Abstract
Poverty models are generally estimated by using sample surveys affected
by missing data problems. Most methods proposed to take account of missing data problems consider point estimators which typically impose restrictive assumptions. However, it is possible to identify a range of logically possible values for the poverty probability, an identification interval, without imposing any assumption. It is then of interest to check whether the point estimates lie within the identification interval. This is a way to check the validity of the assumptions imposed by point estimators. Using the ECHP we perform this check to assess different estimation methods.
by missing data problems. Most methods proposed to take account of missing data problems consider point estimators which typically impose restrictive assumptions. However, it is possible to identify a range of logically possible values for the poverty probability, an identification interval, without imposing any assumption. It is then of interest to check whether the point estimates lie within the identification interval. This is a way to check the validity of the assumptions imposed by point estimators. Using the ECHP we perform this check to assess different estimation methods.
| Original language | English |
|---|---|
| Pages (from-to) | 1 |
| Number of pages | 22 |
| Journal | Empirical Economics |
| Volume | 38 |
| Issue number | 1 |
| Early online date | 8 Jan 2009 |
| DOIs | |
| Publication status | Published - 1 Feb 2010 |
Keywords
- Missing data
- Partial identification
- Propensity score
- Imputation
- Poverty