Abstract
Under attrition, usually panel data are trimmed to include only the individuals with complete records; then, to allow for a unit-specific effect related to regressors, LSE is applied to the first-differenced model. This paper provides conditions for the LSE to be valid.
| Original language | English |
|---|---|
| Pages (from-to) | 251-256 |
| Number of pages | 6 |
| Journal | Economics Letters |
| Volume | 74 |
| Issue number | 2 |
| DOIs | |
| Publication status | Published - Jan 2002 |
| Externally published | Yes |
Keywords
- Attrition
- Panel data
- Sample-selection
ASJC Scopus subject areas
- Finance
- Economics and Econometrics