beneficiary households that have school-aged where economic insecurity is widespread, and children. They can also use the household data boys also face economic barriers. Such assistance collected by the MPC programme and apply can also be differentially targeted according to further education-specific targeting criteria. It may the gender of the child (and the age of child), to be necessary for education actors to still have an provide larger value of assistance to children additional targeting mechanism that identifies that face greater barriers to accessing education households with children that are excluded from – though it is always important to be mindful of the MPC and to fill gaps. potential risks, as highlighted in Tool 4.

Targeting Guidance 29 Key considerations for targeting when linking ·· Is there a need to modify targeting process- with social protection systems. Linking CVA with es to ensure effective and timely inclusion government social protection programmes and of households affected by the emergency? systems can serve as a platform for efficiently Targeting processes of some social transfer delivering humanitarian CVA for education programmes can be bureaucratic and time outcomes through ‘shock responsive social consuming and may create barriers to some protection,’ while also contributing to building vulnerable families being included. In some nationally-owned solutions and providing a contexts, these may be relaxed, or modified, coherent response ‘across the nexus’.18 Experience following an emergency, which can overcome has informed several key considerations, as these constraints. For example, the document described below, to bear in mind in the design and requirements to prove eligibility or the require- implementation of targeting when linked to social ment for a home visit for verification purpos- protection systems. es might be waived to reduce the burden on institutions managing targeting and speed up ·· What is the level of inclusion of the target the targeting process. In other cases, target- group of interest? Achieving education out- ing mechanisms may have been modified to comes from expansion of an existing social incorporate community outreach to increase protection programme requires that: i) the accessibility for isolated and vulnerable house- programme is implemented in those geograph- holds. When the aim is to use social protection ical areas affected by the emergency; and ii) processes to target a new group (for example, the programme uses targeting criteria that refugees), these processes may need to be capture households with out-of-school children adapted to the needs and constraints of this and those at risk of dropping out. Some social new population group. protection schemes are specifically designed to achieve education outcomes and explicitly ·· Will it be possible to access administrative target children of school going age. However, data? Making use of existing social protection programmes targeting based on poverty or programme beneficiary lists or household other demographic groups (e.g. elderly) can records in national social registries can speed still ensure a high degree of overlap with poor up targeting of CVA in emergencies, provided households with school-aged children. the data can be easily accessed by education and other humanitarian cash actors. Education ·· Do targeting mechanisms in national social practitioners must consider data protection protection programmes have sufficient capac- rules governing national social protection data ity to expand to reach new households with and whether beneficiary lists and household children? A social protection programme’s tar- data from social registries can be shared by the geting processes can be used to identify addi- government. Where rules are strict, partner- tional households and children for assistance. ships with government agencies can be a way In this case the institutions involved in target- to overcome these problems. ing must have the capacity to undertake timely and accurate identification of a new caseload following a crisis. Where bottlenecks exist, ed- ucation actors can look at ways to build these capacities. In crisis prone countries, identifying capacities that need to be strengthened can be done as part of emergency preparedness.

18 Cristescu (2018).

30 Targeting Guidance REMEMBER:

·· Targeting is not perfect: all criteria and mechanisms will generate errors of inclusion and exclusion.

·· There is no single ‘best way’ to target CVA for education outcomes, and all options presented in this guidance have their limitations.

·· Targeting decisions must be based on practical considerations of the context — on the realities of funding, on capacities and time available, the availability and quality of data and access constraints.

·· Targeting criteria, mechanisms and selection decisions should be regularly reassessed to account for new information (from monitoring, complaint and feedback mechanisms and other contextual information) to maximize effectiveness and minimize errors.

·· For best results, targeting of CVA should be closely coordinated with targeting of other assistance effectively overcoming barriers to achieving EiE outcomes.

·· Do what is ‘good enough’: select the criteria and mechanisms that allow for the rationing and prioritization of assistance to meet needs as quickly, fairly and transparently as possible. This means striking a balance between accuracy, timeliness and cost, while acknowledging limitations and seeking to mitigate risks, when possible.

·· Incorporate mixed methods: combine criteria and mechanisms to reduce errors and further prioritize resources.

·· In almost all contexts, geographical targeting will be an essential first ‘layer’ of targeting, to prioritize resources to particular areas, combined with other mechanisms.

©UN070695

Targeting Guidance 31 References

Alam (2019). “Impact Assessment of Cash for Edu- REACH and UNICEF Iraq (2016). “Baseline Assessment cation Project Funded by DG ECHO in Egypt.” Plan for Education Cash Transfer Programming for IDPs in International. Dahuk Governorate.”

Cristescu, D. (2018). “Cash and Voucher Assis- Smith (2017a). “Case Study 1: Use of Cash in Con- tance for Education in Emergencies: Synthesis flict-Affected Areas in the MENA Region – Example of Report and Guidelines.” UNICEF and the Global Yemen.” Internal Lessons Learned Case Study Series Education Cluster. Developed for UNICEF MENA Regional Office.

Danish Refugee Council Uganda (2019). Smith (2017b). “UNICEF Case Study 3: Advantages “Multi-Purpose Cash Assistance: Targeting and Challenges of Joint Cash Transfer Programming Guide.” South West Consortium Uganda (Danish in the Middle East and North Africa Region.” Internal Refugee Council, Lutheran World Federation and Lessons Learned Case Study Series Developed for Action Against Hunger). UNICEF MENA Regional Office.

Education Sector and Child Protection Sub-Sector, Snilstveit, B., Stevenson, J., Phillips, D., Vojtkova, Rohingya Refugee Response (2017). “Education and M., Gallagher, E., Schmidt, T., Jobse, H., Geelen, M., Child Protection Joint Rapid Needs Assessment,.” Pastorello, M., and Eyers, J. (2015). “Interventions for Improving Learning Outcomes and Access to Educa- European Commission (2018). “Communication tion in Low- and Middle-Income Countries: A System- from the Commission to the European Parliaments atic Review,” 3ie Systematic Review 24, International and the Council on Education in Emergencies and Initiative for Impact Evaluation (3ie). Protracted Crises.” Brussels. UNESCO (2013). “Education Transforms Lives: Educa- Global Education Cluster (GEC) (2018). “Cash tion for All Global Monitoring Report.” Transfer Programming for Education in Emer- gencies: Report on the Second Instalment of the UNHCR (2017). “Cash for Education: A Global Review Elevating Education in Emergencies Series,” 20 of UNHCR Programmes in Refugee Settings.” November 2018. UNHCR (2018). “Operational Guidance for Cash-Based Juillard et al. (forthcoming). “Contributing to Interventions in Displacement Settings.” Humanitarian Reform through Cash Programming Scale Up.” Research Report by Key Aid Consult- UNHCR et al. (2015). “Operational Guidance and Tool- ing. DFID. kit for Multipurpose Cash Grants.”

Mohiddin and Smith (2016). “A Review of UNICEF (2015). “The Investment Case for Education Needs Assessment Tools, Response Analysis and Equity.” Frameworks, and Targeting Guidance for Urban Humanitarian Response.” IIED Working Paper. UNICEF (2016). “Education Cannot Wait: A Fund for London. Education in Emergencies, Investment Case.”

Patel et al. (2016). “What are the Practices to Iden- UNICEF Nigeria (2017). “Impact Evaluation of UNICEF tify and Prioritize Vulnerable Populations Affected Nigeria’s Girls’ Education Project Phase 3 Cash Trans- by Urban Humanitarian Emergencies? A Systemat- fer Programme in Niger and Sokoto States.” ic Review of Evidence.” Oxfam GB. UNICEF Turkey (2019). “Conditional Cash Transfer for Plan International (undated). “Plan Egypt Vulnerabil- Education Programme for Syrians and other Refugees: ity Assessment Tool.” Fact Sheet Year 2.”

32 Targeting Guidance ©UN0329166

Targeting Guidance 33 Resources for targeting

·· Coady, Grosh and Hoddinott (2004). “Tar- ·· Smith, Mohiddin and Phelps (2017). “Targeting geting of Transfers in Developing Countries: in Urban Displacement Contexts: Guidance Review of Lessons and Experience.” World Note for Humanitarian Practitioners.” Stron- Bank. http://siteresources.worldbank.org/ ger Cities Consortium, IIED. SAFETYNETSANDTRANSFERS/Resourc- es/281945-1138140795625/Targeting_En.pdf. ·· UNHCR et al. (2015). “Operational Guid- ance and Toolkit for Multipurpose Cash ·· Cross and Johnston (2011). “Cash Transfer Pro- Grants. Part 3.3 Targeting strategy and gramming in Urban Emergencies: A Toolkit for determining eligibility.” www.calpnetwork. Practitioners.” Oxford: CaLP. org/wp-content/uploads/2020/01/opera- tional-guidance-and-toolkit-for-multipur- ·· MacAuslan and Farhat (2013). “Review of Ur- pose-cash-grants-web.pdf. ban Food Security Targeting Methodology and Emergency Triggers.” OPM report for Oxfam ·· UNICEF (2018). “UNICEF Humanitarian Cash and ACF. Transfer Programmatic Guidance.”

·· Mohiddin and Smith (2016). “A Review of ·· CaLP’s “Programme Quality Toolbox (targeting Needs Assessment Tools, Response Analy- tab).” www.calpnetwork.org/learning-tools/ sis Frameworks, and Targeting Guidance for programme-quality-toolbox/. Urban Humanitarian Response.” IIED Working Paper. London. http://pubs.iied.org/10796IIED.

34 Targeting Guidance Resources for needs assessments

·· Global Education Cluster (forthcoming). ·· Mohiddin, Smith and Phelps (2017). “Urban “Checklist on Considerations for CVA in EiE Multi-Sector Assessment Tool for Displace- Needs Assessment.” ment Contexts: Guidance Note for Humanitar- ian Practitioners.” Stronger Cities Consortium, ·· ERC Consortium (2018). “Basic Needs Assess- IIED. http://pubs.iied.org/10823IIED/. ment Tool.” https://reliefweb.int/report/world/ basic-needs-assessment-guidance-and-tool- box.

©UN0329265

Targeting Guidance 35 Tool 1 Selection of targeting criteria

Indicators commonly used in EIE programmes are highlighted in red.

WHY USE WHEN TARGETING FOR CONSIDERATIONS TYPES OF CONSIDERATIONS EXAMPLES IN EDUCATION FOR USE INDICATORS FOR USE PRACTICE OUTCOMES

Economic insecurity Requires analytical Income: ··Income is a strong determinant of economic insecurity; Jordan: UNICEF’s is a feature of expertise and good ·· Weekly/monthly income when compared to MEB it shows which households child grant used ‘out vulnerability affecting understanding of context ·· Source of employment/ livelihoods cannot meet minimum basic needs. of school children’ access to education to develop appropriate ·· Number of livelihood sources ··Actual income - difficult to measure in informal sector as an initial proxy and attainment of indicators and timeframes ·· Regularity / predictability of work as it is variable. Livelihood sources are simpler. for economicw other basic needs. of reference. ··Requires contextualization to ensure all relevant vulnerability. Expenditure: income sources are captured. Proxies can be easier to ·· Household weekly/monthly expen- Iraq: UNICEF’s measure than income and diture compared to the Minimum ··Generally viewed as a more accurate indicator cash for education expenditure and are often Expenditure Basket (MEB). than income. programme in Mosul Socio-economic criteria Socio-economic - but not always - good ·· % of overall expenditure for food ··Need to compare expenditure data to the estimated used: TYPE OF CRITERIA OF TYPE predictors of economic ·· % of expenditure for rent cost of meeting MEB to highlight those at risk of not –– current housing vulnerability. ·· Household reports inability to meeting education needs. This can be time consuming. type; meet education needs. ··Difficulties in setting recall period - MEB usually has –– umber of adults Many of these require a monthly or weekly reporting timeframe whereas working; collection of detailed Consumption: education expenditures can be less frequent. –– frequency of work; household data (time ·· Food consumption score ··Alternatively, estimating % of expenditure for survival and consuming - less useful in needs can serve as a proxy. –– coping strategies rapid onset). Other proxy index. indicators: ··Global relevance, in any context. Comparatively easy Some in BOLD can be ·· Out-of-school children and fast to measure. Somalia: TdK’s collected through the ·· At risk of dropping out ··If food consumption is low, can be good evidence that cash for education school. other needs, like education, are not being met. programme in ·· Distance to school ··At risk of dropping out is more subjective to measure Mogadishu used: Many are self-reported and should be triangulated (school attendance data). –– food insecurity indicators, which people ·· Type/condition of housing ··Needs careful analysis to verify this is caused by (meals/day); can misrepresent, under- financial constraints. –– livelihood sources; report or find difficult to ·· Coping Strategies Index (CSI) ··Risk of perverse incentives (see Section Risks). –– main expenditures; estimate - ideally requiring measures what people do, i.e. and triangulation. how they manage to cope when ··Where transportation needed, distance can increase –– CSI. they do not have money to buy financial barriers to access. enough food.

36 Targeting Guidance WHY USE WHEN TARGETING FOR CONSIDERATIONS TYPES OF CONSIDERATIONS EXAMPLES IN EDUCATION FOR USE INDICATORS FOR USE PRACTICE OUTCOMES

Some can be difficult ·· Type of housing can be a good indicator of overall poverty Socio-economic criteria Socio-economic for communities to – ‘condition’ requires careful assessment. TYPE OF CRITERIA OF TYPE understand – need careful communication. ··Can build a more accurate picture of socioeconomic vulnerability as it captures all livelihood capital assets Mechanism most com- available to households and their ability to use these to monly used with scorecard cope with crises. or Proxy Means Test. Self ··Can contextualize to include those linked to education and institutional targeting access - early marriage; child labour; withdrawing too, if additional screening children from school.

through interviews is un- ··Complex, requires good training of enumerators. dertaken by the agency. ··Need to ensure appropriate timeframe of reference to capture the most vulnerable (i.e. those households that have already resorted to such coping strategies).

Demographic Commonly used in CVA Demographic ·· Achieving education outcomes for children means priori- Iraq: UNHCR and characteristics of programmes seeking groups: tizing school-aged children. UNICEF both used the household can education outcomes. ·· School-aged children ·· Needed for education programmes wanting to target CVA criteria of female- influence an individual towards one gender to reduce gender disparities (though headed households, or household’s ability Most are generally easy to ·· Gender (of child) not always necessary or desirable for achieving this out- as evidence showed to access income, verify characteristics that come. See Key Considerations Section). these HH have poor manage risk, and thus are quicker and easier to ·· Disability (of child) Food Consumption likelihood of facing accurately measure than ·· Orphan ·· When marginalization contributes to economic insecurity, Scores, higher social Categorical criteria Categorical TYPE OF CRITERIA OF TYPE economic barriers to income and expenditure. ·· Ethnic minorities/ marginalized this can be relevant. Need to ensure poor access is based vulnerability and worse education. groups on economic rather than social barriers. coping strategies Some (BOLD) can be than male-headed collected through the ·· Can be more robust than simply considering the house- households. school. HH composition: hold head. ·· Dependency ratio (working/non- ·· Large household units can be a key indicator of vulnerabili- Jordan: CARE Easy for community to working members) ty, especially in urban contexts. targeted out-of-school understand – generally ·· Household size children refugees perceived as fair. ·· Households with more school-aged children will face and those at risk of ·· Number of school-aged children greater expenses. dropout, using child Generally good predictors ·· Number of children/ disabled/ ·· Households with these members tend to face greater con- age and a range of of economic vulnerability elderly straints and idiosyncratic shocks, therefore are more likely demographic criteria. (commonly having higher to be (though not always) economically vulnerable. poverty than the average Household head population) though not ·· Female headed household ·· These households tend to face greater constraints and id- every household; can lead ·· Child headed household iosyncratic shocks, therefore are more likely to be (though to inclusion error. ·· Household led by older person not always) economically vulnerable.

37 Targeting Guidance WHY USE WHEN TARGETING FOR CONSIDERATIONS TYPES OF CONSIDERATIONS EXAMPLES IN EDUCATION FOR USE INDICATORS FOR USE PRACTICE OUTCOMES

A household’s The sheer numbers of ·· New arrival ··Newly displaced can often be more vulnerable as they Uganda: ECHO’s ‘displacement status’ displaced means this ·· Time since arrival in have had less time to make economic and social connec- cash for education (IDP, refugee) is a criterion alone is not neighbourhood tions. This can be a proxy for economic vulnerability. programme for necessary criterion sufficient. ··Need to be sure that this holds out in practice or risk refugees targets those - but needs to be excluding more vulnerable long termers. arriving in the last year. further refined and In protracted ·· Host community contextualized to be displacements there ··To be meaningful, needs to be applied alongside addi- Kenya: A study found Status-based criteria Status-based

TYPE OF CRITERIA OF TYPE meaningful. Displaced can be a great range in tional socioeconomic or categorical criteria. Useful to host communities households may be vulnerability of displaced include in programmes moving toward recovery and faced similar less vulnerable than households. resilience in displacement crises, when social cohesion difficulties to IDPs in populations that do is a critical factor. accessing health and not move, for example More meaningful for education, and that if they had money/ targeting when combined targeting only IDPs assets to be able to with indicators of the caused increased afford to move. time or frequency of tension and excluded

19 displacement. the vulnerable. It can also be important to target vulnerable households in host communities that are adversely affected, to avoid social tensions.

A primary objective of Need careful application ·· Children working/children out of ··Out-of-school children can be a proxy for child labour. Somalia: Intersos EiE programmes through protection school ··Indicators identify children exposed to protection risks targeted girls at TYPE OF CRITERIA OF TYPE is to ensure protection. interviews. Requires skill ·· Children who are separated and that can be driven by economic insecurity (e.g. pulling high risk of GBV and

Protection Important to include sets of child protection unaccompanied children out of school and into work as a coping strategy survivors who were children exposed to colleagues. ·· Child marriage increases protection risks for children). referred by members protection risks as a ·· Children associated with armed ··Need to be contextualized through analysis of the of the GBV sub-cluster result of being out of Generally, needs to be forces and groups drivers of protection risks. and received further school. applied discretely to avoid ··All except the first are more specialized areas of screening. risk of doing harm. ·· Girls at high risk of GBV and programming.

survivors who are unable to access school for economic reasons ··Most relevant for programmes seeking reintegration through education.

19 Smith, Mohiddin and Phelps (2017).

38 Targeting Guidance Tool 2 Selection of targeting mechanisms

These mechanisms have a range of benefits and limitations that affect targeting of CVA programmes (education-specific or otherwise); education actors must be aware of these benefits and limitations. Those benefits and limitations specific for cash-for- education are highlighted in red.

©UN0329162

Targeting Guidance 39 BENEFITS OF THE LIMITATIONS/OTHER WHAT IT USE FOR EIE MECHANISM FOR CONSIDERATIONS FOR EXAMPLES LOOKS LIKE PROGRAMMES EDUCATION ACTORS EDUCATION ACTORS

Beneficiaries Starting point ·· When vulnerability ·· Excludes vulnerable chil- In Lebanon, UNICEF’s are selected for all CVA varies by location, can dren living elsewhere. This cash for education from an interventions focus resources on ar- can be addressed through programme for refu- existing (all modalities, eas/schools of greatest area-based coordination gees focused on two administrative all target need. with other education governorates known list (e.g. groups, all ·· The main targeting agencies. to be the poorest and school crises). approach for teacher ·· Geographical targeting of with high concen- enrolment list, incentives. MPC will define the zone trations of refugees. list of social ·· Where needs are high, of intervention based on The programme

Geographical targeting protection provides a pragmatic, other, non-education-spe- targeted all children scheme accountable way of cific indicators which may enrolled in the sec-

MECHANISM MECHANISM beneficiaries, rationing assistance reduce alignment of the ond (refugee) shift at beneficiary that is easy to explain programme with areas of all schools. list from to communities. greatest education need. another CVA ·· Helps to reduce dupli- ·· Any primary data collec- programme, cation and gaps when tion –mapping of neigh- list of part of a well-coordi- bourhoods and services registered nated approach with – can be time consuming displaced other education actors. and requires resources and households, ·· Including capacity of expertise. households schools as an indi- ·· Secondary data must be listed in a cator ensures that accurate and up to date to national social subsequent household be relevant. registry). targeting does not ·· Risk of creating ‘pull fac- overburden. tors’ by targeting partic- ·· Uses some data ular neighbourhoods and already available to ed- schools. ucation actors (school attendance).

Beneficiaries Can be used ·· Makes use of readily ·· Requires data in the list to In Jordan, UNICEF’s are selected in all types of available data, some be accurate and up to date. child cash grant from an crisis, including of which is already Can be addressed by fol- targeted children in existing rapid onset, if accessed by education lowing up with households localities surrounding administrative data is up to actors. to update records (requires double shift schools. list (e.g. date. ·· Low human resources time and resources). The Ministry of school and capacities needed. ·· Need to consider whether Education provided enrolment list, Can be used in ·· Can be used to identify the data fields that are UNICEF with lists list of social a community- households/individu- included match your educa- of enrolled children protection centred als using all types of tion targeting criteria (e.g. used to contact scheme approach criteria. school enrolment list alone caregivers for a beneficiaries, (social ·· School enrolment lists does not inform about household interview beneficiary protection are a way to quickly those at risk of dropping to find children at risk Administrative targeting list from beneficiary list; identify cases (for chil- out). Can be addressed by of drop out. another CVA list of registered dren already in school). following up with house-

MECHANISM MECHANISM programme, refugees; hold interviews to apply In Somalia, Intersos list of list of CVA any missing criteria. used the household registered beneficiaries) ·· Excludes those not on the database of UNHCR’s displaced – relevant list which may be some of refugee repatriation households, for MPC the most vulnerable (e.g. programme to households and sector- school enrolment lists do identify families with listed in a specific or not include out-of-school school-aged children national social school-centred children). for their education registry). approach (list ·· Utilization of such lists CVA. of enrolled requires agreement of the children). managing agency to share In Uganda, targeting the data, which can take cash for education for The latter is time. refugees used data relevant for in UNHCR’s list of targeting cash registered refugees. for education This did not include programmes, data on out-of-school rather than for children; the list was targeting MPC. used to pre-select households with children 6-15 for further screening.

40 Targeting Guidance BENEFITS OF THE LIMITATIONS/OTHER WHAT IT USE FOR EIE MECHANISM FOR CONSIDERATIONS FOR EXAMPLES LOOKS LIKE PROGRAMMES EDUCATION ACTORS EDUCATION ACTORS

Eligible Can be used ·· Easy to explain to ·· Can be difficult to include In Somalia, many beneficiaries in all types of communities and many socioeconomic crite- CVA programmes are identified crisis where perceived as fair, pro- ria, as these need house- have used by the local structures viding the structures hold’s input. community- community. exist. involved are trusted. ·· Needs careful oversight to based targeting, ·· Builds on and can help avoid bias and corruption. including using Can be through strengthen existing ·· Risk of harm – exploitation camp committees a community- education-related com- of children (especially girls) and community centred mittees. due to power dynamics; education approach ·· Educators and parents creating community ten- committees. When (community are among those with sions over selection. committees replicate leaders/ most detailed knowl- ·· Community structures can clan structures and Community-based targeting targeting Community-based authorities) or edge of which children reinforce existing inequal- become dominated school-centred are most vulnerable. ities, excluding margin- by powerful clans, approach ·· Makes use of categor- alized groups (refugees; households from MECHANISM MECHANISM (school ical or status-based lower castes; ethnic mi- minority groups may management indicators. norities). Can be addressed be excluded. Agency committee/ ·· When done through through education commit- validation of lists and Parent Teacher school management tees that include diverse including minority Association). committees, some groups and children. clans in committees

socioeconomic proxy ·· Less appropriate in areas can resolve it. indictors (attendance) with poor social cohesion (Save the Children, could be factored in. and few existing commu- Intersos, TdK, Relief ·· Less resource inten- nity education structures. International) sive than household Can be addressed by in- visits. vesting in developing these In Nigeria, a UNICEF ·· Useful in contexts in structures (in contexts of girls’ education which governance is cyclical/ protracted crisis). programme used fragile and administra- ·· Less appropriate for educa- community groups tive data limited. tion programmes identify- and religious ·· Where access to for- ing protection cases. leaders to support mal civil documenta- CVA targeting by tion is limited, can help verifying the age of to verify eligibility of children without birth households. certificates.

Vulnerable Can be used ·· Takes advantages ·· School staff (and protec- In Egypt, Plan children or in all types of of the teachers’ (or tion service providers) International households crisis. protection service must have sufficient received referrals are identified providers’) knowledge capacity. from staff of informal and referred Can be used in of the population group ·· Services must be known learning centres, who for assistance a community- and their social capital and trusted by the most profiled eligibility of by a particular centred within the community. vulnerable displaced centre attendees for Institutional targeting Institutional institution, approach ·· Can benefit from the households. Some of CVA according to a e.g. school, (protection expertise of trained the most vulnerable lack questionnaire. MECHANISM MECHANISM or a child services – good protection staff. information about services protection for capturing ·· Good way to find very that exist. In Turkey, Concern service out-of-school vulnerable cases and ·· Need firm commitment identified out-of- provider/NGO. children) or a those in need of spe- from the schools (or pro- school children for school-based cialized support (child tection service providers) assistance through

approach (staff protection cases, child to identify and refer those referrals from refer children labour, GBV) in need and share informa- Concern’s protection they consider ·· Screening can be done tion. programme partners. vulnerable for any type of criteria. Cases were then and at risk of screened further by dropping out). CVA programme staff through household Safe schools visits. initiative can be a platform.

Targeting Guidance 41 BENEFITS OF THE LIMITATIONS/OTHER WHAT IT USE FOR EIE MECHANISM FOR CONSIDERATIONS FOR EXAMPLES LOOKS LIKE PROGRAMMES EDUCATION ACTORS EDUCATION ACTORS

Statistical Relevant for ·· Can be considered ‘ob- ·· Requires a representative In Jordan, UNICEF’s analysis longer term jective’ and less prone household data sample child cash grant to identify interventions: to manipulation. and specialist expertise in for refugees was household protracted ·· Can include a range statistical analysis – exper- conceived as a top characteristics crises/social of targeting Indicators tise outside of education up grant to food that correlate protection. – providing there is a actors. and basic needs with poverty proven statistical link ·· Time and resource inten- assistance being (expenditure/ Used on with poverty. Useful sive, requiring detailed provided to families consumption). community- for including socioeco- census-style household by WFP and UNHCR.

Proxy means testing (PMT) testing means Proxy Then centred nomic criteria. survey. This limits regular UNICEF took the screening approaches. ·· Education actors can replicability. Can be ad- list of households household work with other CVA dressed by using existing identified as eligible

MECHANISM MECHANISM characteristics More relevant actors and employ joint administrative data on for assistance against a for large scale targeting approaches, households for the screen- through UNHCR’s formula which programmes which can spread costs ing (though data must be PMT and provided weights, or (not worth and make investments up to date and accurate). the child grant to scores, these investment for more worthwhile. ·· ‘More objective’ does not those beneficiaries characteristics small scale/ ·· Education actors can necessarily mean more that had children. according to short term make use of results accurate. This depends on Alongside this the strength programmes). of a pre-existing PMT robustness of the econo- they targeted child of the administered by other metric model, and the headed households relationship, cash actors which can coverage and consistency and unaccompanied generating be a quick and efficient of data collected in the and separated a poverty way to generate a list of survey. children who were score for each economically insecure ·· Not easy for education not included in household. households for further practitioners to under- UNHCR’s targeting. screening. stand, critique, engage with or adapt. ·· Difficult to explain to com- munities, contributing to tensions.

Scorecards Can be used ·· Any type of criteria can ·· The scoring assigned to In Iraq, UNICEF combine in all types of be included. indicators in the scorecard topped up UNHCR’s a range of crisis, though ·· Scoring, and selec- must be true reflections of MPC programme indicators would need to tion is determined by increasing or decreasing beneficiaries in that are each be kept simple human analysis, which vulnerability. Mosul with a child assigned a in acute phase. is less demanding on ·· Exclusion errors are possi- grant. UNICEF added score. Data resources, time and ble (e.g. vulnerable cases child vulnerability on these Mainly used in expertise than PMT. that ‘buck the trend’). Can questions to the indicators is a community- ·· Allows for human be addressed through existing UNHCR then collected centred adjustments to the in- applying human element questionnaire and MECHANISM MECHANISM

Scorecard through a approach. clusion and exclusion to selection (not based on assigned scores. This household of households that are score alone as this is not was applied through survey to close to the threshold an ‘exact science’). a house to house develop a score. ·· Household data collection survey. cumulative ·· Easier to explain to can be time consuming

score which communities than and resource intensive (es- determines PMT. pecially with larger more eligibility. ·· A more pragmatic, complex scorecards). applicable and lower ·· Risk of exclusion of vulner- cost solution than PMT, able groups most in need aligned with the capac- of education support. Can ities and expertise of be addressed by combining education teams. with institutional referrals. ·· Can be more challenging and time consuming to explain to communities.

42 Targeting Guidance BENEFITS OF THE LIMITATIONS/OTHER BENEFITS OF THE LIMITATIONS/OTHER WHAT IT USE FOR EIE WHAT IT USE FOR EIE MECHANISM FOR CONSIDERATIONS FOR EXAMPLES MECHANISM FOR CONSIDERATIONS FOR EXAMPLES LOOKS LIKE PROGRAMMES LOOKS LIKE PROGRAMMES EDUCATION ACTORS EDUCATION ACTORS EDUCATION ACTORS EDUCATION ACTORS

Statistical Relevant for ·· Can be considered ‘ob- ·· Requires a representative In Jordan, UNICEF’s Agency Useful in ·· Advantages where ·· Demand led – can be risk In Turkey, UNICEF’s analysis longer term jective’ and less prone household data sample child cash grant does not longer term population density, of poor uptake/exclusion CCTE was application to identify interventions: to manipulation. and specialist expertise in for refugees was pro-actively interventions geographical scale and errors if awareness of the based. household protracted ·· Can include a range statistical analysis – exper- conceived as a top identify – protracted insecurity present diffi- social protection scheme characteristics crises/social of targeting Indicators tise outside of education up grant to food households crises and culties in identifying in the target population is In South Sudan, that correlate protection. – providing there is a actors. and basic needs or individuals relief-to- eligible beneficiaries low. Can be addressed with the Girls Education with poverty proven statistical link ·· Time and resource inten- assistance being who fit the development through other means. investments in outreach in South Sudan (expenditure/ Used on with poverty. Useful sive, requiring detailed provided to families criteria. The (targeting social ·· Strict ‘self-targeting’ communication. programme made the consumption). community- for including socioeco- census-style household by WFP and UNHCR. mechanism protection). avoids the need for ·· Where needs are high, the decision to employ

Proxy means testing (PMT) testing means Proxy Then centred nomic criteria. survey. This limits regular UNICEF took the relies on screening of any fur- programme needs suffi- self-targeting, with screening approaches. ·· Education actors can replicability. Can be ad- list of households those within Also used in ther eligibility, saving cient budget to accom- CVA being open for household work with other CVA dressed by using existing identified as eligible affected target rapid onset time. modate the demand, or any girl enrolled in

MECHANISM MECHANISM characteristics More relevant actors and employ joint administrative data on for assistance population to emergencies ·· In school-related cash further additional screen- school. against a for large scale targeting approaches, households for the screen- through UNHCR’s actively come in the case of for work schemes, low ing is still needed. (Charlie Goldsmith Self-targeting formula which programmes which can spread costs ing (though data must be PMT and provided MECHANISM forward to join cash-for-work risk of inclusion errors ·· Risk that households will Associates) weights, or (not worth and make investments up to date and accurate). the child grant to a programme, programmes since wages are gener- not come forward due to scores, these investment for more worthwhile. ·· ‘More objective’ does not those beneficiaries or to apply to linked to school ally low which reduces concerns of stigmatization characteristics small scale/ ·· Education actors can necessarily mean more that had children. join. rehabilitation. applications from less or lack of trust in services. according to short term make use of results accurate. This depends on Alongside this vulnerable households.

the strength programmes). of a pre-existing PMT robustness of the econo- they targeted child School based of the administered by other metric model, and the headed households cash-for-work relationship, cash actors which can coverage and consistency and unaccompanied schemes use a generating be a quick and efficient of data collected in the and separated school-centred a poverty way to generate a list of survey. children who were approach. score for each economically insecure ·· Not easy for education not included in Self-targeting household. households for further practitioners to under- UNHCR’s targeting. on a social screening. stand, critique, engage with protection or adapt. programme ·· Difficult to explain to com- could be munities, contributing to through a tensions. community centred or school-centred approach. Scorecards Can be used ·· Any type of criteria can ·· The scoring assigned to In Iraq, UNICEF combine in all types of be included. indicators in the scorecard topped up UNHCR’s a range of crisis, though ·· Scoring, and selec- must be true reflections of MPC programme indicators would need to tion is determined by increasing or decreasing beneficiaries in that are each be kept simple human analysis, which vulnerability. Mosul with a child assigned a in acute phase. is less demanding on ·· Exclusion errors are possi- grant. UNICEF added score. Data resources, time and ble (e.g. vulnerable cases child vulnerability on these Mainly used in expertise than PMT. that ‘buck the trend’). Can questions to the indicators is a community- ·· Allows for human be addressed through existing UNHCR then collected centred adjustments to the in- applying human element questionnaire and MECHANISM MECHANISM

Scorecard through a approach. clusion and exclusion to selection (not based on assigned scores. This household of households that are score alone as this is not was applied through survey to close to the threshold an ‘exact science’). a house to house develop a score. ·· Household data collection survey. cumulative ·· Easier to explain to can be time consuming

score which communities than and resource intensive (es- determines PMT. pecially with larger more eligibility. ·· A more pragmatic, complex scorecards). applicable and lower ·· Risk of exclusion of vulner- cost solution than PMT, able groups most in need aligned with the capac- of education support. Can ities and expertise of be addressed by combining education teams. with institutional referrals. ·· Can be more challenging and time consuming to explain to communities.

Targeting Guidance 43 ©UNI322051

44 Targeting Guidance Tool 3 Geographic targeting

TYPES OF INDICATOR RATIONALE SOURCES THAT CAN BE USED FOR USE

population of school- Shows areas with highest –– census age children absolute numbers, where cu- –– Humanitarian Needs mulative needs will be great. Overview other estimates Population size

Districts/ neighbourhoods Can compare with enrolment –– UNHCR with high refugee/IDP data to identify hotspots of –– government department caseload (including new out-of-school children. responsible for arrivals) displacement

High presence/ Can identify populations with –– census proportion of greatest need where margin- marginalized groups alization leads to economic (e.g. caste, religion) insecurity.

Districts with high Identify district/administra- –– national household survey incidence of poverty tive unit level. –– other government data Poverty Poverty Neighbourhoods with Identify poverty pockets –– more granular local high concentrations of within settlements/neigh- authority data poor households bourhoods. –– Key Informant Interviews with local authorities –– development committees –– needs assessment data of aid agencies

Districts with poor Education-specific indicators –– Education Management enrolment and commonly used by education Information System attendance data (for actors. Areas with poorer girls/boys) indicators can indicate that economic insecurity Schools with poor is affecting access. Needs enrolment and contextual analysis to ensure Economic insecurity Economic attendance data (for this is due to challenges of girls/boys) economic access rather than other constraints.

High prevalence of food Other indicators commonly –– emergency food security insecurity being collected in the re- assessment sponse that demonstrate High incidence of economic insecurity/inability negative coping to meet basic needs. strategies

High incidence of Can be indicator of economic in- –– multisectoral rapid needs diarrhoeal disease in security (needs contextualizing). assessment young children; Also useful for targeting inte- poor nutrition grated education programmes that focus on WASH.

Targeting Guidance 45 TYPES OF INDICATOR RATIONALE SOURCES THAT CAN BE USED FOR USE

Number of and High level indicator of service –– ministry of education data functionality of services capacity in the area of –– key informant interviews in locality interest. In case of disaster or conflict, Education service Education schools may be closed.

Number of free spaces/ Indicates whether education –– school data overcrowding /child to services are under stress and teacher ratio their capacity to absorb new students.

Distance to school Can define the limits of the –– school catchment plans zone of intervention – to be –– Key Informant Interviews effective, schools need to be with local authorities accessible. –– map of settlements

Other pragmatic indicators Presence and coverage Can highlight gaps in –– 4W data of other interventions assistance and underserved areas. Areas with agency –– strategic documents presence

Government/donor Highlights those geographical –– government action plans priorities areas that are identified –– donor plans as a priority or ‘no go’ for assistance. Insecurity –– conflict hazard reports –– Key Informant Interviews

46 Targeting Guidance Tool 4 Risk analysis on targeting

Those risks that contribute to protection risks for children, especially girls, are highlighted in red.

RISK MITIGATION EXAMPLE FROM EXPERIENCE

Security context ··Continually monitor the In Yemen, insecurity from conflict restricted restricts access operating context. access to communities for UNICEF’s CVA to communities ··Rely on administrative for families with children. UNICEF relied for targeting targeting and referrals. on its existing database of Muhamasheen ··Apply a school-centred beneficiary households. Additional eligible approach rather than out- households were also identified through reach in communities. a community grievance and appeals ··Employ third party orga- mechanism. nization able to operate (Smith, 2017a) in the context.

The targeting ··Protection risk assess- In Uganda, targeting of ECHO’s cash for method used ment accounting for education of refugees was informed by a exposes possible drivers in the protection risk assessment. households/ context. children to ··Clear SOPs for the pro- In Somalia, Intersos engaged gender-based protection risks cess of household visits. violence programme staff to guide the ··Consult child protection method for targeting vulnerable protection (e.g. danger and gender experts on cases. traveling to ways to reduce risks for targeting vulnerable groups includ- site; risk of ing girls and child headed In Jordan, Save the Children’s education exploitation by households. and child protection programme had clear those selecting ··Post-distribution mon- procedures to follow for staff of partners beneficiaries) itoring of beneficiaries conducting household interviews to reduce to check if they were risk of harm, including the need to visit charged, harassed, etc. households in pairs and a mix of male and ··Independent complaints female enumerators. mechanism that caregiv- ers/ children can access.

Targeting Guidance 47 RISK MITIGATION EXAMPLE FROM EXPERIENCE

The targeting ··Risk assessment of tar- In Nigeria, UNICEF’s Girls’ Education criteria create geting approach, espe- Programme did not only target children perverse cially decisions to target that were out of school to avoid perverse incentives out-of-school children, or incentives. However, the evaluation noted detrimental to by gender. that targeting only girls had a negative impact child wellbeing ··Consult child protection on parents’ attitudes to boys’ education. and gender experts. (e.g. targeting ··Include children that are In Uganda, ECHO’s cash for education children out of in school as well as out programme for refugees is targeting out- school means of school; include boys of-school children. This has led other poor children are as well as girls (reduce community members that are struggling removed from geographical scope if financially to comment that they should education; need be). remove children from school to benefit from targeting girls ··Strong sensitization of future rounds of support. NGOs are trying to means boys communities of the ratio- mitigate the risk with strong sensitization. pulled out nale for this approach. of school to ··Careful verification of In Somalia, part of WFP’s school feeding replace girls’ child and caregiver programme included a voucher for a role at home; relationships (through take home ration, targeted at girls. This a family sends documentation or com- successfully promoted girls’ retention but one of their munity). there was anecdotal evidence that for some daughters to live households this led to boys being taken out with another of school. household to (NCA Somalia) ‘double dip’ in the programme, In Nigeria, Save the Children received exposing the anecdotal evidence that families without child to risk of school aged children were borrowing children exploitation and from other relatives to be included in the neglect) cash for education programme. This risk is monitored through community complaints mechanisms.

Corruption ··Segregation of functions In Yemen, UNICEF invested in independent by those in targeting process. accountability functions and independent responsible ··Oversight of targeting third-party monitoring for its CVA for for identifying and third-party verifica- conflict-affected households with children. beneficiaries tion. This mechanism reduced the risk of ··Independent monitoring fraud and targeting errors by programme (e.g. ghost and complaints and feed- implementers. beneficiaries, back mechanisms. (Smith, 2017a) selecting ··Zero tolerance to corrup- non-eligible tion. households, taking bribes)

48 Targeting Guidance RISK MITIGATION EXAMPLE FROM EXPERIENCE

Targeting is ··Protection risk In Somalia, Relief International and Intersos not effective at assessment taking into developed SOPs to ensure consistency including the account possible drivers in application of CVA targeting processes most vulnerable in the context. across partners, including zero tolerance to ··Base selection of criteria corruption. (e.g. criteria are on best evidence not effective available. In Turkey, targeting of UNICEF’s Conditional indicators of ··Avoid over reliance Cash Transfer for Education initiative for vulnerability; on self-reported refugees is based on applications made by households vulnerability indicators families at the offices of social assistance manipulate that are hard to verify. foundations. This created some access the selection ··Invest in strong challenges for vulnerable households that process; most communication with had difficulty filling in the form or traveling vulnerable affected populations. to the centre offices. This was resolved face barriers ··Employ community- through investing in additional activities to to accessing centred approach and support these vulnerable households. To targeting; poor outreach targeting apply, families also had to have a refugee or inconsistent mechanisms (referrals registration card from the government implementation from community-based department responsible. This department of targeting organizations and service was overworked and faced administrative mechanisms; providers; house to backlogs. UNHCR provided staffing and IT bias/ house visits). support to this department to reduce the discrimination ··Do not rely on an bottlenecks and the ministry of education in the selection application-based coordinated with the department to fast track process) approach if the target the registration of families with children, so group will face access they could access the UNICEF Cash Transfer challenges. for Education initiative. ··Oversight of actors (Juillard et al., forthcoming) involved in selection, especially community committees. ··Ensure marginalized groups are represented in selection committees. ··Complaints and feedback mechanisms and post- distribution monitoring to flag erroneous targeting, and feedback loops to inform programme design. ··Develop an SOP for consistency in applying targeting mechanisms.

Targeting Guidance 49 RISK MITIGATION EXAMPLE FROM EXPERIENCE

Targeting ·In displacement contexts, · In Nigeria, for its Girls’ Education Programme, contributes to consider the need to UNICEF blanket targeted all girls of school community include vulnerable age in the selected area to avoid community and household households/children in conflicts. The evaluation noted that the tensions host communities. decision to target only girls led to difficulties ·Consider blanket · within households in which caregivers had to (e.g. targeting targeting inside a make hard choices between girls and boys, is not well geographical area. and that the decision to focus only on girls understood, ·Risk analysis of proposed · left boys feeling frustrated and envious. not considered targeting criteria, taking fair, excludes into account community In Uganda, NGOs implementing ECHO’s cash particular dynamics and social for education programme for refugees found households tensions between groups that targeting only out-of-school children or groups, of interest. led to social tensions, with poor households contributes to ·Risk analysis of · whose children were in school but struggling intra-household geographical targeting, to meet other basic needs feeling aggrieved. tensions if taking account of the restricted by location of different gender or age) ethnic, religious and marginalized groups. ··Strong communication to communities on the chosen criteria and selection processes. ··Use simple, categorical criteria. ··Consult communities on their understanding of vulnerability.

50 Targeting Guidance

52 Targeting Guidance