This is an old revision of the document!


Effects of Family Structure on Income

Family structure has much to do with income levels and asset building, both of which lead to economic prosperity.

A family’s income is used to finance immediate needs and, if it is sufficient, may allow the family to save for future needs. There are two elements in the amount of income received: the dollar value of hours worked and the number of hours worked. These in turn are affected by, among other things, the parents’ education level and work habits that typically are formed in the early years.

The marriage of the parents has much to do with a child’s educational attainment and work ethic. The relationship can be expressed as an equation: Income = (education attained) x (work ethic) x (unity of family structure).

1. Influence of Family Structure

According to research by Dr. Henry Potrykus and Dr. Patrick Fagan of the Marriage and Religion Research Institute, the fraction of intact families and the fraction of high school graduates have similar positive and precisely determined1)2) influences on the employment rate among 25- to 54-year-old males.3)

The fraction of college graduates in a geographic area has a smaller but still positive, precisely determined influence on employment when controlling only for demographics and education, but the influence is indeterminate under controls for earnings. Presumably, once the human capital and credentialing effects of higher education are absorbed and accounted for by earned income itself, college has little otherwise specifiable effect on employment propensity. As mentioned above, this is not the case with high school: High school drop-outs behave very differently with respect to employment than those with high school diplomas.

Income level itself has a less precisely determined but otherwise positive relation to employment. The fraction of Hispanics in a geographic area has a modest, positive, and precisely determined influence on the employment rate among 25- to 54-year-old males under all controls. Hispanics in a geographic area tend to work more, or they seek those areas with work more readily than others do, or both. Blacks have no precisely determinable influence on employment rates in an area once other controls4) have been implemented.

The minor dependency ratio (more children present) has a large, positive, precisely determinable influence on men working. Presumably these men are working to support those additional children. The minor dependency ratio in a geographic area also has the largest consistently positive influence on average earnings for males aged 25 to 54.5) The influence of minor dependency is large, positive, and precisely determinable under controls for demographics, education, and earnings levels (of the whole adult population).

The fractions of blacks or Hispanics in a geographic area have smaller, determinable, negative influences on average earnings for males aged 25 to 54 when controlling for demographics, education, and overall area income. The influences are essentially indeterminable under controls for demographics and education only.6)

2. Marriage, Education, and Income

(See Effects of Marriage on Workforce Participation)

Of course, one does not obtain an adequate and steady income just by marrying. Increasing the number of hours worked at a job valued by the marketplace will provide more income. The number of hours worked is linked directly to educational achievement and family structure (See Charts 1 and 2). Families whose members have lower levels of education normally will have to work longer to reach a modest level of financial security than do those whose members achieve higher levels of education.

Labor Force Participation by Level of Education

Education Attainment of Parents by Family Structure

However, people who are not married and have less education work the fewest hours per year. In general, married couples have higher levels of education and work longer (see Charts 3 and 4 ), and make sure that their children achieve higher levels of education.7)

Total Number of Hours Worked

Hours Worked by Intact Married Families

Although the income of a family household depends on the educational level of parents, it is the parents’ income rather than their level of education that predicts more accurately the level of education their children will achieve.8) In general, children with high-income parents receive more education than do children of lower-income parents.9) But higher income is less likely without marriage (see Chart 5), and poverty is much more likely without marriage.

Median Income of Families with Children by Family Structure

Education gives the child from a high-income family a great advantage. The federal government’s Panel Study of Income Dynamics showed the large economic gains that can be realized by completing high school, both in the level of wages earned and in the longer hours per week that a person will work.10) But family background accounts for at least half the variance in educational attainment.11) Students from intact families score more positively on all measures than do those from both step and single-parent families.12) Adolescents who do not live with both natural parents are at significantly greater risk of leaving high school before graduating.13) And the number of years of education received translates into a better first job and better jobs later at higher salaries.14)

3. Marriage, Work Ethic, and Income

A significant portion of two-parent families have moved out of the poverty range because both parents work, which also increases—and in many cases doubles—the total number of hours worked within the household. Among America’s poor, there has been a significant shift in the number of hours worked per household, which indicates that much of the disparity in young men’s economic status is concentrated in the number of hours worked.

In 1960, nearly two-thirds of households in the bottom quintile of income were headed by individuals who worked—primarily married fathers. By 1991, this figure had fallen to around one third, and only 11 percent of these households were headed by someone who worked full-time throughout the year.

The total number of hours worked in married households has increased significantly over the past 40 years. According to former Congressional Budget Office Director June O’Neill, in 1950 only 18 percent of married mothers with children under 18 worked outside the home. By 1975, 41 percent of married mothers worked and that proportion reached 64 percent in 1992. Yet mothers on welfare appear to work little—only 7 percent report any employment. These data were collected before the enforcement of the Welfare Reform Act (1996).

Not only are those in the lowest quintile generally working fewer hours than their counterparts were in the 1950s and 1960s, but they are doing so despite a national family trend in the rising number of hours worked.

A reverse trend accompanies the disappearance of marriage: The number of hours worked in the family household declines. Present-day single heads of households are working fewer hours than the married heads of poor households in the 1950s (typically, married men). At the same time, married couples are increasing the total number of hours worked, and although there are some unwelcome consequences from this increase in working hours in married households, there is no doubt that it has increased the number of families exiting a life of poverty.

1) Precision has no formal meaning. It indicates how clearly determinable (distinguishable from zero) an influence on an outcome is. Precision is comparable to standard deviation. Low/ no precision indicates a high standard of deviation in which data points spread over a large range of value, signifying that the influence of one variable over another is relatively uncertain. High precision indicates a low standard of deviation in which data points hover around the mean, signifying that the influence of one variable over another is relatively certain. For further elaboration see “Marriage and Economic Well-Being: The Economy of the Family Rises or Falls with Marriage”
2) Henry Potrykus and Patrick Fagan, “U.S. Social Policy Dependence on the Family, Derived from the Index of Belonging,” (2013) available at http://marri.us/policy-2013.
3) This is interesting given the emphasis by some on education when the health of the American workforce is discussed. Family, heretofore ignored in this discussion, must be given equal footing. If we are concerned with the American workforce, the family must be discussed side-by-side with education. See, again, Heckman and Masterov, Cunha and Heckman, “Formulating, Identifying and Estimating the Technology of Cognitive and Noncognitive Skill Formation,“ and Armor.
4) See Table 2 in Henry Potrykus and Patrick Fagan, “U.S. Social Policy Dependence on the Family,” (January 2013) available at http://downloads.frc.org/EF/EF13B33.pdf for a list of the controls and other methodological considerations.
5) The fraction of intact families in a geographic area has the largest consistently positive influence on average earnings under a model without the dependency ratio controls. Hence, we assert that this is a “family effect” most properly attributable to care for dependents. Endogeneity is partially broken by our multi-model specification.
6) Hence, this may be a phenomenon of where the minorities tend to work and not a function of minorities' pay levels themselves. This is a subtle matter touching on the ecological inference problem.
7) Susan J. Popkin, “Welfare: Views from the Bottom,” Social Problems, Vol. 37, No. 1 (February 1990), pp. 64-79.
8) M. S. Hill and G. Duncan, “Parental Family Income and the Socioeconomic Attainment of Children,” Social Science Research, Vol. 16 (1987), pp. 39-73.
9) , 10) , 14) Martha S. Hill and Jodi R. Sandfort, “Effects of Childhood Poverty on Productivity Later in Life: Implications for Public Policy,” Children and Youth Services Review, Vol. 17, Nos. 1 & 2 (1995), pp. 91-126.
11) Hsiang-Hui, Daphne Kuo,and Robert M. Hauser, “Trends in Family Effects on the Education of Black and White Brothers,” Sociology of Education, Vol. 68 (1995), pp. 136-160.
12) Darin R. Featherstone, Bert P. Cundick, and Larry C. Jensen, “Differences in School Behavior and Achievement Between Children from Intact, Reconstituted, and Single-Parent Families,” Adolescence, Vol. 27, No. 105 (1992), pp. 1-12.
13) Herbert Zimiles and Valerie E. Lee, “Adolescent Family Structure and Educational Progress,” Developmental Psychology, Vol. 27, No. 2 (1991), pp. 314-320.


This entry draws heavily from How Broken Families Rob Children of their Chance for Future Prosperity.