pproaches35540 M a r c h 2 0 0 5 N o t e N u m b e r 0 5 Output-based aid Supporting infrastructure delivery through explicit and performance-based subsidies I ncreasing access to basic infrastructure and social targeted and often inefficient. They tend to benefit services is critical to reducing poverty and achieving better-off, mainly urban customers, while leaving the Millennium Development Goals (MDGs). How- poorer people with no service at all. Subsidized infra- ever, increasing access is a challenge because of the gap structure services may be priced well below what cus- between what it costs to deliver a desired level of ser- tomers are both able and willing to pay. Moreover, vice and what can be funded through user charges. these subsidies can divert scarce public resources from Subsidies have often played a role in funding this gap, other valuable uses to which they might be put. for a variety of socio-economic reasons: There may be The OBA approach is performance-based because limited ability to afford a particular infrastructure ser- it strongly links the payment of service providers to vice, especially among specific disadvantaged groups. their delivery of specified services, or outputs. This pay- The service may have "public good" characteristics ment on outputs transfers performance risk to the ser- making it difficult to collect user charges. And there vice provider. The provider largely self-finances the may be important positive economic externalities service, receiving reimbursement mostly after the verifi- where the benefits of one individual's consumption are cation of successful delivery. By contrast, in other ap- felt much more widely in society, for example in the proaches donors or governments (or both) pre-fund case of health and sanitation. "inputs," so there is commensurately less transfer of However, given the political commitment by a num- performance risk to the service provider. ber of countries to increase aid flows, but at the same OBA can help improve aid effectiveness by: time the mounting concerns of aid effectiveness, it is · Increasing accountability. The transfer of performance critical that subsidies be linked to the actual delivery of risk to the service provider maintains pressure to services, or "outputs." deliver the pre-specified outputs. One way to do this is through Output-Based aid · Improving transparency. Explicit recognition and (OBA), a strategy for using explicit performance- identification of subsidy flows reduces scope for based subsidies to deliver basic services--such as corruption. water, sanitation, electricity, transport, telecommuni- cations, education, and health care--where policy con- · Increasing value for money. Competitive award of OBA cerns would justify public funding to complement or subsidies, together with the transfer of performance replace user fees. Two key features distinguish OBA risk to the service provider, can increase the value subsidies from some other forms of publicly funded for money. subsidy: OBA subsidies are explicit, and they are per- · Reducing economic distortions. Explicit recognition and formance based. identification of subsidies can help reduce the OBA interventions are explicit because they ensure economic distortions that subsidies tend to intro- explicit recognition of why the subsidy is being pro- duce. vided, who is receiving the subsidy and who is providing Further, OBA approaches are a mechanism to imple- it, and what is being subsidized--both the activity and ment public-private partnership (PPP): effective use of the financial sums involved. Historically, infrastructure donor funding and the public budget via OBA-type service delivery in both industrial and developing coun- tries has often involved "implicit" rather than explicit This OBApproaches is derived from GPOBA's OBA Working Paper of subsidies. These implicit subsidy approaches are poorly the same title, No. 4, March, 2005. Supporting the delivery of basic services in developing countries oba case study 4.28 1 4/28/05, 3:26 PM pproaches mechanisms can mobilize private capital and efficien- type 2 outputs--or even both--might be appropri- cies for increased service delivery to the poor. ate. But if the scheme took place in, say, one of the Central Asian republics, where most customers are Key principles of design Each application of OBA will differ at least slightly from others. Since 2000, when OBA was officially Box 1 To target or not to target? Two OBA schemes launched, some 30 projects (mostly World Bank) have in water been developed that include an OBA-type component. Targeting subsidies can be difficult, typically be- But many of these projects are at an early stage of ex- cause of the costs involved. Indeed, in many situa- ecution. So evidence supporting the use of OBA ap- tions it could be argued that the costs of targeting proaches as a tool for improving aid effectiveness is simply are not worth it. Still, as examples from still being collected, as are lessons for designing such Cambodia and Paraguay show, some measures can approaches. But wherever OBA is applied, there are ar- help ensure that subsidies go to the intended re- guably three key principles of design to consider: cipients. · Ensuring intelligent subsidy design. This involves sharp- In a water and sanitation project in provincial ening the targeting of subsidies (where desired) so and peri-urban areas of Cambodia, it was decided that they reach the intended beneficiaries and not that OBA subsidies would be targeted directly to others (box 1); providing just the right amount of individual households: poor households deemed subsidy to effect the desired response and minimize eligible would receive a connection (from a com- market distortions (box 2); and maximizing the petitively selected private operator) that would be sustainability of any subsidy intervention that does funded through a grant from the International De- not hinder long-term development or prevent the velopment Association. Once four pilot OBA towns achievement of goals in other areas. were identified, the local communities played a big · Developing a performance regime. Performance risk can part in selecting the households that would receive be transferred to service providers by introducing connection subsidies. A household survey was de- subsidies within a competitive regime or by devel- veloped within the communities to collect data, oping specific, performance-based subsidy con- and village representatives and commune council tracts. A big challenge in designing performance- members determined poverty criteria. Based on based regimes is determining what outputs to link these criteria and the results of the survey, the the payment of subsidies to--that is, which outputs communities themselves identified the poor house- will deliver the desired outcomes. For example, if holds that would receive the subsidy. An indepen- the OBA scheme depicted in figure 1 took place in dent consultant later randomly verified the a rural part of Sub-Saharan Africa where rates of selection. Of the 13,000 households in the four access to electricity are very low, either type 1 or towns, the 3,000 poorest households would re- ceive a subsidized connection. Figure 1: Outcomes and outputs in a hypothetical OBA An OBA project in Paraguay relied on aguateros scheme in electricity (small-scale water entrepreneurs) and local con- struction companies to provide water services to poor rural communities. The providers, selected Desired outcome through a competitive process, would receive part Increased household access of their payment from users and the other part to electricity services from a World Bank­financed subsidy fund. In the Paraguay project, rather than individual house- holds, unserved rural areas and small towns where most residents are poor were selected to receive the subsidies. In addition, the very poorest custom- Type 1 Type 2 ers were given the option to provide labor in dig- Payments linked to connection Payments linked to supply of 100kWh ging trenches as part of their payment to the of new customers to grid of electricity to each household per month service provider. Possible contractual outputs Source: Mumssen 2004 and Drees 2004. Supporting the delivery of basic services in developing countries oba case study 4.28 2 4/28/05, 3:26 PM pproaches donor-funded OBA support will very likely form a Box 2 Getting the design right for subsidies smaller subset of OBA schemes, reflecting the priorities of the donors involved. Finally and importantly, tack- The size and type of subsidy required to fund ser- ing an OBA scheme onto a bad project--one that is vices in the absence of user fees can vary greatly. poorly designed--will not make it a good one. For simplicity, three types of situations can be dis- tinguished: · Those requiring a one-time subsidy, often to Other design considerations support initial connection costs. All OBA schemes thus far have involved private provid- · Those requiring a transitional subsidy, for ers and nonincumbents. But in many situations infra- example, to smooth an increase in tariffs to a structure services will be provided by an incumbent or level that recovers full costs (see figure). a public utility, and it will not be feasible--often for po- · Those requiring an ongoing subsidy, for example, litical but sometimes also for economic reasons--to in- to support a lifeline tariff. stitute an alternative model, at least in the short term. Clearly, introducing competition where there is only one supplier of the desired services is more problem- Transitional subsidy arrangement atic. More analysis as well as OBA pilots are needed to demonstrate the potential gains of OBA with incum- bents or public providers. But the types of measures Transition Subsidy that might be used to help mitigate problems arising from a lack of competition are those commonly used User contribution by regulators to improve efficiency where there are in- Tariff cumbents, including identifying and contracting out contestable activities, employing yardstick or bench- mark approaches, imposing effective regulatory Time mechanisms, and introducing the threat of competi- tion or a Swiss challenge. already connected to the national or municipal electricity grid, tying subsidies to a type 2 output Box 3 Introducing a competitive dynamic in an OBA would probably be much more appropriate. scheme in telecommunications · Creating a competitive dynamic or tension. Creating competitive pressures is important in ensuring value In OBA schemes the potential for improving aid ef- for money, by minimizing the subsidy required or fectiveness--by achieving value for money and tar- maximizing the benefit from a given subsidy geting intended beneficiaries--can be further amount (box 3). It will be critical in schemes improved by introducing a competitive dynamic in involving public and private sector incumbents, awarding or determining the subsidy. In a rural where competition may need to go beyond subsidy telecommunications project Peru used a least-sub- award, even if this competition is "threatened" sidy bidding approach, in which telecommunica- rather than actual. tions operators bid for the minimum government subsidy they required to deliver pay phone service in targeted areas. Winning bidders received a non- In reality, creating "perfect" OBA schemes that in- exclusive concession defining their rights and obli- corporate all these elements will be extremely difficult. gations. Results thus far show that the investment Nonetheless, they can be seen as a starting point when mobilized has been much greater than the subsidy developing schemes, even when real-world practicali- provided. Moreover, in a pilot project the winning ties require a step back from the theoretically desirable. bidder requested a subsidy 41 percent lower than Moreover, OBA is only one way of applying subsidies; that estimated by the sector regulator and 74 per- there may be other legitimate approaches. Nor should cent lower than a previous amount requested by it be expected that just because something fits the the incumbent provider. broad definition of OBA, it should be entitled to spe- cific donor support. The types of projects qualifying for Source: Cannock 2001. Supporting the delivery of basic services in developing countries oba case study 4.28 3 4/28/05, 3:26 PM pproaches Another key design challenge is mitigating payment ment and monitoring systems clearly will attract and risk. For an infrastructure project supported by out- mobilize the private sector. In addition, well-designed put-based aid, cash flows are subject not only to risks OBA approaches can improve the performance and related to tariff payments by customers but also to governance of public utilities and help strengthen the risks related to subsidy payments by the government or partnership between communities and nongovernmen- government entity. In many countries government pay- tal organizations. However, more projects are needed ments are considered unreliable and so are assigned a to demonstrate the potential results--particularly on a low credit rating by financial markets and investors. In larger scale. Designed properly, OBA can become a key these cases the quality of OBA payments needs to be mechanism in improving the delivery of infrastructure enhanced--that is, lifted to a higher level to become services to the poor, helping to achieve the Millennium creditworthy--such as by a partial risk or partial credit Development Goals. guarantee. Administration and monitoring of schemes is also References critical. The disbursement of funds must be considered Cannock, Geoffrey. 2001. "Expanding Rural Telephony: reliable, with mechanisms to mitigate payment risk. Es- Output-Based Contracts for Pay Phones in Peru." crow accounts and offshore banking might be consid- In Penelope J. Brook and Suzanne M. Smith, eds., ered. Monitoring arrangements--for example, to Contracting for Public Services: Output-Based Aid ensure that outputs are satisfactorily delivered--need and Its Applications. Washington, D.C.: World Bank. to be made, taking into account local capacity. Mumssen, Yogita. 2004. "Output-based Aid in Cambo- dia: Private operators and local communities help Next steps deliver water to the poor." OBApproaches No. 01. Subsidies are back on the table as a legitimate policy Washington, D.C.: World Bank. tool for infrastructure development. Full evidence of Drees, Franz, Jordan Schwartz, and Alexander the effectiveness of OBA as an aid instrument still Bakalian. 2004. "Output-based Aid in Water: needs to be collected, but initial results look positive. Lessons in Implementation from a Pilot in Para- OBA is a means to enhance public-private partnership: guay." Viewpoint No. 270. Washington, D.C.: Better-designed subsidy mechanisms with credible pay- World Bank. About the Global Partnership on Output-Based Aid The Global Partnership on Output-Based Aid How can GPOBA help? GPOBA can assist in the (GPOBA) is a multidonor trust fund administered design and development of pilot OBA projects, by the World Bank. Its purpose is to fund, demon- and can help identify and disseminate emerging strate and document output-based aid (OBA) ap- knowledge on issues related to OBA approaches proaches to supporting the sustainable delivery of through studies, publications, workshops, and basic services to those least able to afford them conferences. GPOBA can also contribute to the and to those without access to such services. funding of subsidized payments for the provision of services under OBA schemes. Global Partnership on Output Based Aid World Bank Mailstop: H3-300 600 19th Street, NW Washington, DC 20433, USA To find out more, visit www.gpoba.org The Global Partnership on Output-Based Aid Supporting the delivery of basic services in developing countries oba case study 4.28 4 4/28/05, 3:26 PM