The effect of labor-market attachment The effect of public policies

7 The individual woman's labor-market situation

So far, we have not made use of our information concerning why a childless woman has (practically) no earned income in a given year. In most cases, she is unemployed or enrolled in education, and most often we have then recorded that she has received an unemployment benefit or an educational grant or loan from public sources. (Only negligible groups of childless women are unemployed or study without such allowances, or have little or no income because they are sick or traveling abroad or for some other reason.) To pick up the effect of study and unemployment periods, we have produced another indicator, this time to represent the childless woman's labor-market or activity situation. The details of its definition are seen most easily in [Table 2], which lists the relative "risks" of first birth in an analysis that also includes the woman's age and municipal employment trends. (The effect of the latter covariate is much as in [Table 1] and is not repeated here.) We have subdivided women with an annual income of less than 100 thousand kronor (in fixed prices) according to unemployment benefits and educational allowances received as well as according to their income from work, and have kept the three income levels for those who earned more. As before, the latter subdivision does not reveal much difference in childbearing behavior, but the activity situation of women with low incomes turns out to be very important. (The different effect levels in Tables 1 and 2 are caused by the different choices of baseline in the two tables. Note that each of the three first income categories in Columns 4 and 5 of [Table 1] is replaced by four categories in the upper panel of [Table 2]. This part of our analysis is based on data for childless women at ages 21-38. We have omitted women at ages 17-20 because these days most of them go to school, and very few of them have children.)

As has been shown time and again (see, for instance, [13], Table 25 etc.; [2], [3], [8], [14]), students clearly have by far the lowest fertility among all childless women ([Table 2]). In our data, this is the case at each (low) income level, independently of whether the student has received unemployment benefits as well during the preceding year.

It is intriguing though not really surprising that women who have received unemployment benefits (without also being students), have a relatively high fertility, rather than a low one. People certainly take a cut in income if they become unemployed, but the cut is perhaps not overwhelming, given the generosity of the Swedish benefits system, and so the immediate effect on first births may be rather limited. Unemployment benefits often count on a par with income when entitlements to parental-leave benefits are computed after a birth, so most unemployed women should get a reasonable maternity benefit when they have their first child. An unemployed woman may be pessimistic about her chances of landing a job anytime soon and may want to use the waiting time to have her first child, particularly if she perceives child care as readily available should a job offer come her way after the end of the maternity leave. Being unemployed may thus entail inducements to start childbearing that are strong enough in a woman's eyes to outweigh her lowered chances of getting a job again during pregnancy and after. (Our findings complement the few previous empirical results we have found in the literature about the effect on childbearing of an individual's own unemployment; see [13], Table 25 etc., [10], [11], p. 253. Previous results do not tally across investigations, which is natural since the effects must depend strongly on data characteristics and circumstances in the study population.)

The situation is quite different for women who are enrolled in education (unless they have had a job recently). Their maternity benefits are low (currently 60 kronor or some US $7 per day) and a birth would most likely lead to a break in their studies. Starting childbearing cannot be an enticing prospect under such circumstances. A stream of young women into educational activities is a major intermediary cause of our declining fertility. It represents a strategy to sidestep current difficulties in the job market and is a response to the underlying economic problems that drive the behavior we study. (In 1989, 14.1 percent of all women at ages 21-24 received educational allowances in Sweden. In 1996, the figure was 40.7. At ages 25-28, the figure increased from 9.1 to 21.7 percent.)

The results presented above are the outcomes of analyses based on our preferred model. We have carried out a number of further experiments. To get a better understanding of the character of income effects, for instance, we have replaced our labor-market or activity indicator from Section 7 by the woman's total income from work, unemployment benefits, and educational allowances, taken together, subdivided according to whether she received public benefits and the character of those benefits for women with low incomes, as in [Table 2]. We have also included a woman's income pattern over the two most recent calendar years preceding any current month, not only the single most recent year. None of this changed the character of the results or even the strength of the effects in any major way.


The effect of labor-market attachment The effect of public policies

logo70.gif (2450 bytes)

Entry into motherhood in Sweden:
the influence of economic factors on the rise and fall in fertility, 1986-1997

Britta Hoem
© 2000 Max-Planck-Gesellschaft ISSN 1435-9871