THE COST SHARES MODEL OF CHILD SUPPORT GUIDELINES
Laura W. Morgan
In May 1999, we posted The Use of Economic Data in Child Support Guidelines, which reviewed current economic studies on the costs of raising children. In the intervening 5 years, the "Cost Shares" model, devised by R. Mark Rogers, has received support among non-custodial parents. This month's article is a short review of this model.
As the May 1999 article made apparent, economic studies on the costs of raising children have focused on intact two parent households or one-parent households. Child support in divorce and paternity cases, however, presume two households. This is an important deficit in the economic studies. More importantly, however, economic studies do not make the normative judgment as to how the burden of supporting a child is shared between two parents in two households. That normative judgment is contained within the child support guidelines model a state uses: Income Shares, Percentage of Income, Melson Formula, or Hybrid Formula. Any one particular economic study can be adopted to any other particular guidelines model. One is not tied to the other. Thus, the normative standards employed in child support guidelines models are equally or more important than the estimates of the costs of raising the children.
The Cost Shares model's chief criticism of these other models is that child support in these models is based upon the above described economic data that figure the expenditures on children in intact two parent families. The Cost Shares approach, by contrast, starts with the assumption that child support should be based upon on the actual out of pocket expenses incurred by the custodial parent.
When a family is living together, the family can economize by the sharing or jointly consuming goods and to a lesser extent through bulk purchases. When the parents divorce and two households are formed, the amount of sharing falls. The end result is that cost of maintaining the two households will be greater than one. Therefore, unless total income between the two households increases, the standard of living of at least one the households must fall.
This fact is reflected in the poverty thresholds employed by the Cost Shares model to adjust for differences in family size and composition. Using the poverty threshold employed by the Census Bureau in 2002, for a single individual, the poverty line is $9,359; for a single parent with one child, it is $12,400. Thus, the total cost of maintaining the two households with one child is $21,759 ($9,359 + $12,400), compared to $14,480 for an intact family of three with one child. These figures imply that unless the total income of the two parents rises by 50.3% (($21,759 - $14,480) / $14,480), regardless of how income is divided between the two households, at least one household will suffer a decline in its standard of living after divorce. Stated simply in economic terms, there are going to be losers from divorce. The normative question that guidelines seeks to answer is how best to share this burden.
The Cost Shares model, however, does not focus on the loss and how to share it. Indeed, it seems to ignore it. Instead, the Cost Shares model focuses solely upon the standards of living of the custodial and non custodial households after divorce relative to each other. The Cost Shares model thus discards the notion that children should be supported at the same level before divorce as after divorce.
The Cost Shares model, by comparing one household to another, uses the idea of equivalence scales: how much more income do you need in situation A to make it comparable to situation B. (See discussion in May 1999 article re Engel and Rothbart estimators and the use of equivalence scales.) The Cost Shares model compares the standard of living of the noncustodial parent (one adult) to the custodial parent (one adult and one child). To accomplish this comparison, the Cost Shares model constructs "poverty ratios": the income (net after tax after support) of the unit divided by the poverty line for the unit. If Incp denotes the income of the noncustodial parent, and Icp denotes the income of the custodial parent, then if the two households have an equal standard of living when the custodial parent has 32.5% more income than the noncustodial parent:
The figure 1.325 is the "equivalence scale," which expresses how much more a single parent with a child needs in income to be as well off as a single parent.
There is a relationship between equivalence scales and estimates of the proportion of consumption devoted to children, which the Cost Shares model ignores. Returning to the poverty line example above, if a custodial parent needs $12,400 to be at poverty standard of living, while a noncustodial parent needs $9,359, then the marginal cost of the child is $3,041 at the poverty line. Thus, 24.5% (= $3,041 / $12,400) of the poverty line consumption for a custodial parent is attributed to the child. The relationship between the equivalence scale (S) and the proportion of total consumption devoted to the child (C) can then be expressed:12,400 - 9,359 9,359 1 ES-1
CS = ____________ = 1 - ______ = _____ = ______
12,400 12,400 ES ES
There is a one to one relationship between the equivalence scales used in the analysis and the "cost of children."
From the above examples, 24.5% of net income is being spent on the child in single parent families, while 16.8% is being spent in intact families. We should not conclude, however, that because the percentage for intact families is lower than for custodial parents, then spending in intact families is lower than in single parent families. Because of the income differences in the two households, intact families spend a lower percentage of their net income on the child, but they have substantially more net income. The result is that spending the child will fall if no support is given.
This leads to two key conclusions. First, there is a one to one relationship between estimates of the cost of children and the equivalence scales used to compare households of different composition. One cannot be assumed without the other. The Cost Shares model, however, does exactly that. Thus, if the states make assumptions about the cost of children in the Basic Obligation tables, it does not make sense to use a set of equivalence scales to compare standards of living that are at variance with assumptions that the state has made. Alternatively if one assumes a set of scales, then one should use those cost of children implicit in those scales to compare different models.
Second, in the absence of support, the level of spending on children in single parent families will be less than the level of spending that would have occurred in the intact family. The only exception to this claim is when the custodial parent's share of combined gross income significantly exceeds 50%.
The Cost Shares model has produced significantly lower child support awards at every level of income for the simple reason that guidelines assume one set of implicit equivalence scales or cost sharing in intact and divorced families that differ from the scales used to normalize the net incomes of the two households. This is because the Cost Shares model ignores that the custodial parent is expected to contribute a specified amount to the child(ren) and the government is providing the tax subsidy that should be added on to what the parents provide.If the parents are expected together to provide the amount of support that they would have provided if they had remained married, then the custodial parent has either an explicit obligation to provide sufficient funds to meet this target level of spending. In the Income Shares model, this is an explicit requirement while in the Percentage of Income model is implicit. In the Cost Shares model, however, the normative standard differs from these two approaches. It adopts the standard (target) of spending on the child in the post divorce situation. As noted above, this is a lower normative standard. Thus, if the custodial parent seeks to provide for his/her children at the same level as during the marriage, the Cost Shares model leaves the custodial parent at best equally well off as the non-custodial parent (equal income case), but in most cases worse off.
Finally, the equivalence scales implicit in the poverty lines are arguably too low, for they assume too much sharing; that, too, would lower the impact of both taxes and support on the relative well being of the two households. For example, instead of 1.325 scale, 1.50 is better estimate. Many mainstream economists as well as the National Academy of Science Panel on Poverty measurement came to this conclusion.
In sum, the Cost Shares model of child support guidelines makes many assumptions that are out of the mainstream of economics. Moreover, at every level of income where income is not equal, the Cost Shares model produces child support awards that are lower than any other state's currently existing child support guideline. The Cost Shares model has abandoned a basic tenet of child support: the child should remain at the same standard of living before and after the divorce. For these reasons, states should be wary when presented with a Cost Shares model for consideration.