Published February 2017
| public
Journal Article
Unequal spacing in dynamic panel data: Identification and estimation
- Creators
- Sasaki, Yuya
- Xin, Yi
Abstract
We propose conditions under which parameters of fixed-effect dynamic models are identified with unequally spaced panel data. Under predeterminedness, weak stationarity, and empirically testable rank conditions, AR(1) parameters are identified given the availability of "two pairs of two consecutive time gaps", which generalizes "two pairs of two consecutive time periods". This result extends to models with multiple covariates, higher-order autoregressions, and partial linearity. Applying our method to the NLS Original Cohorts: Older Men, where personal interviews took place in 1966, 67, and 69, we analyze the earnings dynamics in the old time, and compare the results with more recent ones.
Additional Information
© 2016 Elsevier. Received 20 May 2015, Revised 23 January 2016, Accepted 17 October 2016, Available online 31 October 2016. An associate editor and three anonymous referees contributed much to this paper. We are indebted to them. We also benefited from very useful comments obtained in New York Camp Econometrics 2014, AMES 2014, and International Panel Data Conference 2014. The usual disclaimer applies.Additional details
- Eprint ID
- 87054
- Resolver ID
- CaltechAUTHORS:20180613-095307836
- Created
-
2018-06-13Created from EPrint's datestamp field
- Updated
-
2021-11-15Created from EPrint's last_modified field