The impact of employment trends The individual woman's labor-market situation

6 The effect of labor-market attachment

We now turn to our individual-level covariates. To study how first-birth rates are affected by (childless) women's economic situation, we have also included her income from work (in fixed prices) in the last previous year as a time-varying covariate. Column 5 in [Table 1] contains the "relative risks" for a grouped version of this factor. (In that particular regression we have standardized for calendar year but have left out local employment trends to concentrate on individual-level effects for the moment.) It shows that there is a very strong income effect, which seems to work through two thresholds. If we take as our baseline women who have no income from work (or practically none), then childless women with an annual income between 50 and 100 thousand kronor at 1997 prices, have about twice as high first-birth rates. Women with an income of 100 thousand kronor or more have roughly three times as high first-birth rates as the baseline group. SEK 100 000 was worth between 12 and 13 thousand US Dollars in 1997. There is very little apparent effect of additional income beyond this level. It seems that women with a regular job have much higher first-birth rates than those who do not, but that otherwise the amount of the earned income is unimportant.

[Figure 10] shows that the inclusion of the women's earned income from work greatly reduces the period effect. Evidently, some of the drop in fertility in recent years is "explained" by the increase in the number of women who are practically without any earned income.

The analysis just reported left out local employment trends. When we re-introduce this factor into the regression (and correspondingly remove the period factor to avoid semi-collinearity; see Column 3 of [Table 1]), its effect turns out to be somewhat smaller than when we disregarded the woman's income (Column 2 in the same table), but the income effect itself (Column 4) remains just about the same as in our simpler analysis (Column 5). As before, we find that women with an income of 100 thousand kronor or more have around three times as high first-birth rates as comparable women who have no (or practically no) income from work. This result is very stable and it appears in all of our experiments. There is a similar stability in our other results.

 

The impact of employment trends The individual woman's labor-market situation

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
http://www.demographic-research.org/Volumes/Vol2/4