pproaches46700 O c t o b e r 2 0 0 8 N o t e N u m b e r 2 2 Targeting Subsidies Through Output-Based Aid Yogita Mumssen, Geeta Kumar and Lars Johannes O utput-based aid (OBA), or performance-based grants, · Corruption, which motivates people to provide can be used to help target services to the poor. Under false information to obtain service to which they OBA schemes, service providers (usually the private are not entitled, or to exploit vulnerable target sector but in some cases also community or nongovernmental groups. organizations or public sector utilities) are compensated only · Lack of incentives for service providers to connect after delivery of a specified output, such as water connections of poor households eligible for lower "social tariffs" a specified quality, to a targeted beneficiary. In most cases that that are part of cross-subsidy schemes. targeted beneficiary would be a poor household or community. Subsidies are provided in the form of payments for the provision For utility services, additional factors come into of service to targeted groups to help cover the gap between the play. In some poor countries large swaths of the cost of provision and the user's ability to pay. population, including even the better-off, may not be Targeting subsidies involves challenges. OBA can provide connected. Where this is so, politicians and service opportunities to enhance targeting and lead to greater transpar- providers may see greater gains in serving the better-off ency in reaching the poor. While OBA is not yet mainstreamed customers first--even though the services may be sub- in all infrastructure and social services sectors, the principles and sidized through utility tariffs that are inefficiently low, mechanisms are compelling, and initial results promising.1 leading to wasteful use of public resources. The notion that the poor cannot afford to pay sustainable tariffs The challenges may be another factor. Yet many interventions, as well as studies and surveys, have shown that a large share Subsidies have long been used to make basic ser- of poor people are willing and able to pay sustain- vices more affordable and are seen as a vehicle for able tariffs--and usually already spend much more on redistributing resources. But universal subsidies are alternative services such as water from water vendors widely agreed to be inefficient and fiscally unsustain- or batteries for power supply. able. Good targeting practices can help both reduce the need for subsidies and ensure that they reach the Where do subsidies go? intended beneficiaries. The equity and effectiveness of subsidy targeting is commonly evaluated by the extent of coverage of the What are the constraints? target group and the "leakage" to non-target groups If the need for targeting is generally acknowledged, (also known as errors of exclusion or inclusion). why haven't subsidies been more effectively targeted to Experience in many countries suggests that projects reach the poorest? The failure to do so in the infra- in infrastructure and to some extent in social services structure and social services sectors can be attributed have failed to target subsidies effectively to the poor to a host of factors: and vulnerable. Many have lacked an explicit target- · Lack of incentives for politicians to support pro- grams targeted exclusively to the poor. 1This note draws upon lessons from a wide range of OBA · Lack of understanding of the circumstances, prefer- projects, although mostly from GPOBA-funded projects for ences, and types of service useful to the poor. which detailed targeting information is readily available. · Requirements for sophisticated, and often costly, data systems to provide accurate measures of Yogita Mumssen is a senior infrastructure economist with the income. World Bank's Finance, Economics, and Urban Department. · Lack of awareness resulting in low uptake by the Geeta Kumar and Lars Johannes work for the Global target group. Partnership on Output-Based Aid. Supporting the delivery of basic services in developing countries pproaches ing mechanism for subsidies, and any efficiency gains provision of pay phones or internet points of presence. achieved were purely incidental. Projects using OBA connection subsidies usually rely on In the water and energy sectors, subsidies are most the target population being able to afford sustainable commonly channeled through the tariff so that users do tariffs that cover ongoing costs of service provision. But not pay the full cost--sometimes not even the running this is often not a major hurdle; as noted above, the costs--of the service. A seminal study on tariff subsidies poor usually are paying more for alternative services.2 for water and electricity services found that these sub- OBA projects in roads, health, and education nor- sidies are invariably regressive, benefiting the non-poor mally fund the ongoing provision of basic services or more than the poor (Komives and others, 2005). This is maintenance. OBA road maintenance schemes require due largely to one of two reasons: the poor are not con- ongoing subsidies, often funded through road funds. nected to the system in the first place; or there are only OBA health schemes, to ensure continued access to minimal differences in consumption patterns between care for the poor, need to channel subsidies in an the rich and the poor, so that quantity-based subsidy ongoing manner through health care providers as they schemes such as increasing block tariffs actually lead to deliver agreed services (such as well-child visits) over a considerable leakage. The study argues that alternative period of time. But some health projects may focus on forms of targeting--such as connection subsidies that one-off interventions, such as safe-child delivery. support initial access rather than consumption--would be more effective in reaching the poor. Sharpening traditional forms of targeting Targeting also presents a challenge for the social No matter the subsidy design, all OBA schemes must services sectors. This is partly because the poor do not specify the outputs against which subsidies will be dis- have access to facilities--a problem particularly serious bursed--such as household water connections, installed in rural areas--or because funding is concentrated on solar systems, or medical treatments--and consequently services used mostly by the non-poor, such as secondary can identify the beneficiaries more clearly than tradi- and university education. Additionally, public facilities tional input-based schemes. Eligibility criteria for ben- are usually of poor quality due to mismanagement and/ eficiary households can be clearly defined and made a or the diversion of funds. Therefore, the key for subsidy precondition for subsidy disbursements. The third-party targeting in these sectors is to ensure that well-trained verification that triggers disbursement of OBA funds can professionals (e.g., nurses or teachers) are actually de- sometimes even include verification that the targeting livering a service in a form that the poor can and choose criteria have been met, so that subsidies have to be paid to use, for example through contracting schemes that only for eligible households or individuals. pay on the basis of services delivered to the poor. OBA also helps to address the issue of low uptake by the poor, thereby ensuring greater effectiveness of OBA and targeting a subsidy scheme. For example, in the case of utility services, because a proportion of payments to providers OBA schemes can provide subsidies in three ways: tran- (or "supply-side" subsidies) are withheld until proof of sitional tariff subsidies (which taper off as user contri- at least one billing cycle, service providers take on addi- butions increase), ongoing subsidies, or one-off subsi- tional demand risk--and so have an incentive to address dies such as connection subsidies. Which subsidy design the risk of low uptake. For example, both the Rural is chosen will depend on such factors as the sustain- Community Water Project in Andhra Pradesh and the ability of the funding, the capacity for administering the Senegal On-Site Sanitation Project work with NGOs to subsidy scheme, and the type of service to be subsidized. promote community participation to improve uptake. Choosing subsidy design Finally, there are also some questions about the effectiveness of targeting in relation to the quality of For utility services, ongoing OBA schemes, such as a services for the poor. This is another area where an OBA well-known one used in Chile's urban water sector, can design can help ensure quality of service is not com- be effective in reaching the poor only if they are con- nected to the system. Thus in the poorest countries, most OBA schemes in water, energy, and telecommuni- 2Under OBA schemes, tariff affordability issues for the very cations instead involve one-off subsidies enabling initial poorest can be partially mitigated by subsidizing connections access. For example, an OBA subsidy may be used to such as public water points, which have lower running costs help connect a poor household to the water or electric- per capita, and by ensuring appropriate payment schemes ity network or to reduce a community's contribution for for the poor to go along with the subsidy scheme. Supporting the delivery of basic services in developing countries pproaches Box 1 Nepal Biogas Support Program--Combining Geographic and Self-selection Targeting The GPOBA and Community Development Carbon Fund- supported Biogas Support Program in rural Nepal provides household-size biogas plants to families. The subsidies vary according to the plant's size and location. Smaller plants, used by poorer families, receive higher subsidies than larger plants. Wealthier families, with more livestock to provide input, prefer larger plants with greater gas output. Plants in remote mountainous regions, where the population is poorer, receive a higher subsidy than plants in the Terai lowlands, where the population is richer. higher share of subsidies. Subsidies can be targeted progressively by providing higher subsidies for more basic services--for example, smaller solar home sys- An OBA scheme in Armenia limits eligibility to the poorest tems, as in a rural electrification project in Bolivia--or and most vulnerable, using "poverty scores" assigned by the by subsidizing services less attractive to the rich--such government's Poverty Family Benefit Program to identify as public water points, as in a GPOBA-funded water beneficiary households. (Photo © World Bank) project for small towns in Uganda. Self-selection targeting can also be achieved by promised: the definition of "outputs" includes quality introducing a service to an unserved area and starting parameters and in some cases even services that extend a subsidy scheme only after some time. Households out for several years, such as maintenance of and parts that can afford to connect immediately will do so, as supply for solar home systems, to help ensure a certain long as the benefits of service during the time lag will degree of quality and sustainability. outweigh the cost. With this approach, self-selection can be used even for homogeneous outputs such as In sum, output specification and verification, pay- electricity connections, as is being attempted in an ment only on delivery of outputs, and demand man- IDA/GPOBA-supported electricity project in Ethio- agement incentives all allow OBA to more effectively pia. Self-selection targeting is also widely used in the reach the poor using some of the traditional methods health sector. Because wealthier patients tend to favor of targeting, such as geographic, self-selection, and more sophisticated, up-market facilities, OBA projects means-tested targeting. usually finance more basic health care services. Many Geographic targeting is usually the easiest way to reach GPOBA projects have used self-selection to comple- the intended beneficiaries. It is useful in areas in which ment geographic targeting (Box 1).3 intended beneficiaries are concentrated and few people Means-tested targeting is used in several OBA schemes outside the target group live. For projects in such ar- in middle-income countries. Means testing involves eas, excluding unintended beneficiaries can be costlier measuring a beneficiary's wealth to assess whether than including them. One example is a GPOBA-funded a subsidy is warranted. Such schemes require more water project in Uganda that will be focusing on slum advanced administrative systems, as in the case of the areas of Kampala, where most households are very urban water project in Chile mentioned above. For this poor and excluding the few non-poor households reason, OBA schemes that rely on means testing usual- would be too costly. ly piggyback on broader welfare programs that identify An analysis of GPOBA projects shows that most poor households for a variety of public services. This is use geographic targeting. More than half use it as the only targeting mechanism. Many GPOBA projects are pilot projects in very poor countries and their small size 3OBA projects in roads tend to use geographic targeting. makes geographic targeting relatively easy. Because they are a "public good" (meaning that no one can easily be excluded from using them), most forms of targeting Self-selection targeting involves designing projects so are difficult. Ideally, OBA in roads would be complemented by that outputs chosen by poorer beneficiaries receive a self-selection targeting for transport services such as buses. Supporting the delivery of basic services in developing countries pproaches Table 1: Cost and Effectiveness of Targeting Mechanisms Conclusion Targeting Cost/admin. Targeting Targeting can result in improved equity, effectiveness, approach complexity effectiveness and efficiency (including sustainability) in the use of Geographic Low Low ­ Moderate funds, although the costs of targeting must always be Self-selection Low High weighed against its potential benefits. Many existing targeting strategies for channeling subsidies for basic Means testing High High service provision to the poor have not reached those Proxy means testing Moderate Moderate most in need. OBA, while no panacea, is a promising Community-based Moderate ­ High Moderate mechanism for enhancing the effectiveness of schemes for targeting subsidies to the poor. With OBA, the number of beneficiaries reached being done in an urban gas service project in Armenia, through a scheme should be readily apparent. For its for example (see photo). One approach used by some own projects, GPOBA has been able to verify reports OBA projects includes proxy means testing, in which that approximately 258,000 poor people have been easily observable characteristics such as possession of provided with targeted services for about US$8 mil- indicative assets, for example a dwelling of a certain lion of GPOBA subsidies disbursed. As more outputs size, are used as a proxy for income. Marie Stopes are delivered and more services used, the subsidies International, an NGO that serves as management disbursed and the targeted beneficiaries reached agency in a health project in Uganda, jointly funded by should continue to increase. As we gather more KfW and GPOBA, uses a simple questionnaire assess- information on these projects, including from impact ing assets owned, number and quality of meals per evaluations and other reports which should provide day, and other indicators of potential beneficiaries to evidence on the reach, quality, and sustainability of determine eligibility. the services provided, more analysis can be conducted Community-based targeting relies on collaboration with on whether OBA schemes can enhance the target- the local community or its representatives to help iden- ing effectiveness of infrastructure and social services tify the community members most in need of the service. delivery. Community involvement can increase ownership and reduce the risk of targeting criteria being rejected by the References population in the service area. But community-based targeting may have drawbacks, such as the risk of being Komives, K., V. Foster, J. Halpern, and Q. Wodon. hijacked by special interests. Moreover, this form of tar- 2005. Water, Electricity, and the Poor: Who Benefits from geting can be costly and time consuming, as evidenced Utility Subsidies? Directions in Development series. by an OBA water project in Cambodia. Washington, DC: World Bank. About OBApproaches OBApproaches is a forum for discussing and dis- The case studies have been chosen and presented seminating recent experiences and innovations by the authors in agreement with the GPOBA for supporting the delivery of basic services to the management team, and are not to be attributed to poor. The series will focus on the provision of water, GPOBA's donors, the World Bank or any other af- energy, telecommunications, transport, health and filiated organizations. Nor do any of the conclusions education in developing countries, in particular represent official policy of the GPOBA, World Bank, through output, or performance,-based approaches. or the countries they represent. To find out more, visit www.gpoba.org The Global Partnership on Output-Based Aid Supporting the delivery of basic services in developing countries