Summary of the impact

Marco Manacorda's research on social assistance programmes has been a major influence on the design and evaluation of a flagship poverty alleviation initiative in Uruguay known as PANES. Manacorda's work has:

1. enhanced the programme itself by supplying an analytical underpinning for the targeting of resources and a scientific basis for evaluation of its effects;
2. influenced the government's decision to scale up the programme by providing credible evidence of its impact on beneficiaries;
3. shaped the public debate on the design and effects of social assistance measures more generally, both in Uruguay and internationally; and
4. affected administrative practices in the public sector by stressing the importance of collecting good data and using it effectively to improve service provision.

Underpinning research

Over the past decade a number of low-income countries have adopted new social assistance programmes, typically involving cash transfers to beneficiaries. One such programme was Uruguay's Plan Atención Nacional a la Emergencia Social (PANES), a temporary poverty alleviation initiative that from 2005-07 covered 14% of the country's population.

Manacorda has been deeply involved in both the design and evaluation of PANES. His initial contribution (together with colleagues at UC Berkeley and the Universidad de la República in Montevideo) was to the programme's resource targeting mechanism. He later used data from PANES to study the intended and unintended consequences of cash transfer programmes, contributing to a lively debate about the optimal design of social assistance measures in developing countries.

Crucial for the success of the project was establishing good working relationships with and between an array of governmental and non-governmental organisations, including the Uruguayan Ministry of Social Development, Ministry of Public Health, and social security administration (ie, the Banco de Previsión Social); as well as the Inter-American Development Bank (IADB), which provided essential financial support. The research involved both primary data collection and the use of large amounts of detailed administrative data from various government sources.

Initial research focused on the mechanism used to select programme beneficiaries based on an eligibility score. The nature of the scoring rule allowed for an evaluation of the effects on recipients using a regression discontinuity design. This and other empirical strategies were able to demonstrate changes in several aspects of the behaviour and well-being of cash-transfer recipients. For example, there was found to be a significant increase in the proportion of beneficiary households that favoured the incumbent government relative to its predecessor (see reference #1 in Section 3 below).

A subsequent study, funded by the IADB under its project "Improving Early Childhood Development in Latin America," combined PANES data with official birth records in order to show that the poverty alleviation initiative increased weight and other measures of health at birth among children who received cash transfers in utero via their mothers (see reference #2 in Section 3 below).

A third strand of research, again funded by the IADB under its project "Social Assistance and Labor Supply in Latin America and the Caribbean", linked social security and PANES data (see reference #3 in Section 3 below). This work confirmed that subjecting the cash transfers to means testing — under which they are withdrawn when (declared) income exceeds a certain level — created a disincentive to accept formal employment and led would-be beneficiaries to instead take on casual work invisible to the welfare system.

References to the research

Details of the impact

Manacorda has been working in Uruguay since the beginning of the last decade. Following a change of government in the national elections of 2005, he was invited to join colleagues at the Universidad de la República in helping (pro bono) the new government design and evaluate PANES.

Over the intervening years, Manacorda's research on social assistance policy has had an impact in four main areas.

1. Resource targeting.

To address the challenge of alleviating poverty in a population not fully covered by official records, as well as to allow a quasi-experimental approach to evaluating the programme, Manacorda and his colleagues proposed a carefully designed resource-targeting mechanism. This proposal was presented formally by Manacorda to the Minister for Social Development at a 2006 public conference sponsored by the Presidency of the Republic of Uruguay (see source #6 in Section 5 below), and was later adopted by the government for PANES and some other social welfare programmes.

The targeting mechanism functions by computing for each household a poverty index based on verifiable information collected at the time of application for social assistance. Only those households scoring above a pre-defined threshold on the index are deemed eligible for transfer payments. This mechanism has been judged a success in identifying genuinely impoverished Uruguayans. An independent appraisal by the World Bank described PANES as "one of the most successful income transfer programmes in terms of targeting effectiveness," with a higher percentage of beneficiary households in the poorest quintile of the population "than what was achieved in other programs in Latin America and around the world" (see source #4 in Section 5 below; executive summary, point #14).

2. Programme evaluation.

As mentioned above, the resource-targeting mechanism used in PANES was constructed in a way that allowed the programme to be evaluated quasi-experimentally. Analysis of the data revealed clear evidence of success both in poverty alleviation per se and in improvements to the well-being of beneficiaries (whether impoverished or not). These findings were communicated in technical reports (see source #3 in Section 5 below) and through discussions with officials in the Ministry of Social Development, and they were an important factor in the Uruguayan government's decision to scale up PANES and later to adopt the Plan de Equidad (a comprehensive reform of the tax and benefit system introduced in 2008).

3. Welfare policy debates.

Manacorda has actively disseminated the results of his research to a variety of audiences. For example, in April 2011 he participated in a policy dialogue on "Labor Markets and Social Security" at the IADB offices in Washington, DC. This event — attended by 21 ministers and officials from Latin American countries — offered the opportunity to inform policymakers from across the region about the labour supply distortions that can be an unintended consequence of social assistance policies. Indeed, these effects proved to be significant in the Uruguayan case, leading the government to reduce the strictness of means testing for some entitlements and benefit programmes.

In addition to the IADB, Manacorda has presented his research findings at the World Bank and at the OECD's Directorate for Employment, Labour and Social Affairs. His analysis of the effects of PANES has been summarised in a number of policy documents from the international institutions already mentioned as well as from the UK Department for International Development and the UN Development Programme (see sources #7-8 in Section 5 below). He has also given several interviews on social assistance programmes to the local media (eg, Radio el Espectador and Televisión Nacional Uruguaya).

4. Administrative practices.

A side benefit of the PANES project was the establishment of good relationships with and cooperation between various public sector bodies, including the ministries of social development, public health, the social security administration and the university. This has materialized in a formal protocol of data exchange, which has led to the sharing of best practices for data collection, handling, storage, and release. More broadly, the research carried out using PANES data gave Uruguayan policymakers a compelling demonstration of the power of social science to answer questions about the world when provided with a good quality dataset and an astute empirical strategy.

Sources to corroborate the impact

The following individuals can address the impact of Manacorda's work on public policy:

1. Former Minister of Social Development, government of Uruguay; on the impact on PANES targeting criteria and the redesign of income conditionality in the Plan de Equidad.
2. Senior Labor Market Specialist, Labor Markets and Social Security Unit, Inter-American Development Bank (IDB); on analysis of the labour market effects of PANES and its influence on social protection policies across the region.

Technical reports prepared for the Uruguayan government (under the auspices of the Instituto de Economía, Universidad de la República, Montevideo):

3. (a.) "Una propuesta metodológica para la evaluación del impacto del Plan de Atención Nacional a la Emergencia Social (PANES)." (With V. Amarante, R. Arim, and A. Vigorito.) 2006. (b.) "Informe final de la evaluación intermedia de impacto del PANES." (With V. Amarante, G. Burdín, and A. Vigorito.) 2008. (c.) "Informe final de la evaluación de impacto del PANES." (With V. Amarante, G. Burdín, M. Ferrando, A. Vernengo, and A. Vigorito.) 2009.

The World Bank's appraisal of targeting effectiveness under PANES is discussed in:

4. "Income transfer policies in Uruguay: closing the gaps to increase welfare." World Bank Report No. 40084-UY, October 2007. Available at: http://goo.gl/VOlupS
5. "Building on experience: improving social protection in Uruguay and the Plan for Social Equity." World Bank "En Breve" No. 132, July 2008. Available at:
   http://goo.gl/FSSZcs

Other corroborating sources include:

6. "Pautas de evaluación metodológica del impacto del PANES." Press release from the Presidency of the Republic of Uruguay. Available at: http://goo.gl/nKwcyA
7. "Cash transfers: literature review." DFID Policy Division, April 2011. [See discussion of politics of conditional cash transfers on p. 43.] Available at: http://goo.gl/nx7v8b
8. "Poverty reduction and equity reading list." World Bank, June 2011. [See p. 4. Reading list intended "to globally disseminate relevant work among academics, practitioners and civil society."] Available at: http://goo.gl/MtkZgj
9. "Estado asiste al 31% de la población." El Pais, 14 November 2010. [Local media coverage: "Las conclusiones de la investigación son coherentes con los resultados de un estudio...de Andrea Vigorito, Marco Manacorda y Edward Miguel, que concluye que los beneficiarios del PANES `son significativamente más proclives a apoyar al gobierno que los no beneficiarios'."] Available at: http://goo.gl/2CnLTf