Latest research findings
Radio's impact on preferences for patronage benefits vs. education and health provision
Citizens in developing countries support politicians who provide patronage or clientelist benefits, such as government jobs and gifts at the time of elections. Philip Keefer and Stuti Khemani  ask whether access to mass media that broadcasts public interest messages can shift citizens' preferences for such benefits. They look at the impact of community radio on responses to novel survey ‘vignettes’ that make an explicit trade-off between political promises of jobs for a few versus public services for all. They identify the impact of community radio through a natural experiment in the media market in northern Benin, which yields exogenous variation in access across villages. Respondents in villages with greater radio access are found to be less likely to express support for patronage jobs that come at the expense of public health or education. Keefer and Khemani also find that gift-giving is not necessarily traded off against public services; correspondingly, radio access does not reduce preferences for candidates who give gifts. The pattern of results is consistent with a particular mechanism for radio's impact: increasing citizens' demand for broadly delivered health and education and thereby shaping their preferences for clientelist candidates.
Does aid increase growth? Evidence from a quasi-experiment
The literature on aid and growth has not found a convincing instrumental variable to identify the causal effects of aid. Sebastian Galiani, Stephen Knack, Lixin Colin Xu and Ben Zou  exploit an instrumental variable based on the fact that since 1987, eligibility for aid from the International Development Association (IDA) has been based partly on whether or not a country is below a certain threshold of per capita income. The paper finds evidence that other donors tend to reinforce rather than compensate for reductions in IDA aid following threshold crossings. Overall, aid as a share of gross national income (GNI) drops about 59 percent on average after countries cross the threshold. Focusing on the 35 countries that have crossed the income threshold from below between 1987 and 2010, a positive, statistically significant, and economically sizable effect of aid on growth is found. A one percentage point increase in the aid to GNI ratio from the sample mean raises annual real per capita growth in GDP by approximately 0.35 percentage points. The analysis shows that the main channel through which aid promotes growth is by increasing physical investment.
Does World Bank lending raise the quality of economic policy?
Lodewijk Smets and Stephen Knack  investigate the impact of World Bank development policy lending on the quality of economic policy. They measure the quality of economic policy using the World Bank's Country Policy and Institutional Assessments of macro, debt, fiscal and structural policies, and focus on policy loans targeted at improvements in those areas. Smets and Knack find that the quality of policy increases, but at a diminishing rate as the cumulative number of policy loans increases. Similar results hold for the cumulative number of conditions attached to policy loans, although quadratic specifications indicate that additional conditions may even reduce the quality of policy beyond some point.
Son preference, fertility and family structure: Evidence from reproductive behavior among Nigerian women
Strong boy-bias and its consequences for young and unborn girls have been widely documented for Asia. Annamaria Milazzo  finds that in Nigeria parental gender preferences affect fertility behavior, and shape traditional social institutions, with negative effects on adult women's health and well-being. Compared to women with first-born sons, women with first-born daughters have (and desire) more children, and are less likely to use contraceptives. Women with daughters among earlier-born children are also more likely to have shorter birth intervals, a behavior medically known to increase the risk of child and maternal mortality. Moreover, they are more likely to end up in a polygynous union, to be divorced, and to be head of the household. The preference for sons is also supported by child-fostering patterns in which daughters are ‘substitutes’ for foster girls, while the same does not hold for sons and foster boys. These results can partly explain excess female mortality among adult women in Sub-Saharan Africa.
Progress toward the health MDGs: Are the poor being left behind?
Adam Wagstaff, Caryn Bredenkamp and Leander R. Buisman  look at differential progress on the health MDGs between the poor and better-off within countries. Their findings are based on original analysis of 235 Demographic and Health Surveys and Multiple Indicator Cluster Surveys, spanning 64 developing countries over the period 1990-2011. Five health status indicators and seven intervention indicators are tracked for all the health Millennium Development Goals. In most countries, the poorest 40 percent have made faster progress than the richest 60 percent. On average, relative inequality in the MDG indicators has been falling. However, the opposite is true in a sizable minority of countries, especially on child health status indicators (40-50 percent in the cases of child malnutrition and mortality), and on some intervention indicators (almost 40 percent in the case of immunizations). Absolute inequality has been rising in a larger fraction of countries, and in around one-quarter of countries the poorest 40 percent have been slipping backward in absolute terms. Despite reductions in most countries, relative inequalities in the MDG health indicators are still appreciable, with the poor facing higher risks of malnutrition and death in childhood and lower odds of receiving key health interventions.
What a difference a state makes: Health reform in Andhra Pradesh
In the mid-2000s, India began rolling out large-scale, publicly-financed health insurance schemes mostly targeting the poor. Sofi Bergkvist, Adam Wagstaff, Anuradha Katyal, Prabal V. Singh, Amit Samarth and Mala Rao  describe and analyze Andhra Pradesh's ‘Aarogyasri’ scheme, which covers against the costs of around 900 high-cost procedures delivered in secondary and tertiary hospitals. Using a new household survey, they find that 80 percent of families are eligible, equal to about 68 million people, and 85 percent of these families know they are covered; only one-quarter, however, know that the benefit package is limited. Bergkvist et al. also find that, contrary to the rules of the program, patients incur quite large out-of-pocket payments during inpatient episodes thought to be covered by Aarogyasri. In the absence of data and program design features that would allow for a rigorous impact evaluation, the authors make a comparison between Andhra Pradesh and neighboring Maharashtra over an eight-year period spanning the scheme's introduction. During this period, Maharashtra did not introduce any at-scale health initiative that was not also introduced in Andhra Pradesh. Andhra Pradesh's other health initiatives were considerably less ambitious and costly than Aarogyasri. The authors find that Andhra Pradesh recorded faster growth than Maharashtra (even after adjusting for confounders) in inpatient admissions per capita (for all income groups) and in surgery admissions (among the poor only), slower growth in out-of-pocket payments for inpatient care (in total and per admission, but only among the better off), and slower growth in transport and outpatient out-of-pocket costs. Bergkvist et al. argue that these results are consistent with Aarogyasri having the intended effects, but also with minor health initiatives in Andhra Pradesh (especially the ambulance program) playing a role.
Interventions to raise voluntary enrollment in a social health insurance scheme: Evidence from a cluster randomized trial in the Philippines
Joseph J. Capuno, Aleli D. Kraft, Stella Quimbo, Carlos R. Tan Jr. and Adam Wagstaff  report the results of a cluster randomized controlled trial that tested two sets of interventions to encourage enrollment in the Philippines’ Individual Payer Program (IPP). Of 243 municipalities, 179 were randomly assigned as intervention sites and 64 as controls. In early 2011, 2,950 families were interviewed; unenrolled IPP-eligible families in intervention sites were given an information kit and a 50 percent premium subsidy until the end of 2011. In February 2012, the “non-compliers” had their voucher extended, were re-sent the enrollment kit, and received SMS reminders. Half were told that in the upcoming end-line interview the enumerator could help complete the enrollment form, deliver it to the insurer, and have identification cards mailed. The control and intervention sites were balanced at baseline. In the control sites, 9.9 percent (32/323) of eligible individuals had enrolled by January 2012, compared with 14.9 percent (119/801) in intervention sites. In the sub-experiment, enrollment was 3.4 percent (10/290) among eligible non-compliers and who did not receive assistance but 39.7 percent (124/312) among those who did. The conclusion is that a premium subsidy combined with information can increase voluntary enrollment in a social health insurance program, but less than an intervention that reduces the enrollment burden; even that leaves enrollment below 50 percent.
Encouraging health insurance for the informal sector: Evidence from a cluster randomized trial in Vietnam
Subsidized voluntary enrollment in government-run health insurance schemes is often proposed as a way of increasing coverage among informal sector workers and their families. Adam Wagstaff, Ha Thi Hong Nguyen, Huyen Dao and Sarah Bales  report the results of a cluster randomized control trial in which 3,000 households in 20 communes in Vietnam were randomly assigned at baseline to a control group or one of three treatments: an information leaflet about Vietnam’s government-run scheme and the benefits of health insurance; a voucher entitling eligible household members to 25 percent off their annual premium; and both. The four groups were balanced at baseline. In the control group, 6.3 percent (82/1296) of individuals were enrolled in the endline, compared with 6.3 percent (79/1257), 7.2 percent (96/1327), and 7.0 percent (87/1245) in the information, subsidy, and combined intervention groups; the adjusted odds ratios were 0.94, 1.12, and 1.15, respectively. Only among those reporting poor health were any significant intervention effects found, and only for the combined intervention: an enrollment rate of 16.3 percent (33/202) compared with 8.3 percent (18/218) in the control group, and an adjusted odds ratio of 2.50. The results suggest limited opportunities to raise voluntary health insurance enrollment through information campaigns and subsidies, and that these interventions exacerbate adverse selection.
The impact of a pay-for-performance scheme on prescription quality in rural China: An impact evaluation
In China, health care providers have traditionally been paid fee-for-service, and overprescribing and high out-of-pocket spending are common. Xiaojie Sun, Xiaoyun Liu, Qiang Sun, Winnie Yip, Adam Wagstaff and Qingyue Meng  assigned township health centers in two counties (almost) randomly to two groups: in one, fee-for-service was replaced by a global capitated budget; in the other, by a mix of global capitated budget and pay-for-performance (P4P). Performance captured inter alia "irrational" drug prescribing; 20 percent of the global capitated budget was withheld each quarter, points were deducted for failure to meet targets, and some of the withheld budget was returned in line with the points deducted. The authors examined the effects of P4P on appropriate prescribing and prescription cost, data on which were obtained by digitizing prescriptions from a month just before the reform and from the same month a year later. Impacts were assessed via multivariate differences-in-differences with township health center fixed effects; to reduce bias from non-randomness in assignment, the sample was trimmed by coarsened exact matching. P4P reduced inappropriate prescribing significantly and substantially in the county where the initial level was above the penalty threshold, but endline rates were still appreciable; no effects were seen in the county where initial levels were around or below the threshold, or on out-of-pocket spending in either county.
CATA meets IMPOV: a unified approach to measuring financial protection in health
Up to now catastrophic and impoverishing payments have been seen as two alternative approaches to measuring financial protection in health. Adam Wagstaff and Patrick Hoang-Vu Eozenou  propose a unified methodology in which impoverishing and catastrophic payments are mutually exclusive outcomes. They achieve this by expressing out-of-pocket payments as a ratio of 'discretionary' consumption, defined as the amount by which total consumption (gross of out-of-pocket payments) exceeds the poverty line. This allows Wagstaff and Eozenou to identify both households who are impoverished by out-of-pocket payments (their ratio exceeds one) and households who are pushed even further into poverty by out-of-pocket payments (their ratio is negative); the authors call such payments 'immiserizing'. Households experiencing 'catastrophic' payments are a subset of those who incur out-of-pocket payments but who are neither impoverished nor immiserized by them. Two alternative definitions of catastrophic payments are offered: those that absorb more than a pre-specified fraction of discretionary consumption; and those that leave a household's nonmedical consumption (total consumption net of out-of-pocket spending) below a pre-specified multiple of the poverty line. The authors also offer a simple financial protection index that reflects the percentages of households incurring immiserizing, impoverishing, catastrophic, non-catastrophic, and zero out-of-pocket payments. They illustrate their unified approach with data from the World Health Survey, using international poverty lines and a catastrophic payment threshold of 40 percent.
New articles and books
Are health shocks different? Evidence from a multishock survey in Laos
Using primary data from Laos, Adam Wagstaff and Magnus Lindelow  compare a broad range of shocks in terms of their incidence, distribution between the poor and the better off, idiosyncrasy, costs, coping responses, and self-reported impacts on well-being. Health shocks are more common than most other shocks, more concentrated among the poor, more idiosyncratic, more costly, trigger more coping strategies, and highly likely to lead to a cut in consumption. Household members experiencing a health shock lost, on average, 0.6 point on a five-point health scale; the wealthier are better able to limit the health impacts of a health shock.
Changes in addressing inequalities in access to hospital care in Andhra Pradesh and Maharashtra states of India: A difference-in-differences study using repeated cross-sectional surveys
Mala Rao, Anuradha Katyal, Prabal V Singh, Amit Samarth, Sofi Bergkvist, Manjusha Kancharla, Adam Wagstaff, Gopalakrishnan Netuveli and Adrian Renton  compare the effects of the Rajiv Aarogyasri Health Insurance Scheme of Andhra Pradesh (AP) with health financing innovations in Maharashtra (MH) over time on access to and out-of-pocket expenditure on hospital inpatient care. The authors use two cross-sectional surveys: as a baseline, the data from the NSSO 2004 survey collected before the Aarogyasri and other schemes were launched; and as post-intervention, a survey using the same methodology conducted in 2012. Participants included 8,623 households in AP and 10,073 in MH. The main outcome measures were average out-of-pocket payments, large out-of-pocket payments, large borrowing per household per year for inpatient care, hospitalization rate per 1000 population per year. Average expenditure, large expenditures and large borrowings for inpatient care increased in MH and AP, but the increase was smaller in AP across these three measures. Differential changes between the states for average expenditure and large borrowing were significant favored AP for the rural and the poorest households. Hospitalization rates also increased in both states but more so in AP, although the differential growth was not significant and the subgroup analysis presented a mixed picture.
In the news
An op-ed in The Hindu by Sudhir Krishnaswamy and Rajgopal Saikumar arguing that legitimacy needs to be restored to India’s Public Interest Litigation cites Varun Gauri’s  empirical work on how the poor fare in grievance redress in India and elsewhere.
Stuti Khemani’s  work on the spending implications of relations between states and central government in India is picked up in an editorial in The Financial Express. Khemani’s earlier work  on the political economy of state budget deficits in India is picked up in an op-ed in Mint and another in Mint Asia.
And on the blogs
On Development Impact Damien De Walque  gives FIFA a lesson on stratified randomization and shows what the World Cup groups should have looked like.
Also on Development Impact Berk Ozler  argues that CCTs are at best a ‘treatment’ for poverty – not a ‘cure’. He argues for a competition to test the long-term poverty-reduction effects of the most promising ‘cures’ for poverty.
Also writing on Development Impact Jed Friedman  looks at how other social sciences address the issue of external validity. In another post, Friedman looks at a specific threat to external validity: site-selection bias.
In yet another Development Impact post Vijayendra Rao  argues that in order to build ‘adaptive capacity’ impact evaluations need to be integrated with monitoring and decision-support systems, methods to understand mechanisms of change, and efforts to build feedback loops that pay attention both to everyday and long-term learning.