QUARTERLY JOURNAL OF ECONOMICS
births] are often insignicant, and when they are not, they are small [pp. 129 –130].” On the other hand, another recent review, Moftt , suggests that the weight of the evidence implies some effect of welfare benets on marriage and fertility. As to location, Meyer  concludes that there is a signicant but small effect of welfare on migration. Overall, it is likely that endogenous single motherhood and location exert a small bias on our results.
C. Accounting for Individual and State Characteristics
As mentioned above, the results in Table II could be partly explained by differential changes over time in characteristics such as age and education for single women with and without children. Moreover, business cycles may differentially affect sin- gle women with and without children, thereby leading to employ- ment shifts unrelated to policy changes. Consequently, Table III presents probit employment estimates for single women control- ling for demographic and business cycle changes. We include a large number of controls for differences between the two groups, and we include the unemployment rate as well as its interaction with whether or not a woman has children. The specication that we estimate is
(6) Pr(Eit 1)
tYEARt t(YEARt ANYCHILDRENi)},
where Eit equals one if woman i from year t reports positive hours worked in the reference week for the ORG (or the previous year for the March CPS), Xit is a vector that includes demographic and business cycle variables, YEARt is an indicator variable for year t, and ANYCHILDRENi equals one for a woman with children. The year dummies control for labor market trends in overall female employment and the X vector controls for demographic and business cycle effect differences between the groups, espe- cially compositional shifts over time. Thus, differences between coefcients give difference-in-differences estimates controlling for these other factors. These differences can be interpreted as esti- mates of the combined effect of changes in all factors affecting the employment of single mothers relative to single women without children. t
The demographic and business cycle variables accounted for in Table III include controls for state, race, ethnicity, age, educa-