but are not as focused on additional earnings, whereas women without health
insurance increase their hours and earnings, possibly to afford health costs.15
We also look at the transitions that these wives make as a result of the
change in their husbands’ health insurance availability. Table 6 reports marginal
effects from multinomial logit regressions that examine transitions into and out of
not working, full- and part-time work and self-employment. Panel I replicates our
main regression in a multinomial logit framework for all wives in our sample (i.e.
not conditioning on the wife’s labor force experiences in the previous year).
Consistent with the main results, not working outcomes decrease about 1.95 ppt,
and much of this decrease appears to be women increasing their full-time labor
force participation. However, these transitions are different for women who were
unattached to the labor force in the previous period. Panel II limits our sample to
women who were not working at all in the previous year. These individuals are
more likely to enter the labor force upon the husbands’ receipt of VA insurance,
and they appear to predominately enter part-time work for an employer or self-
employment. Women who worked part-time in the previous year, on the other
hand, as shown in Panel III, do not change their labor force attachment, but
instead work more hours, moving from part-time work into full-time work. This
transition potentially makes them eligible for employer-provided health insurance
15 These results should be interpreted with some degree of caution, as the Current Population
Survey variables for health insurance may not be reliable or consistent during these years (Fronstin
2000, Nelson and Mills 2001).
21