Click here for search results

Newsletter

Site Tools

Research Highlights 2006: Human Development and Public Services

Themes
Research Highlights
 A new approach to evaluating whether project-based aid works
  Social health insurance can reduce household costs under the right circumstances
  Cash incentives to poor families can increase the number of girls in school
Team
Publications in 2006
Forthcoming in 2007

Human development and the delivery of basic services are at the core of Bank’s strategy to improve people’s lives and support countries’ economic development. The research program aims to deepen understanding of the factors affecting human development, improve the analysis of service delivery and related political and economic institutions, and examine the effectiveness of aid in developing countries.

Themes

New research on aid effectiveness began this year. The idea that aid is likely to work better in a good policy environment continues to have common sense appeal, despite disagreement over the existing cross-country evidence.

Research continues on developing measures, models, and methods to understand dimensions of human development that are harder to quantify, such as learning, good health, and vulnerability. Ongoing studies have collected and analyzed data on students’ academic performance and on the nutritional and health status and cognitive development of very young children; measured differences in the schooling of orphans and nonorphans across countries; and examined multiple dimensions of inequalities, including among ethnic groups, a source of inequality that is not often studied due to data limitations. Researchers contributed to the World Development Report 2007: Development and the Next Generation (www.worldbank.org/wdr2007)

Several impact evaluations were undertaken this year—using a range of experimental and nonexperimental methods—to examine how households and individuals respond to changes in incentives embedded in new policies or programs. These include an evaluation of a Bank project in China to strengthen rural health insurance and improve health care provision; an evaluation of Ecuador’s largest social transfer program and its impact on the cognitive and socioemotional development of very young children; and an evaluation of the health and socioeconomic impacts of antiretroviral therapy programs for AIDS patients in six African countries and India.

The effects of broad institutional reforms on service delivery are being investigated. For example, one study evaluates how mass media can promote competition among communities in India to improve the allocation of public resources; another focuses on how international norms about social and economic rights and the enforcement of those rights by courts affect the delivery of basic social services.

Research Highlights

A new approach to evaluating whether project-based aid works

Considerable disagreement persists about the degree to which earmarked aid by donors substitutes rather than supplements local spending for that purpose and, given the scope for fungibility, whether aid really ends up funding what donors intend. Yet these issues have major implications for aid effectiveness and for donor policies on project versus program aid. Aid fungibility has typically been studied using cross-country regressions on macro aggregates over time. For a World Bank-financed rural road rehabilitation project in Vietnam, new research took a novel approach by using impact evaluation methods—difference-in-difference estimation coupled with propensity score matching and weighting methods applied to panel data.

Although the road project stipulated that no new roads be built and that the work focus on earth road rehabilitation, local implementing agents in project areas ignored most of these stipulations and diverted some of the funding to road building and other methods of road rehabilitation instead. However, the study also concludes that project funds did stay in the roads sector, giving a sectoral "flypaper effect." [forthcoming 55]

The research provides cautious support for project-based aid, but more such research is needed in other settings to strengthen this conclusion. It shows that, with care, it is possible to ascertain whether an intervention actually funded what it intended and whether it supplemented local spending—first steps in assessing the impacts of development projects.

Social health insurance can reduce household costs under the right circumstances

Health shocks—in the form of a death, the onset of disability, the arrival of a new chronic illness, even an acute illness—can have devastating effects on households. The incidence of especially large or "catastrophic" out-of-pocket spending can push a household below the poverty line. In Vietnam, a long hospitalization increases annual out-of-pocket medical spending by 130 percent [forthcoming 53]. Earlier work on health shocks in China shows that the poverty head count would have been nearly 15 percent lower in the absence of out-of-pocket spending for health.

Can insurance help protect households? Vietnam’s social health insurance program has reduced the incidence of catastrophic health spending there, but not so China’s urban scheme. Two possible reasons: People who anticipate costly health episodes might be the most eager to join an insurance scheme—the problem of adverse selection. Also, where providers are paid fee-for-service, providers can realize a higher profit margin by shifting insured patients from basic care to high-tech, high-margin services and drugs. To discourage providers from overproviding care, a World Bank-supported project in rural China operated on both the demand-side by making the village-based rural health insurance system more attractive through subsidized premiums, and on the supply-side, by introducing treatment protocols, drug lists, and training programs to reduce "demand inducement" by providers [forthcoming 54]. Research shows that the project reduced catastrophic health spending, especially for drugs, for the whole population. This was achieved by regulating what doctors prescribe—drugs on an approved list and strict treatment protocols—rather than by reducing the share of health care costs households paid out of pocket.

Cash incentives to poor families can increase the number of girls in school

Cash incentives for families to enroll girls in school can work even in low-income countries. An evaluation of a government-run scholarship program in Cambodia launched in 2002-03 to raise the enrollment of girls in secondary school showed that the program functioned much like the conditional cash transfer programs used in many Latin American countries. It transferred cash to families as long as their daughters enrolled in school, maintained a passing grade, and missed no more than 10 days a year without a good reason. The evaluation results have been used to design a new program that is better-targeted to poor families and selects beneficiaries on the basis of clearer rules [134].

The evaluation estimated the impact on school enrollment and attendance—measured through unannounced school visits—using three evaluation approaches: a linear probability model that included available data on the characteristics of students plus school fixed effects; propensity score matching which compared enrollment (or attendance) between matched pairs of recipients and nonrecipients; and regression discontinuity analysis. The estimates suggest that the program has had large positive effects.

The findings suggest that a demand-side incentive can be effective in raising girls’ enrollment and attendance in a low-income country. Evaluations of conditional cash transfer programs typically come from middle-income countries and have shown smaller effects than this study does. The larger effect in Cambodia is perhaps not surprising since enrollment rates tend to be much lower in low-income than in middle-income countries—and the scope for improvement is larger.

Team

Elizabeth King, Research Manager


 




Permanent URL for this page: http://go.worldbank.org/J7029K5CN0