Business Services Industry
The effects of family structure on educational attainment: do the effects vary by the age of the child?
American Journal of Economics and Sociology, The, Jan, 1995 by Steven Garasky
I
Introduction
Projections indicate that one-half or more of the children living in the United States will spend at least part of their childhood in a single parent family (Bum-pass, 1984). One in four children are expected to live in stepparent families (Hofferth, 1985). Should we be concerned?
Growing up with only one parent is found to be related to attaining lower levels of education, becoming a parent earlier, being more likely to have premarital births, marrying earlier, and being more likely to divorce when compared to children who lived with both biological parents throughout their childhood (Astone & McLanahan, 1991; Bumpass & McLanahan, 1989; Haurin, 1992; Haveman, Wolfe, & Spaulding, 1991; Hogan & Kitagawa, 1985; Krein, 1986; Krein & Belief, 1988; McLanahan, 1985, 1988; McLanahan & Bumpass, 1988; Mueller & Cooper, 1986; Sandefur, McLanahan, & Wojtkiewicz, 1992). Similarly, children who at some time lived with a stepparent are less likely to graduate from high school, more likely to marry earlier, and more likely to become parents earlier than other children (Astone & McLanahan, 1991; Haurin, 1992; Michael & Tuma, 1985).
For these children, many of the outcomes mentioned above, especially not finishing high school, lead to reduced employment, lower earnings, lower occupational and economic attainment, lower family incomes, and greater likelihood of receiving welfare assistance or going without material necessities during their adulthood (Krein, 1986; Mueller & Cooper, 1986; Veum & Weiss, 1993). As an example, of all persons 25 years and over in 1992, 11 percent lived in poverty. But the poverty rate was 25.6 percent for those of this group who had not received a high school diploma (Census, 1993).
This study considers how the various family structures children experience throughout childhood are related to the likelihood of their graduation from high school. Specifically, the impact of family structure is considered by the type of structure experienced, the age of the child when the experience occurred, and the age of the child when a change in family structure occurred.
II
Previous Studies
Research examining the effects of family structure on the educational attainment of children is abundant. McLanahan (1985) considers the effects of different types of single mother families. Astone and McLanahan (1991), Li and Wojtkiewicz (1992), and Sandefur et al (1992) all study the effects of stepparent families. Krein (1986), Krein and Beller (1988) and Li and Wojtkiewicz (1992) examine duration effects related to experiences within nontraditional families. Each of these studies concludes that periods of childhood spent outside a two parent structure adversely affect educational attainment.
How researchers model childhood experiences varies across studies. Often, analyses examine family structure at only one time during childhood (Astone & McLanahan, 1991; Bumpass & McLanahan, 1989; Dawson, 1991; Greenberg & Wolfe, 1982; Hogan & Kitagawa, 1985; Keith & Finlay, 1988; McLanahan, 1983; Michael & Tuma, 1985; Sandefur, McLanahan, & Wojtkiewicz, 1989). While Krein (1986) expands the modeling by using indicator variables to designate if any time was spent in a single parent family during preschool, elementary school or high school years, she does not consider other family structures. Similarly, Krein (1986) and Krein and Beller (1988) include as an explanatory variable the number of years the child spent in a single parent family.
A one time observation of family structure combined with a measure of the timing of a family formation disruption is used by McLanahan (1985, 1988) and McLanahan and Bumpass (1988) to model childhood experiences and their effects on educational attainment. Building upon these studies, Haveman et al (1991) combine a measure of family structure with measures of the number of marital disruptions and remarriages. Finally, Haurin (1992) adds more complexity to the modeling by including continuous variables for: (1) the number of years a child lived with two (biological, step, or adoptive) parents and (2) the number of years a child lived with only his or her mother; and dichotomous variables for whether a child ever experienced residence: (1) with a stepparent, (2) without a parent but with other relatives, and (3) in foster care. As with the simpler models that use only one observation of family structure during childhood, these studies generally find that time spent away from a two biological parental structure negatively affects educational attainment.
Wojtkiewicz (1993) steps away from the modeling of family structure to ask if the additional complexity that can be achieved with the available longitudinal data sets is necessary. He compares several of the modeling constructs discussed above with a simpler model of whether a family disruption occurred during childhood to determine if more complex models add additional explanatory power. Unfortunately, he considers only mother-father, mother-only and mother-stepfather structures in his combined models. Wojtkiewicz (1993) finds that any time spent in a nontraditional family has a negative effect on the likelihood of graduating from high school. He concludes that while more complex measures of parental structure have explanatory power in a few instances, they do not contribute to fuller explanations in others.
