This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

« Back to Results

New Methods for Measuring Poverty and Welfare

Paper Session

Friday, Jan. 5, 2018 10:15 AM - 12:15 PM

Marriott Philadelphia Downtown, Meeting Room 306
Hosted By: American Economic Association
  • Chair: Brian Dillon, University of Washington

Measuring Seasonal Poverty

Brian Dillon
,
University of Washington
Paul Christian
,
World Bank

Abstract

Food consumption in developing countries is highly seasonal. A typical household consumes significantly more calories and nutrients in the months after harvest than in the months preceding. Lean season consumption sometimes falls below critical thresholds necessary for children to reach their full potential in both cognitive and physical development. In Christian and Dillon (2017) we show in a long-term panel from Tanzania that conditional on average annual consumption, seasonality of consumption in childhood is associated with reduced human capital (height and educational attainment) in adulthood. This association matters for welfare measurement and targeting regardless of whether the link is causal (though we provide evidence that it is). The takeaway from that paper is that two children who are indistinguishable using standard measures of poverty, which are based on annualized consumption, will have significantly different human capital outcomes if one child's diet is more seasonal.

Having established that seasonality matters for welfare (conditional on average consumption), in this paper we develop an empirically tractable poverty measure that accounts for seasonality. Our measure is an axiomatic generalization of the Foster-Greer-Thorbecke. The intuition behind our approach is simple: a seasonality-adjusted poverty measure is an integral over the household's poverty status on each day of the year. To make the measure empirically tractable, we use an empirical approach related to that in our first paper, which relies on variation in survey timing to distinguish between predictable seasonal variation in consumption and idiosyncratic shocks. This approach differs from others in the literature, such as Ligon and Schechter (2003), in that we use a flexible, parametric specification to model seasonality. We apply our seasonally adjusted measure of poverty to consumption data from Malawi, and compare our poverty rankings to those based on the standard approach and on Ligon and Schechter (2003).

Measuring Poverty and Vulnerability in Real-time

Joshua Blumenstock
,
University of California-Berkeley
Michael Callen
,
University of California-San Diego
Tarek Ghani
,
Washington University-St. Louis
Niall Keleher
,
University of California-Berkeley
Jacob Shapiro
,
Princeton University

Abstract

In wealthy nations, novel sources of data from the internet and social media are enabling new approaches to social science research and creating new opportunities for public policy. In developing countries, by contrast, fewer sources of such data exist, and researchers and policymakers often rely on data that are unreliable or out of date. Here, we develop a new approach for measuring the dynamic welfare of individuals remotely by analyzing their patterns of mobile phone use. To benchmark these methods, we conducted high-frequency panel surveys with 1,200 Afghan citizens, and with the respondent's consent, matched each individual's responses to his or her entire history of mobile phone-based communication, which we obtained from Afghanistan's largest mobile operator. We show that mobile phone data can be used to accurately estimate the social and economic welfare of respondents, and that machine learning models can be used to infer the onset and magnitude of positive and negative shocks. These results have the potential to transform current practices of policy monitoring and impact evaluation.

Measuring Poverty With Satellites

Marshall Burke
,
Stanford University

Abstract

Reliable data on economic well-being in developing countries remain scarce, hampering efforts to understand variation in economic performance and to design policies to improve it. Here we show how the combination of high-resolution imagery from satellites and powerful new tools from machine learning can be used to make accurate, inexpensive measurements of local-level well-being across multiple countries in Africa. We discuss how using multiple spectral wavelengths, and simultaneously incorporating multiple zoom levels in the imagery, can lead to substantially improved predictions. Finally, we quantitatively explore the extent to which satellites can be used to improve the efficiency with which social programs are targeted to the poor.

Detecting Land-use Change and On-farm Investments at the Plot Scale Using Remote Sensing

Jennifer Burney
,
University of California-San Diego
Gordon H. Hanson
,
University of California-San Diego

Abstract

The ability to remotely monitor agro-ecosystems over large spatial scales, at high spatial and temporal resolution, promises to open new and previously un-tractable lines of inquiry about the relationships between management practices, welfare, and resilience in coupled human-natural systems. We use remotely-sensed data (from vegetation indices to synthetic aperture radar) to infer when and where land-use and management changes take place at the farm level, including processes leading to degradation, like overgrazing or tree removal, as well as processes intended to boost resilience, like irrigation and conservation agriculture. Here, we first show how ecosystem health metrics can be used as indicators of both poverty and vulnerability. This is especially important because many other remotely-sensed economic proxies exhibit hysteresis in one direction; that is, they may respond quickly to positive income shocks (e.g., a change in income may rapidly lead to more construction and an expansion of the urban environment), but little if at all to negative shocks (a drop in income does not lead to deconstruction of buildings). We then present results from two field projects that show how these techniques can be used to detect management changes — reflecting changes in household welfare — in both field and quasi/natural experiments.
Discussant(s)
Berk Ozler
,
World Bank
Teevrat Garg
,
University of California-San Diego
Pamela Jakiela
,
University of Maryland
Paul Christian
,
World Bank
JEL Classifications
  • O1 - Economic Development
  • I3 - Welfare, Well-Being, and Poverty