4. Welfare reform and labor supplyO

4. Welfare reform and labor supply
One key implication of the findings reported in the previous section is that the welfare reform
legislation must have influenced the labor supply decisions of the targeted immigrants.
I now examine if such a labor supply effect can indeed be documented in the immigrant
population.
I restrict my study of the labor supply decision to the sample of persons aged 18–64 years.
I focus on three alternative measures of labor supply. The first indicates if the person is in
the labor force during the survey week. The second gives the log of annual hours worked
in the past calendar year (calculated only in the sample of workers). The third indicates if a
person is working full time, which is defined as working at least 35 h per week (again, this
variable is only calculated in the sample of workers). It is well known that relatively few
part-time workers have access to ESI and other employee benefits.23 The study of full-time
status can then provide an understanding of how workers respond to policy changes on a
labor supply margin that has important implications for health insurance coverage. Finally,
the analysis will be carried out separately for men and women.
The top panel of Table 6 summarizes some of the key trends in labor supply before and
after PRWORA, again classified according to the generosity of the state’s welfare offer
to immigrants. Consider initially the trends in labor supply experienced by native men.
The labor force participation rate of native men was stable over the 1994–2000 period
in both the less and more generous states. In contrast, the labor force participation rate of
immigrants increased slightly from 84.2 to 85.8% in the more generous states, but increased
much faster (from 83.1 to 86.9%) in the less generous states. Put differently, the labor
supply of immigrant men seemed to be extremely responsive to the welfare cutbacks;
immigrants living in states that did not provide state-funded assistance to replace the federal
cutbacks were the ones who experienced the largest increase in labor force participation
rates. Moreover, this increase in labor supply occurred almost entirely among non-citizens.
The labor force participation rate of naturalized citizens, for example, rose slightly from 83.7
to 84.3% in the more generous states, and was stable at 84.0% in the less generous states.
In contrast, the labor force participation rate of non-citizens rose from 84.4 to 86.9% in the
more generous states, but increased by 6 percentage points (from 82.5 to 88.5%) in the less
generous states. The descriptive evidence, therefore, clearly indicates that the immigrant men who could have been most adversely affected by welfare reform substantially increased
their labor supply.
The other measures of male labor supply reported in Table 6 reinforce this pattern. For
example, the annual hours of work of working native men changed by only 3% or 4%,
regardless of where they lived. In contrast, the annual hours of work of non-citizen men
rose by about 9% if they lived in the more generous states and by 13% if they lived in
the less generous states. Interestingly, the behavioral labor supply response in the affected
immigrant population included a sizable increase in the fraction of immigrant men who
worked full time. The fraction of native men who worked in full-time jobs was relatively
stable over the period, increasing by only about 1 percentage point in both the more and
less generous states. In contrast, the fraction of non-citizens who worked full-time jobs
rose by 3.5 percentage points (from 88.0 to 91.5%) in the more generous states, but by 6.2
percentage points (from 84.8 to 91.0%) in the less generous states.
The trends in female labor supply are not as striking as those documented in the male
sample. The data generally suggest that female immigrants living in the less generous states
increased their labor supply relatively more, but the results are not very consistent. For
example, the labor force participation rate of non-citizen women rose by about 3 percentage
points regardless of the state where they lived. In contrast, annual hours of work of
non-citizen women rose by 10% if they lived in the more generous states and by 14% if
they lived in the less generous states. The discrepancy between the labor supply trends of
immigrant men and women may indicate the existence of spillover labor supply effects
within families (since typically only one family member needs to be covered by ESI), as
well as suggest the possibility that female labor force participation plays a different role in
native and immigrant families, a proposition that has not been sufficiently analyzed in the
existing literature.24
To investigate the extent to which these labor supply trends can be explained by differences
in socioeconomic characteristics among the groups or by state-specific trends in economic
or social conditions, consider again the triple-difference regression model: