450G ECO Notes Fall 2006 Semester University of Kentucky Professor Aaron Yelowitz Chapter 7 Family Size and Structure
Poverty Rates and Family Size Figure 7.1 shows poverty rates increase dramatically with number of children. For female headed families, poverty rates are higher than 80% for those with 5 or more children; among married couples, it is 30%. What are the potential mechanisms that relate poverty rates and family size? Child care costs: Higher costs imply reduced labor force participation (and hence, earnings). How should child care costs affect female heads versus married couples? Since childcare at home has a large fixed cost, but a small variable cost, it is not much more expensive to take care of two or three kids at home than it is to take care of one. As a consequence, this cannot explain increases in the poverty rate by number of kids. Poverty rates rise more dramatically for female heads than for two-parent families yet once the female head is out of the labor force, it is not clear why earnings or income would vary much from child care costs. Mechanical effects: The poverty line is mechanically higher for larger families; holding all else constant, increasing the threshold would increase poverty rates. Schiller reports that if we adjusted the poverty threshold downward by for each family by one person (e.g., give a 5-person family the 4-person threshold, etc.), then 330,000 poor families (or 2 million people) would no longer be in poverty. If we imposed a uniform, 4-person threshold for all families with 5 or more people, 500,000 families would be removed from poverty. This kind of statistical exercise suggest that mild to severe limits on family size may reduce poverty levels. Unlike past discussions where Schiller is content to assume that there is no behavioral response, Schiller makes a number of interesting points here on the reactions that may happen to removing a family member. It is refreshing to see him worry about other reactions, but it would be nice if
he used this kind of economic analysis throughout. He says The statistical elimination of one family member is presumed to have no effect on family income. But this is not a realistic assumption. The following effects are possible: Welfare income would fall with smaller families because it goes up with family size. In reality, the reduction in welfare benefits is small relative to the reduction in the poverty line. Family income may fall with smaller families. Schiller states The father ... may be impelled to work overtime or at a second job ... hence, larger families may constitute a special kind of incentive to work. This probably wouldnt hold for female heads. The mother may be able to go back to work because Schiller states older children who can contribute ... by assuming child-care responsibilities. Welfare: Benefits increase with family size, as do the disincentives to work from the welfare system. Selection bias: Those who have larger families are doing so because their labor market prospects are particularly bad instead of choosing market work, they choose child rearing. Put differently, there are other factors, such as motivation and ability, which affect both family size and earnings. Thus, it would not necessarily be the case that if we gave 1-child families another child, that their poverty rates would go up as dramatically. Declining Family Size Many women are waiting longer to start having children, and ultimately have fewer children than in earlier decades. 1960s: 33% of all children lived in families with at least four children 1990: 15%. Causal Relation Schiller asserts There is no evidence yet that larger families now in poverty enjoyed a higher standard of living when they were smaller. He argues that larger families may increase the duration in poverty, but not necessarily the number of different families in poverty. Given the selection issues (and other behavioral responses), how could you try to scientifically assess the impact of larger family size on poverty rates? A recent study by Angrist and Evans (American Economic Review, June 1998), uses the sex composition of the two oldest kids in the family to predict
subsequent fertility behavior and labor supply behavior. The idea is that families in the U.S. with either two boys or two girls are more likely to have a third child than families with one boy and girl, and that this serves as a sort of randomization or natural experiment. They find that married women who have a third child reduce their labor supply by as much as women in the full sample, while there is no relationship between wives' child-bearing and husbands' labor supply. See http://papers.nber.org/papers/W5778 for discussion. How does access to birth control affect fertility of the poor? Schiller notes that: - Public schools in low income areas are reluctant to provide birth control information - Welfare authorities are usually prohibited from providing such information - Low income families are less able to afford abortion - Medicaid for payments abortion are greatly limited Therefore, Schiller concludes Poor families often end up with more children than other families, not because they want them, but because they are unable to prevent them. Does it follow that family planning policy necessarily reduces family size? An article published by Thomas Kane and Douglas Staiger (Quarterly Journal of Economics, 1996), suggests that recent restrictions on abortion access, including the closing of clinics and restrictions on Medicaid funding led to lower teen birth rates. See http://www.dartmouth.edu/~dstaiger/Papers/KaneStaiger_qje1996.pdf. Another provocative recent article by Gerald Oettinger (1999) uses data on individual enrollment in sex education classes and finds that taking such a class increases the likelihood of becoming sexually active earlier, but has only a weak effect on the likelihood of early pregnancy. See Oettinger, The Effects of Sex Education on Teen Sexual Activity and Teen Pregnancy", Journal of Political Economy, June 1999. Therefore, it isnt obvious that increased access to abortion, greater access to birth control, or additional sex education will reduce fertility, and could increase it.
Changing Family Patterns The number of children that live with both parents has fallen by about 20 percentage points in the last 40 years, with a corresponding increase in the number living in female-headed households. The fraction of children born to unmarried (e.g., divorced, separated, widowed, or never-married) women has skyrocketed: 1960: 5.3 out of 100 children were born to unwed mothers 2000: 33.2 out of 100.
Among African-Americans, the rate has increased from 22% in 1960 to 69% in 2000. See Figure 7.2. Economic Implications Family income potentially falls by more than one-half with the departure of one parent because of the fixed costs of work (e.g., child care). Never-married women usually face greater obstacles in securing child support (e.g., establishing paternity), and are not entitled to alimony. Teen mothers are more likely to drop out of school, and fail to accumulate human capital. Schillers footnote 9 says Teen childbearing reduces the mothers later wages by 13 to 23 percent according to Daniel Klepinger, Shelly Lundberg, and Robert Plotnick. Is teen motherhood that bad? Having a baby as a teenager is a clearly a choice, and it is likely that those who carry to term as a teenager have worse job prospects than those who abort or dont get pregnant. A recent study by Bronars and Grogger (American Economic Review, December 1994) examines the impact of teen motherhood on economic outcomes by comparing teen mothers who had twins versus a singleton. They find that having an additional child did not have an effect on economic outcomes. Other, provocative studies have examined siblings (sisters), and teens who had miscarriages as control groups. For an overall evaluation of natural experiments like these, see Teenage Childbearing Is Not So Bad After All...Or Is It? A Review of ...