Alessandro Tarozzi

Using Census and Survey Data to Estimate Poverty and Inequality for Small Areas

Household expenditure survey data cannot yield precise estimates of poverty or inequality for small areas for which no or few observations are available. Census data are more plentiful, but typically exclude income and expenditure data. Recent years have seen a widespread use of small-area “poverty maps” based on census data enriched by relationships estimated from household surveys that predict variables not covered by the census. These methods are used to estimate putatively precise estimates of poverty and inequality for areas as small as 20,000 households. In this paper we argue that to usefully match survey and census data in this way requires a degree of spatial homogeneity for which the method provides no basis, and which is unlikely to be satisfied in practice. The relationships that are used to bridge the surveys and censuses are not structural but are projections of missing variables on a subset of those variables that happen to be common to the survey and the census supplemented by local census means appended to the survey. As such, the coefficients of the projections will generally vary from area to area in response to variables that are not included in the analysis. Estimates of poverty and inequality that assume homogeneity will generally be inconsistent in the presence of spatial heterogeneity, and error variances calculated on the assumption of homogeneity will underestimate mean squared errors and overestimate the coverage of calculated confidence intervals. We use data from the 2000 census of Mexico to construct synthetic “household surveys” and to simulate the poverty mapping process using a robust method of estimation; our simulations show that while the poverty maps contain useful information, their nominal confidence intervals give a misleading idea of precision.

Semiparametric Efficiency in GMM Models with Auxiliary Data

We study semiparametric efficiency bounds and efficient estimation of parameters defined through general moment restrictions with missing data. Identification relies on auxiliary data containing information about the distribution of the missing variables conditional on proxy variables that are observed in both the primary and the auxiliary database, when such distribution is common to the two data sets. The auxiliary sample can be independent of the primary sample, or can be a subset of it. For both cases, we derive bounds when the probability of missing data given the proxy variables is unknown, or known, or belongs to a correctly specified parametric family. We find that the conditional probability is not ancillary when the two samples are independent. For all cases, we discuss efficient semiparametric estimators. An estimator based on a conditional expectation projection is shown to require milder regularity conditions than one based on inverse probability weighting.

Calculating Comparable Statistics From Incomparable Surveys, With an Application to Poverty in India

We develop an intuitive and easily implemented procedure to recover comparability over time of statistics computed using databases made incomparable by changes in survey design. Our methodology can be adopted whenever the statistic of interest satisfies a certain simple moment condition. The moment condition is satisfied by many interesting economic indicators, including a broad range of poverty and inequality measures. The procedure we propose requires the existence of a set of auxiliary variables whose reports are not affected by the different survey design, and whose relation with the main variable of interest is stable across the surveys. The adjusted estimates can be recovered by using a two-step method of moments framework. Root-n consistency follows easily under regularity conditions. Because most household surveys adopt a multi-stage design, we provide expressions for the asymptotic variance which are robust to the presence of clustering and stratification. We use our adjustment procedure to estimate poverty counts from the 55th Round of the Indian National Sample Survey, a large household survey carried out in 1999-2000. Due to important changes in the adopted questionnaire the unadjusted figures are likely to understate poverty relative to the previous rounds. We provide evidence supporting the plausibility of the identifying assumptions and we conclude that most of the very large reduction in poverty implied by the unadjusted figures is real

Child Nutrition in India in the Nineties

Abstract India experienced several years of fast economic growth during the 1990s, and according to many observers this period also saw a considerable decline in poverty, especially in urban areas. We use data from two rounds of the National Family and Health Survey to evaluate changes in nutritional status between 1992–93 and 1998–99 among children ages 0–3. We find that measures of short‐term nutritional status based on weight given height show large improvements, especially in urban areas. Height‐for‐age, an indicator of long‐term nutritional status, also shows improvement, but one limited to urban areas. However, we also document that the changes in nutritional status are much more favorable for boys than for girls. The gender differences in the changes over time appear to be driven by states in North India, where the existence of widespread son preference has been documented by an immense body of research.

Microcredit, Family Planning Programs, and Contraceptive Behavior: Evidence From a Field Experiment in Ethiopia

The impact of community-based family planning programs and access to credit on contraceptive use, fertility, and family size preferences has not been established conclusively in the literature. We provide additional evidence on the possible effect of such programs by describing the results of a randomized field experiment whose main purpose was to increase the use of contraceptive methods in rural areas of Ethiopia. In the experiment, administrative areas were randomly allocated to one of three intervention groups or to a fourth control group. In the first intervention group, both credit and family planning services were provided and the credit officers also provided information on family planning. Only credit or family planning services, but not both, were provided in the other two intervention groups, while areas in the control group received neither type of service. Using pre- and post-intervention surveys, we find that neither type of program, combined or in isolation, led to an increase in contraceptive use that is significantly greater than that observed in the control group. We conjecture that the lack of impact has much to do with the mismatch between women’s preferred contraceptive method (injectibles) and the contraceptives provided by community-based agents (pills and condoms).

A Cluster-Based Randomized Controlled Trial Promoting Community Participation in Arsenic Mitigation Efforts in Bangladesh

ObjectiveTo reduce arsenic (As) exposure, we evaluated the effectiveness of training community members to perform water arsenic (WAs) testing and provide As education compared to sending representatives from outside communities to conduct these tasks.MethodsWe conducted a cluster based randomized controlled trial of 20 villages in Singair, Bangladesh. Fifty eligible respondents were randomly selected in each village. In 10 villages, a community member provided As education and WAs testing. In a second set of 10 villages an outside representative performed these tasks.ResultsOverall, 53% of respondents using As contaminated wells, relative to the Bangladesh As standard of 50 μg/L, at baseline switched after receiving the intervention. Further, when there was less than 60% arsenic contaminated wells in a village, the classification used by the Bangladeshi and UNICEF, 74% of study households in the community tester villages, and 72% of households in the outside tester villages reported switching to an As safe drinking water source . Switching was more common in the outside-tester (63%) versus community-tester villages (44%). However, after adjusting for the availability of arsenic safe drinking water sources, well switching did not differ significantly by type of As tester (Odds ratio =0.86[95% confidence interval 0.42-1.77). At follow-up, among those using As contaminated wells who switched to safe wells, average urinary As concentrations significantly decreased.ConclusionThe overall intervention was effective in reducing As exposure provided there were As-safe drinking water sources available. However, there was not a significant difference observed in the ability of the community and outside testers to encourage study households to use As-safe water sources. The findings of this study suggest that As education and WAs testing programs provided by As testers, irrespective of their residence, could be used as an effective, low cost approach to reduce As exposure in many As-affected areas of Bangladesh.

Bednets, Information and Malaria in Orissa

We study the identification and estimation of key parameters in a basic model of technology adoption when specifically collected information on subjective beliefs and expectations about the technology’s impact is available. We discuss identification with both non-parametrically and parametrically specified utility as well as parametric and semi-parametric specifications for unobserved heterogeneity. We propose parametric and semi-parametric estimation methods to recover underlying preferences and use the model to study the adoption of bednets among poor households in rural Orissa (India). We carry out counterfactual exercises to examine the effects of price and belief changes on net ownership decisions. The results suggest that net purchase decisions are relatively insensitive to changes from current prices and beliefs. The methods proposes here should have applicability to other discrete choice settings with non-linear indices.

High prevalence of Wuchereria bancrofti infection as detected by immunochromatographic card testing in five districts of Orissa, India, previously considered to be non-endemic

India accounts for over one-third of the worlds burden of lymphatic filariasis (LF). Although most coastal districts of Orissa state (eastern India) are LF-endemic, the western districts of Orissa are considered non-endemic. During a large-scale insecticide-treated bed net/microfinance trial, we tested one randomly selected adult (age 15-60 years) for LF from a random sample of microfinance-member households in five districts of western Orissa, using immunochromatographic card testing (ICT). Overall, 354 (adjusted prevalence 21%, 95%CI 17-25%) of 1563 persons were ICT positive, with district-wide prevalence rates ranging from 15-32%. This finding was not explained by immigration, as only 3% of subjects had ever lived in previously known LF-endemic districts. These results therefore suggest ongoing autochthonous transmission in districts where LF control programs are not operational. Our results highlight the importance of broad, systematic surveillance for LF in India and call for the implementation of LF control programs in our study districts.

Evolution of households' responses to the groundwater arsenic crisis in Bangladesh: information on environmental health risks can have increasing behavioral impact over time

A national campaign of well testing through 2003 enabled households in rural Bangladesh to switch, at least for drinking, from high-arsenic wells to neighboring lower-arsenic wells. We study the well-switching dynamics over time by re-interviewing, in 2008, a randomly selected subset of households in the Araihazar region who had been interviewed in 2005. Contrary to concerns that the impact of arsenic information on switching behavior would erode over time, we find that not only was 2003-2005 switching highly persistent but also new switching by 2008 doubled the share of households at unsafe wells who had switched. The passage of time also had a cost: 22% of households did not recall test results by 2008. The loss of arsenic knowledge led to staying at unsafe wells and switching from safe wells. Our results support ongoing well testing for arsenic to reinforce this beneficial information.

Child Nutrition in India in the Nineties: A Story of Increased Gender Inequality?

We establish some new interesting stylized facts on the changes in boy versus girl nutritional status in India during the nineties, a period of rapid economic growth. Our analysis is based on the comparison, over time and across genders, of the distribution of z-scores calculated for height and weight measures. Overall, we find that child nutrition improved substantially, but we also find that gender differences in nutritional status increased as well, with nutritional status improving substantially more for boys than for girls. Consistent with a large literature that shows the existence of a steep North-South gradient in gender inequality in India, we find that changes in nutritional status appear to be much more similar between genders in the South. We also estimate predicted changes in nutritional status based on changes in the distribution of household wealth (proxied by asset ownership) and a few other observed household characteristics. Actual changes appear to be relatively close to predicted ones in urban areas. For children living in the rural sector the results are more mixed, and we observe that actual changes in weight are quite larger than predicted ones for boys, while they are much worse than the predicted ones for girl height. We also estimate that the predicted changes are generally larger for boys than for girls.