Study: Child custody policies and divorce rates
Richard Kuhn, Children's Rights Council, Washington, D.C.
John Guidubaldi, D.Ed., John Carroll University (Cleveland, OH)
and Kent State University (Kent, OH)
11th Annual Conference of the Children's Rights Council October 23-26, 1997. Washington, D.C. Copyright 1997, Children's Rights Council. Reproduction with attribution is permitted.
This paper compares divorce rate trends in the United States in states that encourage joint physical custody (shared parenting) with those in states that favor sole custody. States with high levels of joint physical custody awards (over 30%) in 1989 and 1990 have shown significantly greater declines in divorce rates in following years through 1995, compared with other states. Divorce rates declined nearly four times faster in high joint custody states, compared with states where joint physical custody is rare. As a result, the states with high levels of joint custody now have significantly lower divorce rates on average than other states. States that favored sole custody also had more divorces involving children. These findings indicate that public policies promoting sole custody may be contributing to the high divorce rate. Both social and economic factors are considered to explain these results.
Empirical evidence shows that children raised by a divorced single parent are significantly more likely than average to have problems in school, run away from home, develop drug dependency, or experience other serious problems (e.g., Amato and Keith, 1991; Guidubaldi, Cleminshaw, Perry, and McLoughlin,1983; Hetherington and Cox, 1982). Although many single parent families are created as a result of unwed motherhood, far more are the result of divorce. Of 18.6 million children in the United States living with only one parent, approximately two thirds are with divorced or separated parents (Census, 1994). This paper investigates the relationship between child custody policies and changes in the U.S. divorce rate, using data from a 19 state sample collected by the National Center for Health Statistics of the Centers for Disease Control.
States differ widely in their policies toward joint custody. Many states routinely grant joint legal custody, which gives the non-residential parent the right to participate in major decisions about the children's upbringing and to view certain records. Joint legal custody does not affect the child's living arrangements. Often it is granted with the traditional residence arrangement, in which the child lives with one parent but visits the other parent four days per month.
Less commonly, joint physical custody is awarded. With joint physical custody (also called shared parenting), the child lives with both parents, often on an alternating week basis. Joint physical custody is usually defined as a schedule where the child has at least a 30/70 time share between parents, although 50/50 arrangements are a common form of shared parenting (Ricci, 1981). Some form of joint custody is a preference or presumption in a few states, while in some other states with no preferred custody option, judges have favorable attitudes toward joint custody and frequently grant it. For the 19 states in the NCHS sample, the average rate of joint physical custody awards in 1990 was 15.7%, and in two states joint physical custody was awarded in nearly half of the cases.
State policies on joint custody have changed significantly in the past 25 years. Because of maternal preference policies, joint custody was unusual before 1970, although divorced families in times past sometimes worked out arrangements that were equivalent to modern joint custody (Ricci, 1981). For example, the Maryland Court of Appeals considered a case in 1934 in which the division of time between parents was equivalent to joint physical custody (McCann v. McCann), although the term joint custody had not yet been invented. As maternal preference laws were found to violate the 14th Amendment guarantee of equal protection under the law in the 1960s and 1970s (Roth, 1976), joint custody began to increase. Although much has been written about links between "no-fault'" divorce laws and the divorce rate, there has been little discussion of the effect of child custody policies on the divorce rate.
Custody Policies and the Divorce Rate
It might be argued that joint custody could encourage divorce, by making divorce "easier." On the other hand, widespread acceptance of joint physical custody might be expected to reduce the divorce rate, because joint custody makes it difficult for an angry parent to hurt the other by taking away the children, or to relocate and thereby eliminate interaction with the other parent. In addition, an economic argument has been advanced that high levels of child support associated with sole custody may encourage divorce, because custody of children represents an asset for the custodial parent to the extent that child support payments exceed the cost of raising a child (Muhtaseb, 1995). Because joint physical custody results in a more equal division of parenting time, child support payments may be lower, although there are still payments unless both parents have the same income. States that more frequently award joint physical custody may thus see a decline in the divorce rate. To date, no study has provided empirical evidence to support either hypothesis about the effect of joint custody policies on the divorce rate.
State divorce rates and other vital statistics are maintained by the National Center for Health Statistics (NCHS), a division of the Centers for Disease Control, U.S. Department of Health and Human Services. The divorce rate measure used is the number of divorces per thousand population. A 1995 NCHS report (Clarke, 1995) gives data on physical custody awards for 19 participating states for the years 1989 and 1990. This NCHS report is the first of its kind to report figures for physical custody of children. Values given are percentages of sole custody father, sole custody mother, and joint custody awards. Figures for 1989 and 1990 are given, separated by a "/". In some cases the total may be slightly less than 100% because awards to persons other than mother or father (generally from 0 to 2% in the NCHS report) are not included in Table 1. More recent data are not yet available. Table 1 shows the physical custody awards for these states. The definition of joint physical custody used in the NCHS study is a minimum of 30% time share with each parent (Clarke, 1996). Figures for 1989 and 1990 are similar, although the percentages for joint custody are slightly higher in 1990 for those states reporting both years. For five states, 1989 figures were not available; these are indicated as "NA". States were divided into categories of High (above 30%), Medium (10% to 30%), or Low (below 10%) levels of joint physical custody awards, as shown in Table 1.
Montana 8.1/8.4 47.8/46.4 43.3/44.0 High Kansas 7.8/6.8 50.1/47.2 39.5/43.6 Connecticut 5.3/5.3 58.7/58.1 35.8/36.4 Idaho 9.8/10.4 57.9/55.3 31.9/33.2 Rhode Island NA/5.4 NA/62.2 NA/31.7 Alaska NA/14.2 NA/63.1 NA/19.5 Medium Vermont NA/10.6 NA/71.4 NA/17.1 Illinois 8.7/9.2 77.4/75.4 13.7/15.1 Wyoming 11.0/9.5 73.0/74.4 14.1/15.1 Missouri 10.4/11.0 74.4/73.1 14.0/14.8 Oregon 10.7/12.6 74.1/71.7 14.9/14.0 Michigan 9.5/11.2 76.4/73.9 12.5/14.2 Virginia NA/11.6 NA/70.9 NA/13.8 Pennsylvania 10.5/10.0 78.6/76.7 9.4/10.1 Low Utah 10.5/9.7 79.3/81.1 10.1/9.0 Tennessee 11.1/11.3 78.9/78.9 8.1/8.6 Alabama 9.7/10.7 79.5/80.2 9.3/8.6 New Hampshire 12.2/11.0 79.9/80.4 6.6/7.1 Nebraska NA/12.2 NA/81.3 NA/4.1 Table 1. Physical Custody Awarded (percent), 1989/1990 Source: Monthly Vital Statistics Report, Vol. 43, No. 9 (March 22, 1995), National Center for Health Statistics.
Findings and Discussion
Divorce rates for 1989, 1990 and 1991 were compared with 1993, 1994 and 1995 levels, as shown in Table 2. Comparisons between basal values of 1989/1990/1991 and values for 1993/1994/1995 are used rather than absolute values in order to factor out differences that may be unrelated to custody policies. For example, states differ in their ethnic, religious, and racial compositions, factors that can affect the divorce rate. The effect of custody policies can be more precisely isolated by using differences across time, just as the effect of a medication is isolated by comparing before and after treatment values for subjects whose initial (and final) values for blood pressure, heart rate, or other measures may be significantly different. Initial values and values four years later for the state groups are shown in Table 3. Table 3 also shows 1980 divorce rate averages for the three groups. Joint custody had begun to emerge as a custody option in 1980, although its adoption into state policies occurred at different points. Rate changes between 1980 and 1990, therefore, are likely to contain some effects of policies regarding joint custody. Note that the High and Medium joint custody groups had very similar divorce rate declines between 1980 and 1994 (by approximately 1.1 and 1.2 per thousand respectively), while the states with low levels of joint custody had a decline of only 0.4 per thousand between 1980 and 1994.
----------- Divorce Rates by Year----------- State 1989 1990 1991 1993 1994 1995 93-89 94-90 95-91 Change Montana 5.1 5.1 5.4 5.1 4.9 4.8 0 -.2 -.6 Kansas 5.0 5.1 5.5 4.8 4.7 4.2 -.2 -.4 -1.3 Connecticut 3.7 3.5 3.5 3.1 2.8 2.9 -.6 -.7 -.6 Idaho 6.3 6.4 6.3 6.3 6.2 5.8 0 -.2 -.5 Rhode Island 3.6 3.7 3.3 3.4 3.2 3.7 -.2 -.5 +.4 -.37 Alaska 6.3 5.7 6.4 5.3 5.5 5.0 -1.0 -.2 -1.4 Vermont 4.5 4.4 4.6 4.8 4.0 4.8 +.3 -.4 +.2 Illinois 4.0 4.0 4.0 3.7 3.7 3.3 -.3 -.3 -.7 Wyoming 6.6 6.9 7.0 6.5 6.5 6.7 -.1 -.4 -.3 Missouri 4.9 5.0 5.1 5.1 5.0 5.0 +.2 0 -.1 Oregon 5.4 5.5 5.2 5.3 5.3 4.8 -.1 -.2 -.4 Michigan 4.4 4.4 4.3 4.1 4.1 4.2 -.3 -.3 -.1 Virginia 4.2 4.4 4.5 4.5 4.6 4.4 +.3 +.2 -.1 -.23 Pennsylvania 3.2 3.4 3.3 3.3 3.3 3.3 +.1 -.1 0 Utah 4.8 5.2 4.8 4.8 4.7 4.6 0 -.5 -.2 Tennessee 6.6 6.6 6.5 6.5 6.6 6.3 -.1 0 -.2 Alabama 6.2 6.3 6.5 6.5 6.2 6.1 +.3 -.1 -.4 New Hampshire4.5 4.5 4.3 4.5 4.4 4.2 0 -.1 -.1 Nebraska 4.0 4.1 4.0 3.9 4.0 3.8 -.1 -.1 -.2 -.10 Table 2. Divorce Rates and Four-year Difference in Rates Sources: Monthly Vital Statistics Report, Vol. 43, No. 13 (October 23, 1995), National Center for Health Statistics. Statistical Abstract of the United States, 1993.
As can be seen from Table 2 and Table 3, the states with high levels of joint custody had significantly lower divorce rates four years later. States with higher levels of joint custody had an average four-year decline in the divorce rate approximately double that for states with medium levels of joint custody. On a percentage basis, between 1989 and 1994 the rate in the High joint custody group declined by 8%, in the Medium group by 4%, and in the Low group by less than 1%.
Joint Custody ---------- Year -------------------- Level 1980 1989 1990 1993 1994 High 5.42 4.74 4.76 4.54 4.36 Medium 6.06 5.04 5.04 4.94 4.84 Low 5.25 4.88 5.02 4.92 4.87 Table 3. Changes in Divorce Rates Over Time
Figure 1 shows joint custody awards and divorce rate changes for the 19-state NCHS sample. States are ordered by level of joint custody awards in 1990. As joint custody awards increase, states in general have greater declines in divorce rates. Figure 2 summarizes the changes in divorce rates for states in the three joint custody categories. Statistical analysis shows that the correlation between joint physical custody and reduced divorce is almost certainly not due to chance fluctuation. The statistical measure used is a correlation of the average of joint custody awards per state in 1989 and 1990 with the average decline in divorce rate from 1989 through 1991 to 1993 through 1995 (i.e., difference between the average of 1993, 1994, and 1995 rates and the average of 1989, 1990, and 1991 rates.) This is the average of the "Joint" column of Table 1 correlated with the difference between the average of 1993 to 1995 rates and 1989 to 1991 rates in the "Divorce Rates by Year" column of Table 2. There is less than a five percent probability that this correlation is due to chance (correlation coefficient r = .47, p < .05). (Note: Wisconsin reported numbers in 1989 but not in 1990, so it was not included in this analysis. However, separate calculations show that inclusion of the Wisconsin data does not affect the statistical significance of the results.)
One possible explanation to consider for the difference in divorce rates between high and low joint custody states is an effect resulting from changes in marriage rates. If marriage rates per thousand population increase, then divorce rates per thousand population in following years can increase if marriages fail at the same rate. Similarly, divorce rates can decrease during a particular period if marriage rates decreased in previous years, because fewer marriages were created. Thus it is important to look at whether the greater decline in divorce rates in high joint custody states during the early 1990s results from a decrease in marriage rates during the early 1980s. Table 4 shows the change in marriage rates between 1980 and 1985, a decade before the period under study.
- Marriage Rates -- Chang State 1980 1985 per 1000 per 1000 percent Montana 10.6 8.7 -1.9 Kansas 10.5 9.5 -1.0 Connecticut 8.4 8.6 + .2 Idaho 14.2 12.2 -2.0 Rhode Island 7.9 8.3 + .4 -.86 -2.7% Alaska 13.3 11.8 -1.5 Vermont 10.2 10.4 + .2 Illinois 9.6 8.5 -1.1 Wyoming 14.6 10.6 -4.0 Missouri 11.1 9.8 -1.3 Oregon 8.7 8.3 - .4 Michigan 9.4 8.7 - .7 Virginia 11.3 11.7 + .4 -1.05 -9.6% Pennsylvania 7.9 7.5 - .4 Utah 11.6 10.6 -1.0 Tennessee 12.9 11.5 -1.4 Alabama 12.6 11.5 -1.1 New Hampshire 10.0 11.4 +1.4 Nebraska 9.1 7.9 -1.2 -1.21 -6.3% Table 4. Change in Marriage Rates Source: Statistical Abstract of the United States.
If the greater decline in divorce rates for High joint custody states results from declining marriage rates in previous years, then we would expect marriage rates for these states to show larger decreases in the early 1980s than the Low joint custody states. As can be seen from Table 4, the reverse is true. The low joint custody states actually had greater declines in marriage rates during the early 1980s. If marriages continued to fail at the same rate during the decade, then these states should also show greater declines in divorce rates during the early 1990s. The fact that they did not suggests that other factors may be at work. It is not reasonable to conclude that the decrease in divorce rates associated with joint custody is simply a result of declines in marriage rates. A second explanation proposed here considers both social and economic factors.
Before the 1960s, social pressures and legal requirements made divorce relatively uncommon in the U.S. Divorce typically required grounds severe enough that a reasonable person could not expect the marriage to continue: adultery, desertion, abuse, insanity or imprisonment of a spouse. With a few exceptions, states adopted unilateral "no-fault'" divorce laws in the 1960s and 1970s, which allowed a spouse to abandon a marriage without traditional grounds. Divorce was actually encouraged by some as an antidote to boredom, or for other reasons that might have been considered frivolous a generation before. About 80% of U.S. divorces today result from the unilateral decision of one spouse, rather than the joint decision of both (Gallagher, 1996), with the spouse who files for divorce first often having an advantage.
If one investigates the simple question, "who initiates divorce," we find from the Monthly Vital Statistics Report May 21, 1991 (NCHS, 1991), that from 1975 to 1988, in families with children present, wives file for divorce in approximately 2/3 of the cases each year. In 1975, 71.4% of the cases were filed by women, and in 1988, 65% were filed by women. While these statistics alone do not compel a conclusion that women anticipate advantages to being single, rather than remaining in the marriage, they do raise that reasonable hypothesis. If women can anticipate a clear gender bias in the courts regarding custody, they can expect to be the primary residential parent for the children. If they can anticipate enforcement of financial child support by the courts, they can expect a high probability of support monies without the need to account for their expenditures. Clearly they can also anticipate maintaining the marital residence, receiving half of all marital property, and gaining total freedom to establish new social relationships. Weighing these gains against the alternative of remaining in an unhappy marriage may result in a seductive enticement to obtain a divorce, rather than to resolve problems and remain married.
States that favor sole custody in divorce may thus expect to see more divorce than states that encourage joint custody. On a practical level, joint physical custody makes it less likely that a parent can move to another city to eliminate interaction with the other parent. Because both parents provide for the child directly, child support payments may be somewhat lower with joint custody, reducing financial motives for divorce. Perhaps most significant, joint custody also removes the capacity for one spouse to hurt the other by denying participation in raising the children. The correlation between joint custody and reduced divorce may have a simple explanation. If a parent considering a divorce is told by an attorney that a judge will probably not permit him or her to relocate with the children, and that the other parent will continue to be involved, he or she may decide that it is easier to work out problems and remain married.
State 1989 1990 1989 Average 1990 Average Montana 55.1 55.3 54.9 54.4 Kansas 55.5 55.2 Connecticut 49.1 49.5 Idaho 55.4 54.8 Rhode Island 59.5 57.3 Alaska 52.4 49.1 54.4 53.3 Vermont 60.2 57.4 Illinois 55.6 55.5 Wyoming 58.0 56.8 Missouri 52.0 54.2 Oregon 52.4 51.8 Michigan 55.9 53.7 Virginia 49.3 48.7 Pennsylvania 57.3 56.8 57.0 57.3 Utah 62.2 63.2 Tennessee 49.9 49.9 Alabama 51.8 51.1 New Hampshire 57.7 59.4 Nebraska 62.9 63.6 Table 5. Percentage of Divorces Involving Children, 1989/1990 Source: Monthly Vital Statistics Report, Vol. 43, No. 9 (March 22, 1995), National Center for Health Statistics.
Put simply, when divorce becomes a less attractive alternative to marriage, we should expect less divorce. As can be seen from the findings, this appears to be happening in states with higher levels of joint custody. If sole custody reduces incentives to continue marriage, then we should also expect states that favor sole custody to have more divorces involving children. As can be seen from Table 5, the low joint/high sole custody states also had more divorces involving children, although the difference is not statistically significant.
Summary and Conclusions
The evidence reported in this paper indicates that widespread acceptance of joint physical custody will not increase the divorce rate, and may in fact reduce divorce. States whose family law policies - either by statute or through judicial practice - encourage joint custody have shown a much greater decline in their divorce rates than those that favor sole custody.
Both social and economic factors may explain the differences between divorce rates. Sole custody allows one spouse to relocate easily and to hurt the other by taking away the children. Potentially higher child support payments with sole custody may provide an economic motive for divorce as well. With joint physical custody, both social and economic motives for divorce are reduced, so parents considering divorce may simply decide it is easier to remain married. States whose policies result in more joint custody and less sole custody should thus see a reduction in divorce rates. The findings reported in this paper indicate that this is in fact happening.