Whistle While You Work

In addition to information about downsizing, we use the employment growth rate in the industry of employment in the state of residency as a second type of instrumental variable for the subjective probability of job loss. [...] Regarding the CESD scale, we find that a 25-percentage-points increase in the subjective probability of job loss (about one standard deviation, see Table A1 in the appendix) results in an increase of 0.14 in the CESD index using the OLS estimate, and in an increase of 0.59 using the IV estimate; it also results in an increase of 2 percentage points or 9 percentage points in the probability of scor [...] Table A3 in the Appendix shows the job insecurity affects each of the components of the CESD index and, as in Table 2, the IV estimates are much larger than the OLS estimates. [...] The instruments in the IV regression includes whether the employer downsized (and whether it affected jobs similar to the respondent) and the employment growth in the individuals' industry in their state of residency. [...] This explanation can help to understand the difference in the size of the effects between the OLS and IV estimates, since the complier population in the IV estimation is composed of workers whose job loss expectations were increased because of external reasons, namely that their employers are downsizing (or recently downsized) or that their industries are underperforming in their state of residenc