The Future of Children, Princeton - Brookings: Providing research and analysis to promote effective policies and programs for children.

Large national surveys of women and men can be creatively manipulated to estimate the resources that absent fathers could provide to their children, and thus predict potential support, although several obstacles make this a difficult challenge. The most fundamental problem is that information linking fathers and mothers who do not live together is seldom available, especially if the couple was never married. Many longitudinal data sets collect detailed information from respondents (male and female) about their parenthood, but if the father and mother live separately, little information is collected from one about the other.

One source of data that provides information on both the mother and father is the National Maternal and Infant Health Survey (NMIHS). This data set focuses on a random sample of 9,953 women between 15 and 49 years of age who had a live birth in 1988, and it includes a limited set of characteristics of both the mother and father, such as age, marital status, education, race, receipt of government assistance after the birth, and the father's employment status and residence with the mother. That information makes it possible to characterize the partners of women who receive government assistance, and to ask whether these men are able to provide reasonable amounts of child support.

One can calculate using these data, for a woman of a given age, race, and AFDC status, the probability that her partner was in a given age, education, and marital status group.²⁴ These probabilities allow one to know something about the men with whom women are having children. To calculate potential child support payments, information is needed on both the earnings and the number of children these men will have over the first 18 years of the child's life, the years in which the fathers would be obligated to pay support. For this information, the study uses a sample of men who participated in another survey—the National Longitudinal Survey of Youth (NLSY), which includes extensive demographic and financial data gathered yearly. Because the NLSY documents any child a man has had outside of marriage, and records his annual earnings and the births of additional children, it can be used to construct fatherhood, marriage, income, and education histories for this sample of 4,231 men.

Of course, any survey on fatherhood is limited by the likelihood that some men will fail to report at least some of the children they father, especially men who are unwed and have little contact with the child. One study²⁵ suggests that the men in their early twenties underreport births to the NLSY interviewers by 15% to 23%. This issue is significant because the men who do not report the birth to survey researchers may be those who also walk away from parenting responsibility. When such fathers cannot be included in research samples, study findings may overestimate the involvement of absent fathers with their children.

The information from these two data sets (the NMIHS and the NLSY) can be combined to link mothers with different characteristics to estimates of the financial resources available from the absent fathers of their children. For instance, consider the case of an 18-year-old mother who participates in a welfare program. The NMIHS establishes the probability that her partner has particular characteristics, and the NLSY indicates the likely earnings of that type of man. One can then apply a formula for calculating child support obligations to determine the child support obligation that would confront such a man (based on his income and the number of children he has fathered). Here, child support obligations were computed using Wisconsin's child support statute. This relatively straightforward standard requires that a father with one child pay 17% of his income as child support. If the man has a second child, his obligation toward each child is 12.5% of his income. With three children, this obligation is 9.7%, and with four children it becomes 7.75% per child.²⁶ Finally, to calculate the support available to the 18-year-old mother, one simply multiplies the probability that the partner was in a particular category by the associated potential child support, and sums these products across all types of men.

The data from the NMIHS present a picture not only of the mothers who gave birth in 1988, but also of the men who fathered their children. The mothers who reported receiving AFDC were younger, had lower levels of education, came from households with less income, and had more children than the total group of women who gave birth that year. Given the eligibility rules for AFDC, women who receive welfare are unlikely to be married. Nevertheless, as Table 1 shows, a surprising proportion of welfare recipients report living with the father during the pregnancy and afterwards (36% of nonblack women and 15% of black women coresided with the father of their child when the survey was conducted an average of 16 months after the birth).²⁷ Table 1 also shows that the male partners of AFDC recipients are younger and have less education than the total group of men who became fathers in 1988.²⁸ A much smaller percentage were working, and they worked fewer hours. It must be noted that many women, particularly unmarried women, are unable or reluctant to provide information about their partners: More than 7% of AFDC mothers did not report the work status of their partners.

When the earnings of a sample of men who match the characteristics of absent fathers are examined using the NLSY, it becomes clear that the incomes even of absent fathers vary with the mother's age and race.²⁹ Fathers' earnings are higher for women who delay childbearing to later ages, and the incomes of nonblack women's partners exceed those of black women's partners. Table 2 shows the incomes of the mother's predicted partner over the first 18 years of a child's life, comparing mothers who received AFDC after the birth with those who did not. The table shows the discounted present value of the man's projected 18-year earnings, in 1994 dollars.

Focusing first on women who did not collect AFDC, Table 2 indicates that the likely partner of a nonblack woman who had a child when she was under 18 would earn $239,536 over 18 years, or $1,708 per month. The partner of a woman with similar characteristics who had her child after age 25 would earn $436,622 over those 18 years, or $3,113 per month. Partners of black women who did not receive AFDC earned from $1,622 to $2,510 per month.

Still considering only the women who did not receive AFDC, one finds that if a nonblack woman under 18 was awarded and collected child support under the Wisconsin statutes, she could look forward to $232 per month in child support. Her black counterpart would be owed $220 per month in child support. In general, the value of the expected support increases with the age of the woman and is higher for nonblack women.

Less support is available from partners to women who received AFDC after the child's birth. This is not surprising, given the evidence shown in Table 1 that the partners of AFDC mothers work less and have less education than the typical father. Moreover, the potential child support available to AFDC mothers is less affected by the age of the mother at childbirth. For nonblack women who did not depend on AFDC, the potential monthly child support estimates doubled when comparing the youngest and the oldest mothers. Among nonblack AFDC recipients, the older women could tap only 61% more potential support than was available to the younger women.