Journal Issue: Children and Divorce Volume 4 Number 1 Spring/Summer 1994
The Economic Consequences of Divorce
A growing body of literature has developed around the economic consequences of divorce.2 While this literature is diverse in terms of data, definitions employed, and analytic strategies, a number of conclusions can be reached. First, women and children experience a significant decline in income following divorce. Second, men are much less likely to experience a decline in income following divorce. Third, the event most associated with a rebound in economic well-being following divorce for women and children is remarriage.
Beyond these simple points, though, it is difficult to draw firm conclusions about the magnitude and duration of the economic consequences of divorce, the diversity of economic outcomes, and the reasons for these outcomes. Below we attempt to provide as clear a portrait as possible, while recognizing the need for additional research to provide more complete answers. We do so while discussing some of the most recent data on the economic consequences of divorce for women and children.
One reason we know so little about the economic consequences of divorce is that the data needed to answer our questions are not available. What is needed are longitudinal data on a large number of men, women, and children (families) that are representative of the United States population—data that contain information on the marital transitions experienced by these families, as well as precise information about the economic situation of spouses and children for relatively fine intervals of time. Moreover, these data should span a broad period of time so that both the short-term and the long-term consequences of divorce can be considered.
Unfortunately, such data do not exist. Consequently, all research on the economic consequences of divorce involves tradeoffs concerning data content. For example, the Census Bureau's Current Population Survey (CPS) provides the most current, nationally representative data on divorced female-headed families. Table 1 presents an overview of their economic situation. CPS data indicate that divorced women with children have a high likelihood of living in poverty: 39% of all divorced women with children and 55% of those with children under six were poor in 1991. (Poverty thresholds by size of family and number of children are provided in Table 2.) From Table 1 it is also evident that, although the average amount of child support received is only $3,143, child support payments comprise almost one-fifth of the total income of divorced mothers with children. Unfortunately, no detailed data are available for father-headed families. However, in 1991, only about 4% of single-parent families were headed by fathers.3
Although CPS data are informative, they provide only a snapshot or cross-sectional picture of divorced families. To better understand the economic dynamics of divorce, longitudinal data are needed. Most of the results we choose to report come from the U.S. Census Bureau's published reports based on the 1984 Survey of Income and Program Participation (SIPP).
The 1984 SIPP is a sample of over 20,000 households first interviewed in October of 1983 and interviewed every four months thereafter for a period of two and one-half years.4 Thus, a large number of households are followed for up to eight panels (four-month periods), or 32 months. While the SIPP cannot provide information on the long-term consequences of divorce, detailed monthly information on sources of income and marital status make it one of the best available sources of longitudinal information on the short-term consequences of marital disruption.5
Table 3 presents four measures of economic well-being based on data from the 1984 SIPP; mean family income, mean per capita family income, mean ratio of family income to the poverty level, and percent of families with incomes below the poverty level.6 The first measure is simply the average monthly income available to a family. The second measure adjusts for the fact that a given income must be shared by a variable number of family members. The third measure relates family income to the poverty level.
Different poverty thresholds are calculated for families of different size and composition. These thresholds rise gradually with increasing family size, reflecting the fact that it is not necessary to increase income proportionately to increases in family size to maintain a family's standard of living as family size increases. Hence, measures three and four implement control for family size but less dramatically than measure two. A ratio of 1.0 indicates that a family is at the poverty line. Ratios above and below 1.0 indicate the degree to which family income is above or below the poverty line. The fourth measure indicates the proportion of families living in poverty. We use four different measures of economic well-being because, although they are all interrelated, they each provide a slightly different perspective on the magnitude of income loss following divorce. This varied perspective is important when there is no commonly accepted cutoff for determining what constitutes a significant change in income.7
The data from SIPP indicate that, on average, children were better off economically at the end of the SIPP sample period (1986) than they were at the initial interviews in 1983-84 (see Table 3). Mean monthly family income increased by about 6.9%, mean per capita family income increased 6.1%, mean ratio of family income to the poverty level increased 6.5%, and the percent of families with incomes below the poverty line decreased 12.3%. For children whose fathers leave, however, the picture is much different. Family income dropped by about 23% (a figure that is consistent with many of the estimates reported in prior literature).8 Per capita family income fell, but only by about 8%. The ratio of family income to the poverty level also dropped (about 13%), but the average family remained above the poverty line (approximately 69% had income-to-needs ratios above the poverty level). Note, however, that the percent of families below the poverty line increased by nearly 10 percentage points (about a 46% increase in the total number of families).
It might seem odd that the ratio of family income to the poverty level drops relatively little yet there is such a substantial increase in the percent of families with incomes below the poverty line (a point that illustrates the need to view the data in various ways). This seeming discrepancy is generated by variation in economic well-being preceding father absence. As indicated in prior studies, women and children who are in higher-income families prior to disruption subsequently suffer the most substantial decline in income.9 Yet, the reduction in income experienced in these families often is not sufficient to leave them below the poverty line following divorce.
The substantial increase in the percentage of families below the poverty line following disruption results from the large fraction of families close to the poverty line prior to disruption. Compared to children in continuously married families, children who lived in families that experienced marital disruption were economically disadvantaged even before their fathers left. Mean monthly family income was 17% lower ($2,346 versus $2,834), and the percent of families below the poverty line was 1.75 times greater (21.3% versus 12.1%). This difference is consistent with an extensive body of research which indicates that economic stress and deprivation are positively associated with subsequent marital dissolution.10
The economic vulnerability of families without fathers present is further indicated by the desperate financial circumstances of children who lived the entire length of the 1984 SIPP in mother-only families. While these families did not experience any decrease in their financial well-being over the period covered, they remained the group least well-off. More than 50% of the mother-only families were in poverty, and the mean ratio of family income to the poverty level was only slightly greater than 1.0.
The last panel of data in Table 3 indicates the role a father's income can play in reducing economic distress associated with mother-only families. Family income increased 115% for families into which a father entered, and the increase remains substantial when considering both income per capita (about 50%) and the mean ratio of family income to the poverty level (about 90%). As might be expected, the notable increase in family income associated with gaining a father substantially reduces the incidence of poverty (62%).
When data are considered separately by race, both blacks and whites experience a significant decline in income following marital disruption. Using data from the Panel Study of Income Dynamics (PSID), Corcoran reported that black mothers and white mothers experience a similar decline in the percent change in family income from disruption (comparing the year prior to marital disruption with the year following marital disruption).11 However, because intact black families generally have lower income levels, a greater proportion of black women fall below the poverty line following marital disruption. Also using the PSID, Duncan and Hoffman found that black women who had been relatively well-off during marriage experienced larger percentage declines in income following divorce than comparable white women.12
Not surprisingly, the composition of all families is not equal to the composition of families living in poverty. For example, female-headed families with children comprised 48% of families living in poverty in 1990, but constituted only 12% of all families.13 Conversely, married couples with children comprised only 28% of poor families, while they made up 38% of all families. Over the past 30 years, these two types of families have exchanged places as the predominant type of poor family. As the percentage of two-parent families has declined with respect to female-headed families, female-headed families have increased as a percentage of all poor families.
Eggebeen and Lichter examined shifts in family structure from 1960 to 1988 with respect to changes in child poverty.14 They found that, had the proportions of children in married-couple, female-headed, and male-headed households remained the same as in 1960, the child poverty rate in 1988 would be one-third less than the child poverty rate actually observed. Analyses conducted by race indicate that changes in black family structure during the 1980s were responsible for 65% of the increase in poverty of black children, while changing family structure among whites accounted for 37% of the rise in official child poverty.
In addition to being more likely to be living in poverty at a given point in time, divorced women with children are likely to remain poor longer. Bane and Elwood15 found that, during the late 1970s, female-headed families (both ever and never married) experienced an average poverty spell of seven years, in contrast to a less-than-two-year spell experienced by others in poverty.16 Thus, poverty among families headed by divorced women is not only more prevalent than for two-parent families, it is also more likely to be chronic.
While the data in Table 3 are informative, they are limited by the fact that fathers can leave during any of the eight panels. Thus, one cannot interpret the income levels at time 1 and time 8 in a longitudinal framework for this important group—that is, one cannot obtain a firm, unambiguous idea about how income might change across time following a father's departure. Table 4 presents a longitudinal representation of the data for families experiencing marital disruption. For children who lived with both parents at the beginning of the 1984 SIPP, the four measures of economic well-being used in Table 3 are shown for five points in time: the panel immediately prior to the father's leaving and the four panels immediately following the father's departure (or up to about 12 months after departure).17 Two different groups are considered. The first group consists of the families of all children who experienced a marital disruption, and the second consists of the families of children who experienced a marital disruption and whose mothers did not remarry or reconcile.
The data in Table 4 show a decline in the income available to families after a father departs, a decline that appears immediately after marital disruption. In particular, note that the percent of families in poverty nearly doubled from the panel immediately prior to marital disruption to the panel immediately following dissolution. If one compares the two sets of figures (for all children and children whose mother did not remarry/reconcile), it is evident that, for the first year following marital disruption, there is no improvement in the economic well-being of families unless the mother remarries or reconciles.18 Income at time 4 is no higher than at time 1 for women who have not remarried or reconciled, and the proportion of families below the poverty line remains at about 35%.
The data in Table 4 show that the economic consequences of marital disruption are not limited to the immediate turmoil surrounding a divorce and that, without remarriage or reconciliation, there is no clear trend toward improvement in economic well-being for at least the first two to three years after divorce. While these data cannot be construed to provide information on the long-term economic consequences of marital disruption, other research suggests that economic deprivation may be long-term, up to five years or more.19 In large part, however, the sample sizes in these studies tend to be quite small, making it difficult to draw firm conclusions.
As indicated above, remarriage or reconciliation is the most immediate mechanism alleviating economic deprivation associated with marital disruption. Over the past decade, though, rates of remarriage have declined (from a high of 166 remarriages per 1,000 divorced or widowed women age 15 to 54 in 1966–1968 to 109 remarriages per 1,000 similar women in 1987–1989).20 Lower remarriage rates would generally result in longer periods between marriage and thus longer periods of poverty. However, the recent rise in the cohabitation rate of single-parent families and the income sharing this may imply may bring about some economic relief.21
The substantial difference in economic well-being between disrupted families according to whether the mother remarries suggests that the employment of divorced mothers is not a particularly effective buffer against economic deprivation. Evidence to that effect is given in Table 5, again using data from the 1984 SIPP. The difference is probably partly attributable to the facts that most fathers have had prior continuous work experience, whereas at least some proportion of mothers have not, and that most mothers have lower-status jobs and are victims of wage discrimination.22
Following the presentation in Table 4, employment and income figures are given for the panel immediately prior to marital disruption and for the four panels immediately subsequent to the father's departure (covering approximately two years). Two sets of figures are given: one for all children and one for children whose mothers did not remarry or reconcile.
Prior to marital disruption, about 56% to 58% of children had mothers who were in the labor force and who earned about $930 per month, or about 37% of family income (using family income values reported in Table 4). In the first panel after marital disruption, more than 70% of mothers were working. A first response of many newly divorced mothers to economic stress, therefore, appears to be entry into the labor force. Note, however, that the average amount earned declined from immediately before to immediately after disruption, although the amount earned constituted about 60% of postdisruption family income. This decline occurred because the jobs available to many new entrants into the labor force are often less than full-time employment and/or pay wages below those earned by mothers already in the labor force.
Over time, the proportion of children whose mothers were in the labor force declined to a point only slightly higher than observed just prior to marital disruption. In addition, average income increased as the proportion of mothers working declined. While part of the increase in earnings may be attributed to annual pay increases (in real dollars), it is likely that most of the increment is associated with the fact that many women earning low wages elect to leave the labor force. Thus, while the initial response to economic uncertainty following marital disruption may be to find employment, for many women this shift is short-lived. In the long run, only those women who are successful in the labor market are likely to remain employed.
Because many women are new labor force entrants or are returning to work after being absent for some time, it is difficult for them to find jobs that pay enough to support a family. Many of these women must take shift work to find employment or to be home with their children at least part of the time. Presser reports that unmarried mothers are nearly twice as likely as married mothers to work non-day shifts, often in jobs that are unpleasant and pay the minimum wage.23 When combined with the costs and constraints of child care and discrimination against single parents in the workplace, low-paying, dead-end jobs force many working single mothers out of the labor market.24 The availability of welfare benefits, especially those that erode with earned income, acts to reduce the likelihood that divorced mothers will remain active in the labor force.25