Welfare Dynamics in Pakistan and Ethiopia – Does the Estimation Method Matter?
利用巴基斯坦和埃塞俄比亚的农户面板数据，检验现有文献中不同估计方法是否导致对贫困陷阱的不同结论，并提出一种新的半参数面板数据估计方法。
Identifying household-level welfare dynamics and associated dynamic poverty trap thresholds can have important implications for the targeting of poverty reduction policies. The small existing empirical microeconomic literature has found evidence both for and against poverty traps. Using household panel data from rural Pakistan and Ethiopia, this article examines whether these different results are likely driven by differences in estimation methods or whether they reflect actual differences across settings. It applies the estimation methods from the existing literature to the same two datasets and also proposes a novel semiparametric panel data estimator that combines the advantages of the previous fully parametric and nonparametric approaches. The results suggest that absent any dynamic poverty trap thresholds the effect of using different estimation methods is secondary, having a small influence on the estimated long-term level of household well-being but not the identification of multiple dynamic welfare equilibria and associated dynamic poverty thresholds. Households in rural Pakistan and Ethiopia seem to be stuck in a static, structural-type poverty trap facing an expected level of long-term well-being that places them squarely in poverty.