September 2008 · Number 135 A regular series of notes highlighting recent lessons emerging from the operational and analytical program of the World Bank`s Latin America and Caribbean Region. 46983 PERU Making Accountability Work ­ Lessons from RECURSO Daniel Cotlear Even the best accountability rules will not lead to to distinguish three phases in the evolution of this improvements in the quality of public services if citizens portfolio. The first focused on coverage, the second on do not know what they should expect from their schools improving rules of accountability, and the third on making and clinics. the accountability rules enforceable by fostering the establishment of standards by which citizens can measure A mother who does not know how a well-educated second the quality of public services. grader should read, for example, has no reason to demand more from her child's teacher. A father who does not know Hyperinflation and political violence in the late 1980s and how much his 6-month-old should grow in the coming year early 1990s set the stage for the first phase: Bank programs has no reason to demand better primary care services. sought to repair damage to physical infrastructure such as schools, clinics, hospitals, and rural sanitation services, to Such was the experience of Peru in the early 2000s. High train new teachers and nurses and to make available key levels of coverage had been reached, abstract transparency inputs such as textbooks and basic medication. and accountability rules were put in place, but the quality of public services remained The second phase of Bank engagement low. The lack of quality benchmarks in human development in Peru coincided prevented parents from demanding better with the global economic slowdown of nutrition and education for their children. 1997. In Peru, the negative effects of this Their expectations for their children's slowdown were amplified by another development remained low, and parents political crisis ­ the exposure of evidence were not effective agents of change. of corruption, political manipulation of public expenditures, and increasing heavy- This note describes the experience of handedness. This political tumult forced RECURSO, a successful effort to create the president to resign in November 2000, the missing link: high expectations. The only five months after his controversial RECURSO program provided the impetus re-election to a third term. A Transition for the establishment of easy-to-understand Government carried the country to new standards by which citizens can measure elections. The crisis contributed to a the quality of basic public services. It enables parents growing appetite among Peruvians for transparency and to take advantage of accountability mechanisms and, the Bank assisted the Government in developing a plan ultimately, ensure better health and education services to strengthen transparency and accountability in public for their children. This note describes the genesis, expenditure, with an emphasis on social expenditure. implementation, and impact of RECURSO. Once the Transition Government took office, the Bank rapidly agreed with the Transition Government on a series World Bank engagement in human development in Peru: a of Programmatic Social Reform Loans (PSRL), the first brief history being signed with the Transition Government in 2001; The Bank has had an active human development another three were signed in the following four years with portfolio in Peru since 1990. Looking back, it is possible the administration of President Alejandro Toledo. Much of the emphasis of these PSRL loans was on ratings, or of health outcomes, such as the nutritional backing legislation and regulations attempting to impact of nutrition programs, all suggest poor average strengthen the three sides of the accountability triangle quality of service. Quality is also highly unequal and (as described in the World Development Report 2004). highly unpredictable, with the poor receiving particularly The goals were (1) to strengthen voice through support bad and uneven services. for a public information law, implementation of an Internet site with user-friendly budget information about A special challenge for policy makers is the stickiness social programs, and design of participatory budgeting; of this low-quality equilibrium, which results from (2) to strengthen client power by providing support for a mediocrity pact that binds the hands of frontline programs that transferred the management of primary care providers (teachers, nurses, doctors), their local-level clinics to community groups, programs that gave parents supervisors, middle managers (regional or municipal school oversight capabilities, and the development of toll- authorities), and Ministerial authorities. The knot is free complaint lines; and (3) to strengthen the compact tied in a way that penalizes any party that attempts through the implementation of performance contracts, quality improvements. The low-quality equilibrium the development of output-linked payment systems, and results from a combination of complex rules based on the improvement of the central government's monitoring legislation, norms and agreements with unions, practices systems for the social sectors. and ­ crucially ­ in low expectations of performance. The implication of the stickiness is that it is impossible By the mid-2000s, it became clear that while many of to achieve change through incremental improvements. these interventions successfully increased stakeholder A shock is needed to free the system from low-quality participation in social services, amplified the government's equilibrium; only once the mediocrity pact is broken will capacity to monitor and influence providers' behavior, and incremental measures have an effect. raised public expenditure and salaries, the changes were not enough to generate an improvement in quality. It was The analysis found that while users and community apparent to analysts, government officials, media, and the leaders contributed to the expansion of coverage of public that the reforms' primary objectives were not being services, they were not active in improving the quality achieved ­ the quality of education and health services was of these services. Parents built schools, hired teachers not improving. In 2005, the Bank agreed with the Prime and lobbied for improvements to school infrastructure. Minister to undertake analytical and advisory work (known Community leaders built clinics and actively equipped in the Bank as AAA) to identify the source of the problems and staffed the clinics. Local leaders became known and identify options for change, with the goal of presenting for their efforts to establish schools and clinics. Why these to the public during the 2006 presidential campaign. weren't these stakeholders also pressing for improved quality? The analysis found that while it is easy to see While transparency and accountability improved, service and measure coverage ­ every parent knows if there is a quality remained low. Why? The third phase of Bank school in town ­ it is difficult to see or measure quality. intervention attempted to answer this question with a The low quality of services is further protected by a veil cluster of programs jointly known as RECURSO ­ the of low expectations. RECURSO proposed to lift that veil Spanish acronym for Accountability for Social Reform by providing benchmarks by which poor people could (REndicion de CUentas para la Reforma SOcial). measure the quality of the public services they receive. Explaining the persistence of low quality Rules of accountability will not lead to improvements Peru has experienced huge increases in the coverage in quality as long as quality remains immeasurable. To of education, health care and some social assistance make accountability work, there is a need for standards ­ programs. Comparisons to countries with similar income accountability metrics. In the long run, detailed standards levels in LAC and elsewhere indicate that Peru has higher are needed; in the short run, the team looked for simple coverage in primary, secondary, and tertiary education. standards that could also be used to quickly change This is also the case for several primary health care culture. The next section describes that experience. programs, such as the administration of immunizations, and several social assistance programs. In contrast to RECURSO: defining quality, creating high expectations these achievements in coverage, however, service quality While coverage is concrete and therefore easy to see is very low. Measures of education quality, such as PISA and measure, the quality of education or health services is an abstract concept. Users find it hard to measure 2 · September 2008 · Number 135 quality, so political pronouncements about "poor quality" as to how to measure this generate minimal enthusiasm among parents and voters. with any watch. The video3 Only specialists participate in discussions about quality, challenges parents to find out and they tend to make the discussions increasingly how well their children are complex, complicating the participation of users and reading and tells them that they community leaders. have the right to demand a good education. The low quality trap is largely created by users' low expectations. Parents don't know what to expect The nutrition video compares from schools and nutrition programs. Since there are two neighboring villages in a no benchmarks by which to measure their children's poor rural area in Peru. One achievements, parents believe their children are doing village benefited from a health well as long as they get passing grades and show some clinic with a good nutrition improvement. If they knew that their children take five support program; this program years to learn to read at a standard that should be achieved reduced chronic malnutrition from 80 percent to 20 percent after one or two years, they might demand change ­ but in four years. The other has a conventional health clinic they have no way of knowing. that operates in a village where malnutrition remains at 80 percent. The video compares the sizes and attitudes Similarly, the studies found that the extremely high of children in the two villages. The video4 then defines a prevalence and persistence of chronic malnutrition (25 standard: children should grow 24 centimeters in their first percent of under-5 children nationwide, 80 percent in poor year and 12 centimeters in their second. The video asks areas) is also due to low expectations. Parents, nurses, parents if they know how well their children are growing. teachers and society at large expect children of the poor to be small, quiet, unhappy, with little appetite for fun, The video created a strong impression on the government. adventure or learning and with fragile health, so no real The Prime Minister said he repeatedly tried to raise the effort is made to make changes. profile of malnutrition but the media and public opinion were uninterested because the problem was largely invisible The findings of the RECURSO AAA were published in and poorly understood. To take advantage of the video, he three books1. A key recommendation was that stakeholders asked the Bank to produce TV and radio spots to be shown be provided with standards or other instruments that during regular TV hours. allow them to understand and demand quality. In order to [see http://www.worldbank.org/lacnutrition]. illustrate how this could be done, RECURSO produced a number of instruments for the general public and, more Conclusion specifically, directed to the parents of poor children. These The lessons emerging from RECURSO highlight the included three videos, a radio theater series, and numerous importance of providing parents - and all agents involved brochures and posters, many of which have been produced in monitoring service delivery - - with simple, clear and in Spanish, Quechua and Aymara. The videos in particular measurable expectations about how their children should had a strong impact on public opinion. learn, grow, and be healthy. This requires that clear and simple standards are produced for education and The education video demonstrates poor education quality health and that all children are measured periodically and by showing children who cannot read or struggle to compared with these standards. Without this, any rules of read. These dramatic scenes are followed by images of accountability will be difficult to enforce as instruments high-quality education ­ poor rural children of the same to improve results. The Recurso experience also shows age who can read fluently, sometimes in Quechua and that service providers are likely to resist the development Spanish2. The video then defines a standard ­ "children of such standards and the use of periodic tests - precisely finishing the second grade should be able to read 60 words because it increases the pressures of accountability. The per minute" ­ and the video gives clear, simple instructions impact of RECURSO is described in Box 1. 1 A New Social Contract for Peru: An Agenda for Improving Education, Health care, and the Social Safety Net. The World Bank. 2006. (http://go.worldbank.org/ 7755UFLMS0) Peru.; Toward High-Quality Education in Peru : Standards, Accountability, and Capacity Building : The World Bank. 2007 (http://go.worldbank. org/7GZHOY8AB0); Social Safety Nets in Peru. The World Bank. September 2007 (http://go.worldbank.org/GP9OHHD940) 2 The large variance found in the lower tail of the distribution means that there are a few schools and clinics that are achieving good results for the poor. The vid- eos sought these examples to lend credibility to the optimists view stating that even in the most adverse conditions, good quality of service will produce good results. This view is needed to raise expectations and produce change. 3 http://www.youtube.com/watch?v=-BxL1aqb6mY September 2008 · Number 135 · 3 4 http://www.youtube.com/watch?v=mJieb2Xgt9U BOx 1 - IMPACT On politics and policy: · Six months before the presidential election, the RECURSO team held a half-day workshop with each of the four frontrunners. Three months later, all parties presented their plans for government, and a summary prepared by TRANSPARENCIA found that each of the four parties included the main RECURSO recommendations in their respective plans for the social sectors; · In his inauguration speech, President Alan Garcia announced that his government would introduce universal test- ing for second-grade students (a central recommendation of RECURSO and a break with the past, as the Education Ministry always opposed this); in several speeches President Garcia later advised mothers to "become aware of how many words per minute their children were able to read;" · Universal testing of students began in 2006 and is now an annual policy; beginning in 2008, feedback will be pro- vided to all schools and parents; · Congress approved new legislation to provide pay incentives to teachers who approve evaluations; teachers went on strike and, for the first time ever, public opinion sided with the government and the legislation passed; · The Ministry of Health approved a new technical norm that includes the nutrition standards as an important factor of the health communication package; and · The national CRECER program has included the nutrition standards in the communication material that will be distributed to communities for health and nutrition-counseling activities. On Bank programs: · The new Country Partnership Strategy includes an accountability cluster with various activities recommended by RECURSO; · One new activity is a series of Results and Accountability DPLs with strong emphases on targets, testing and man- agement of parental expectations in education, nutrition and health; and · RECURSO evolved into a multi-year programmatic intervention that will accompany the Bank's program for the duration of the CPS. Creating energy in civil society · Several NGOs and private foundations are using the RECURSO framework; · Youth NGOs are making agreements with mayors to attempt to improve the fluency of reading in their districts (and measure it!); · Solaris ­ an international NGO managing six public schools with a system similar to Fe y Alegrķa ­ used the RECURSO video to motivate the development of a plan that measured fluency, developed targets, and proposed instruments; the NGO saw significant success in four months (a video documenting this experience is available at: www.youtube/worldbank · A private-sector foundation is considering a competition to provide prizes to rural teachers whose second-grade students read at the level of the standard; · The education video is being used to motivate discussion in Mexico, Honduras, Bolivia, South Africa and India; · The national CRECER program, together with the Nutrition Initiative in Peru, will be distributing approximately 4,000 copies of the nutrition video to local health centers; · Private radio programs have developed communication campaigns based on the main messages of the nutrition video; and · The Nutrition video has been used as a tool to improve the visibility of chronic malnutrition and to create discus- sion in Bolivia, Ecuador, Nicaragua, Guatemala and Mexico. Adaptation of the nutrition video to additional coun- tries is under way. About the Authors Daniel Cotlear is the Lead Economist in the Human Development Department of the Latin America and the Caribbean Region of the World Bank "en breve" is produced by the Knowledge and Learning Team of the Operations Services Department of the Latin America and the Caribbean Region of the World Bank - http://www.worldbank.org/lac · September 2008 · Number 135