Sociology 229A: Event History Analysis

Short Assignment #4: More EHA Models

Due: Start of class (2:00) May 6

This assignment requires a dataset described in last week’s reading:

http://www.soziologie-blossfeld.de/eha/stata/do_files/Data/rrdat1.dta

-- The reading includes information on all the variables.

In addition, you will need the following do-file:

http://www.soziologie-blossfeld.de/eha/stata/do_files/Chapter%204/ehd3.do

The dataset describes individual employment experiences. The authors initially model the duration of individual employment spells. However, jobs end for VERY different reasons. Therefore, they explore a more complex state space – considering different types of job transition separately. Rather than just looking at timing of “job end”, they look at upward vs. lateral vs. downward moves.

Use the accompanying do-file to run the models (which are exponential/constant-rate models). Modify the do-file to run Cox models, in addition.

Answer the following questions:

1. What is the effect of a 1-unit change in education on each of the outcomes (based on exponential models)?
2. What factors have a distinctive effect on upward job moves (based on exponential models)? Provide an interpretation of those effects.
3. What variables have (more or less) similar effects across all job moves? What does that tell you about who is most mobile in the labor market?
4. Comment generally (briefly) on the difference in results between the exponential and Cox models. Assuming both the exponential and Cox models fit well, which approach is more efficient? Why?

Turn in answers to the questions, plus output of your Cox model for upward job transitions.