This is an old revision of the document!

Effects of Single Parents on Poverty

Synthesis Paper: The Effects of Marriage on Child Poverty

The 1960s War on Poverty was intended to eliminate child poverty nationwide through a variety of income transfers and human development programs. However, during the first three decades of the War on Poverty, there was little net decline in the child poverty rate. The lack of progress in reducing child poverty since 1965 can be explained in part by the erosion of marriage and the growth of poverty-prone single-parent families.

What share of the current level of child poverty in the United States can be attributed to the growth of single parenthood since the 1960s? Or, phrased in a slightly different way: What would the child poverty rate be today if single-parent families had remained at the levels that existed before the beginning of the War on Poverty? To answer the question, researchers attempted to simulate the effects of higher marriage rates using data from the U.S. Census Bureau for the year 2000.1)

1. Simulation Study

To measure the effect of single parenthood on poverty rates, researchers paired 7.6 million single parents with similar non-married persons within the Current Population Survey (CPS) to form 7.6 million new simulated “marriages.” In creating these simulated marriages, researchers attempted to mimic socially realistic marriage patterns. Single mothers were paired with non-married men who were identical in race, age, and level of education.2) Similarly, single fathers were paired with non-married women who were identical in race, age, and level of education.

When a simulated married family is created, a new total family income variable is also created by combining the original incomes of the two “partners.” Any cash welfare assistance is subtracted from the total money income of the newly created family, since most individuals lose their cash welfare/ Temporary Assistance to Needy Families (TANF) assistance benefits when they marry. The analysis next takes into account the large differences in earnings between married and non-married men ( CPS data reveals that married men, on average, earn 40 percent more than non-married men who are similar in race, age, and education).

After simulating a higher level of marriage in society, we calculated the effect of increased marriage on child poverty—first by measuring poverty based on a family’s “money income,” and secondly based on a broader definition of income that includes the Earned Income Tax Credit (EITC) and the value of food stamps.

1.1. Effects of Restoring Marriage to 1960 Levels

As Chart 1 shows, the proportion of children in single-parent families rose substantially from1960 to 2000. In 1960, only 11.7 percent of children were in single-parent or other “broken” home settings, a percentage that ballooned to 27.6 percent by the year 2000.

As Table 1 and Table 2 show, if marriages in 2000 were restored to 1960 levels, the percentage of children residing in married couple families would rise from 72.4 percent to 88.3 percent. A total of 11.49 million children would reside in married couple families rather than single-parent homes (See Table 3 and Table 4). Among black children, the share residing in married homes would rise from 51 percent to 71 percent; nearly 3 million additional black children would reside in married-couple homes.3)

Table 5 shows that restoring marriage to 1960 levels would have a considerable impact on child poverty. Using the money income definition of income, the child poverty rate is now at 15.7 percent. However, if marriage were restored to 1960 levels, the total child poverty rate would be cut by nearly a third to 11.2 percent. Restoration of marriage to 1960 levels among blacks would have a similar effect; their child poverty rate would fall by nearly a third from the current level of 27.5 percent to 20.5 percent.

1.2. Effects of Marriage on Children

As noted, restoration of marriage to 1960 levels would move 11.49 million children from single-parent homes into married-couple homes. The reduction of poverty within this group of children is striking. As residents in single-parent homes, the current poverty rate of these children is 34.2 percent. This analysis matched single parents with demographically similar spouses to create simulated married-couple homes. Table 6 shows that, when the 11.49 million children are placed in these simulated married-couple homes, their poverty rate falls to 6.7 percent. Thus, marriage cuts poverty full 80 percent (from 34.2 percent to 6.7 percent). Among the 11.49 million children moved into married families, about 3 million are black children. The poverty rate among these children is 45.4 percent before marriage and 9.0 percent after marriage; thus, marriage reduces poverty among the affected black children by 80 percent.

1.3 Effects of Marriage on Poverty Using an Expanded Definition of Income

As the final step in this analysis, researchers calculated the effects of marriage on poverty using an expanded definition of income. The Earned Income Tax Credit and food stamps were counted as part of a family’s financial resources when determining whether the family was poor.4) Naturally, poverty rates both before and after the marriage simulation were lower according to this method of measurement.

Using the expanded definition of income, the current child poverty rate is 12.9 percent. If marriage were restored to 1960 levels, the rate falls to 9.1 percent. A total of 2.7 million children would be removed from poverty. Among black children, the poverty rate falls from 22.6 percent to 16.3 percent, a drop of 28 percent (See Table 7 and Table 8).

2. Prior Research

These results are consistent with other microsimulation models of marriage and poverty. In an influential 1995 article entitled “The Impact of the Changing U.S. Family Structure on Child Poverty and Income Inequality,” Dr. Robert Lerman of the Urban Institute estimated that restoring marriage to 1971 levels would have reduced child poverty in 1989 by approximately 23 percent.5) This analysis6) differs from that done by Dr. Lerman in two respects. First, due to the difference in years of comparison (2000 to 1960 rather than 1989 to 1971), this analysis moved more children from single-parent to simulated married homes. Second, Dr. Lerman’s analysis employed a larger adjustment in male wages due to marriage than we did.

More recently, Adam Thomas and Isabel Sawhill of the Brookings Institution simulated marriage rates similarly by restoring marriage to 1970 levels.7) Using money income to gauge poverty, the Thomas and Sawhill simulation would drop the child poverty rate from 18.3 percent to 14.8 percent, a 19.1 percent change. As noted above, the analysis drops the child poverty rate from 15.7 percent to 11.2 percent, or nearly a 30 percent change. The difference between this analysis and theirs can be explained in part by the fact that Brookings raised marriage to 1970 levels, whereas this analysis raises marriage to the higher 1960 levels; thus, this study moved more children from single-parent to married families. In addition, the Brookings analysis made no adjustment at all for the effects of marriage on male wages, through either a selection effect or earnings inducement. As a consequence, Thomas and Sawhill underestimate the impact of marriage on poverty.

1) For a complete analysis of the methodology used, see “The Effects of Marriage on Child Poverty”
2) In matching potential spouses by race, three categories were used: white, black, and other.
3) In all the tables and accompanying text in “The Effects of Marriage on Child Poverty,” the term “black” refers to “non-whites.” “Non-whites” is a category that covers blacks and a smaller group of “others.” In 1960, the Census Bureau divided the population into “whites” and “non-whites”; no separate data on blacks were available. In hypothetically restoring marriage to 1960 levels, we have been forced to mimic this 1960 racial categorization. The use of the categories “whites” and “non-whites” rather than “whites” and “blacks” has no effect on the overall analysis of the impact of marriage on child poverty and marginal effect on the figures given on racial subgroups.
4) Many researchers view this augmented definition of income as superior to traditional money income because it includes cash (EITC) and near cash (fungible food stamp) transfers from the government to families. See, for example, Gregory Acs, Norma Coe, Keith Watson, and Robert Lerman, “Does Work Pay: An Analysis of the Work Incentives Under TANF,” Urban Institute Occasional Paper No. 9, July 1998.
5) Lerman, “The Impact of the Changing U.S. Family Structure on Child Poverty and Income Inequality,” Table 10, p. S136.
7) Adam Thomas and Isabel Sawhill, “For Richer or for Poorer: Marriage as an Antipoverty Strategy,” Journal of Policy Analysis and Management, forthcoming, available at