Evaluating Poverty Targeting Methods: Community-Based Approaches vs. Self-Targeting Ordeals

This lecture delves into the complexities of identifying the poor for redistribution programs, focusing on two distinct methodologies: community-based targeting and self-targeting through 'ordeals.' The discussion begins with an experiment in Indonesia evaluating community-based approaches, where local villagers rank individuals based on their perception of poverty. The findings reveal that while this method might show more 'errors' when judged solely by consumption metrics, it aligns significantly better with local notions of poverty, private rankings within households, and individuals' self-assessed well-being, leading to higher community satisfaction. This suggests that communities capture a broader, more nuanced understanding of welfare than purely objective economic indicators.

The second part of the lecture introduces the concept of self-targeting, inspired by Nichols and Zeckhauser, which proposes using costly application processes or 'ordeals' to screen out non-needy individuals. The underlying principle is that these costs are differentially borne by the poor (e.g., lower opportunity cost of time for the unemployed). However, the lecture critically examines several potential pitfalls: the ordeal might target an unintended proxy (like wage instead of wealth), wealthier individuals might find ways to circumvent the costs (e.g., hiring agents), and concave utility functions could make the ordeal disproportionately burdensome for the very poor, potentially leading to their exclusion. The discussion also highlights the added complexity when benefit receipt is stochastic, requiring applicants to forecast their probability of success, which can reveal private information to the government.

From a practical standpoint, the research indicates that community-based targeting, despite its potential for 'mistakes' by strict economic definitions, fosters greater local acceptance and satisfaction, implying that program design should consider and integrate local welfare metrics. For self-targeting, the lecture emphasizes that theoretical benefits are not always realized in practice, necessitating rigorous empirical testing. It underscores the importance of designing experiments with a strong theoretical foundation to precisely test hypotheses, while also acknowledging real-world challenges such as spontaneous community efforts to mitigate the burden of ordeals, which can complicate experimental measurements.

Ultimately, both targeting mechanisms aim to enhance efficiency in welfare provision. While ordeals impose costs, the rationale is that by effectively screening out non-poor beneficiaries, the saved resources can be reallocated to increase benefits for the truly needy, potentially making the program more impactful overall. The research moves beyond a simplistic trade-off between 'elite capture' and 'better information,' revealing the critical role of diverse welfare metrics and the intricate behavioral responses to policy design. This work, conducted in a high-stakes environment in Indonesia, provides valuable insights for governments designing and implementing large-scale conditional cash transfer programs.

if you're poor and we gave you the program that's an error — if we're rich and you do sorry poor and you do not get the program that's an error recently you got that program as an error that's kind of discreet

there's some local notion of kind of what poverty is in this Village which is highly correlated with but not perfectly correlated with income and when you tell communities to kind of tell you who's poor that's what you get

the community has a somewhat you know we started off by actually designing this whole project as like a trip we thought it was a trade-off between Elite capture and and uh and and better information right

if your goal is to maximize utility and you kind of believe that the social utility is a function the individual utilities and you think that's what this is capturing then this approach might work better

ordeals can be used to Target the poor

the downside of that approach is that it's imposing a cost on the unemployed people right which is you know in principle you can think that cost is just totally kind of deadweight loss

it may be that the thing you're targeting is you know what you want to Target say is you know maybe wealth and what you're actually targeting is kind of wage and those may not be performed correlated

the point of that was to say that actually the first point of I think that we want to say is that look there's nickel zech has our idea is a nice idea in theory but it's not exante totally obvious