Estimation of a Labour Supply Model with Censoring Due to Unemployment and Underemployment
John C. Ham
Abstract
John C. Ham
Abstract
This study proposes and implements a method of labour supply estimation which is appropriate when the sample contains unemployed and underemployed workers. The estimation method consists of excluding unemployed and underemployed workers from the sample and then using (to avoid selection bias) an extension of Heckman's approach to the case where two correlated selection rules generate the sample. Hausman's specification test is then used to determine whether ignoring constrained workers has led to biases in traditional labour supply estimates, and the empirical results suggest that previous estimates of several important parameters are biased. Since the biases go in the direction that would be predicted by the hypothesis that the unemployed and underemployed are constrained, the results support this hypothesis.
OpenAlex reports 270 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This study proposes and implements a method of labour supply estimation which is appropriate when the sample contains unemployed and underemployed workers. The estimation method consists of excluding unemployed and underemployed workers from the sample and then using (to avoid selection bias) an extension of Heckman's approach to the case where two correlated selection rules generate the sample. Hausman's specification test is then used to determine whether ignoring constrained workers has led to biases in traditional labour supply estimates, and the empirical results suggest that previous estimates of several important parameters are biased. Since the biases go in the direction that would be predicted by the hypothesis that the unemployed and underemployed are constrained, the results support this hypothesis.
Key concepts: Underemployment, Unemployment, Economics, Econometrics, Censoring (clinical trials), Estimation, Sample (material), Selection bias