| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 9549211 | Economics Letters | 2005 | 5 Pages |
Abstract
Many panel data models include a lagged dependent variable as a regressor. Analytical results in this note show that common forms of misspecification induce negative bias in the coefficient estimate. An analysis of welfare caseloads illustrates these results.
Related Topics
Social Sciences and Humanities
Economics, Econometrics and Finance
Economics and Econometrics
Authors
Terra McKinnish,