Previous studies of Latin American social policies have focused mainly on their origins or on their effects. In Uneven Social Policies, Sara Niedzwiecki explores their implementation. While confirming that recent nationwide conditional cash transfer and primary health care programs in Argentina and Brazil have been more inclusive of the poor, and more resistant to clientelistic manipulation, than previous income support and health care policies, Professor Niedzwiecki finds that conditional cash transfer programs in particular have been obstructed at the implementation stage by political rivalries between national authorities and governors. In the Argentine province of San Luis (123-27) and the Brazilian state of Goias (139-44), governors who were political rivals of national authorities, and who were claiming credit for competing provincial-level cash transfer programs, obstructed the implementation of the national program, reducing its coverage.
Professor Niedzwiecki finds that credit claiming, although a critical issue for the successful implementation of conditional cash transfer programs, does not affect the successful implementation of primary health care policies. It is harder for authorities at a specific level of government to claim credit for primary health care than to claim it for conditional cash transfers. In conditional cash transfers, actual sums of money are transferred conspicuously from the federal, state, or local government to individuals. In a primary health care program, the flow of money is to the largely invisible funding of salaries, facilities, and supplies. Moreover, it easier for many people to judge the value of a cash transfer than the value of a primary health care service (201). Accordingly, Brazil’s Programa Saude da Familia (1994—, now Estratégia Saude da Familia) and Argentina’s Plan Nacer (2005-, now Programa Sumar) were implemented effectively even in localities where rival subnational authorities governed. This analysis of the “dog that didn’t bark” supports Professor Niedzwiecki’s hypotheses about credit claiming and policy implementation.
Unlike the Argentine and Brazilian primary health care programs, the U.S. Affordable Care Act has been relentlessly subverted in several states with Republican governors (208-12). The Affordable Care Act involves health, but in terms of the criteria that inform Professor Niedzwiecki’s analysis, it is, in some respects, more like a cash transfer program than a primary health care program. The interface of enrolled U.S. residents with the Affordable Care Act revolves around premiums paid, claims reimbursed, and noncompliance fees assessed. The interface of Latin Americans served by Brazil’s Estratégia Saude da Familia and Argentina’s Programa Sumar is indirect, through the services rendered by health care personnel and facilities. The subsidy in these primary health care programs is to largely invisible salaries, facilities, and expenses, rather than directly to transfer-eligible voters, whom rival politicians are courting.
Claiming political credit for a cash transfer program is also easier, Professor Niedzwiecki argues, if the program is new. As a program ages, people start to forget which person, party, or level of administration first established it, complicating credit claiming. Over time, moreover, national authorities sometimes decide to share credit explicitly with subnational authorities in exchange for improved collaboration. In Brazil under Dilma RoussefF (2011-16), the national government explicitly offered to share credit with state governments in return for complimentary state-level programs. The state program logo was inscribed alongside the national program logo on recipients’ ATM cards, and periodic accounting to recipients indicated the relative contributions of the national and subnational governments to the transfer (32, 138).
Professor Niedzwiecki argues that the implementation of both primary health care and conditional cash transfer policies also depends in part on the legacies of previous health care arrangements and income support programs. If subnational authorities believe that the new social policies resemble earlier policies that they have already accepted and helped to implement, they tend to opt for collaboration. If, however, subnational authorities (or supporters on whom they depend) believe that the new policies contravene their interests, they may try to obstruct the new policies. In post-1994 Brazil, for example, the new family health teams associated with the Estratégia Saude da Familia had to compete with traditional hospitals and clinics for resources and personnel (167-70, 205-6).
Also, if the new policies have requirements that the old policies lacked, such as the need to monitor school attendance and health visits for conditional cash transfer programs, the new policies will be less likely to function effectively. Local income support policies in Brazil in the 1990s often involved school attendance or health visit conditionalities, so when the national Bolsa Familia program was introduced in 2003, its “conditionalities were not a complete novelty for schools, health clinics, and municipal officials” (156). Before the implementation of Asignación Universal por Hijo in Argentina in 2009, however, clinics and schools had no analogous experience (133) and therefore “did not develop the institutional learning to ensure compliance with health and education conditionalities” (206).
Also thought to be critical to the delivery of cash transfers or public health services is the degree to which subnational authorities may rely for policy implementation on efficient bureaucracies, qualified staff”, and collaborating civil society organizations (16-17, 82). In the city of Porto Alegre, Brazil, neighborhood health councils, established in the late 1980s, relied on civil society organizations to “notify the government if health professionals are late or absent, if they do not conduct house visits, if there is not enough medicine in a particular health facility, or if the health facility is not in adequate condition” (177). Effective territorial infrastructure may also be used for obstruction, however. In the Argentine province of San Luis, where territorial infrastructure is fairly strong, the dissident Peronist governor
Alberto Rodriguez Saa refused to give the federal government officials who were responsible for implementing the Asignación Universal por Hijo program a list of persons qualified to receive transfers from a provincial income support program. Every six months, moreover, participants in the provincial program were required to submit paperwork certifying nonreceipt of Asignación funds. This requirement was a deterrent to enrolling in the national program (124).
The well-designed quantitative analyses in chapter 4 draw on an original time series cross-sectional database of Argentine provinces and Brazilian states. The conjecture explored is that political rivalries between national and subnational authorities will manifest themselves in lower coverage of conditional cash transfer programs, but not of primary health care programs. The quantitative analyses yield findings that support this conjecture. Complementing this quantitative inquiry, and enabling the coding scheme on which the quantitative analyses depend, is the knowledge and experience acquired from 15 months of field research in 8 municipalities, 2 each in the Argentine provinces of Mendoza and San Luis and 2 each in the Brazilian states of Goias and Rio Grande do Sul. This field research produced 235 interviews with politicians and 148 interviews with policy recipients. The interviews are used to good effect to trace the causal processes behind the associations detected in the quantitative analysis.
Uneven Social Policies includes quantitative analyses but is devoted mainly to historical comparisons at the subnational level—across provinces and municipalities within countries, rather than across countries themselves. Subnational comparisons have a couple of advantages over cross-country comparisons. Provinces and municipalities are less heterogeneous than countries in culture and history, so subnational comparisons are less vulnerable than cross-national comparisons to omitted variable bias. Administrative agencies in provinces and municipalities usually use similar methods to measure economic and social indicators, so subnational comparisons are less vulnerable to measurement error than are cross-national comparisons. Relative to cross-country comparisons, subnational comparisons have been underutilized in the literature on social policy.
The dependent variable in the study, the program coverage headcount, depends not only on provision but also on take-up, which depends, in turn, on decisions by the eligible as to whether program benefits are worth the information, compliance, and psychological costs of acquiring them. This is an important topic for additional research. More research is also needed on whether program coverage affects distal outcomes like income poverty, infant mortality, and so on (213).
Professor Niedzwiecki is the first scholar to show rigorously how credit claiming, political alignment, policy legacies, and state capacity can improve or diminish the coverage of cash transfer and primary health care programs in federalized polities. The research reported in Uneven Social Policies advances the understanding of social policy implementation and contributes to meeting the practical challenges of poverty reduction and improvement in population health.
One of Professor Niedzwiecki’s main conclusions, that blurring responsibility for income support policies can contribute to their effective implementation, has implications not only for middle-income developing countries but for federal systems around the world, including those of poorer countries like India and richer countries like the United States. By illuminating forces and conditions that help to bring about, or obstruct, social policy reforms that benefit the previously excluded, Uneven Social Policies helps to identify critical points at which policymakers and the public can intervene to improve the human condition.
