ABSTRACT
To understand economic insecurity generated by job loss it is crucial to analyze who is affected by this event. This is important for two reasons: first, such knowledge helps to understand the distribution of insecurity that job loss and unemployment generate. In other words, while displacements almost certainly have negative consequences, some people may have lower chances of losing their jobs and hence a lower probability of economic instability. Second, understanding the selectivity of job loss and unemployment is important for estimating the consequences correctly. If the unemployed are selected on characteristics that also influence income trajectories after job loss, the difference-in-difference models are biased, as discussed in Chapter 4. The remedy for selection bias I use is statistical matching, as mentioned earlier. This, however, hinges on knowledge about the selection processes. Therefore, I use the results from this chapter to select the variables for matching.
