The Child Support Guideline Problem*
May 6, 1998
ABSTRACTThis paper discusses current child support formulae and policy in the context of the history of their development. Major flaws in current formulae and their use are apparent. Moreover, the flaws are so serious that existing child support guidelines do not meet federal requirements upon which eligibility for funding of state programs is based. An approach for developing child support guidelines is provided that will meet all federal requirements and should lead to dramatic improvement in the design and use of child support formulae.
Efforts to produce mathematical formulae to assist in child support decisions have gone on for decades in the United States. Historically, mathematics as an integrated part of child support law has been more prevalent in other countries. This is especially true for child support law in socialist countries where award decisions are made in relation to an array of welfare state benefits. In the United States, their formal statutory use as a presumptive calculator of awards prior to 1990 was limited primarily to welfare cases. As in other countries, states and the federal government wanted a highly consistent formalized way of calculating an appropriate amount to be recouped from non-custodial parents for assistance provided to custodial parents and children by government programs.
In a 1981 case that sparked national interest, the Oregon Supreme Court went to lengths to explain established child support doctrine(1) (Smith, see footnote). At issue was whether the child support formula used in Oregon welfare cases should be used in non-welfare cases. The Oregon Supreme Court found that the welfare formula did not correspond to child support law written for non-welfare cases and it was therefore inappropriate to apply it outside the limits of the welfare system.
Formulae used in the welfare program are designed to maximize reimbursement of public money given as entitlements. Simply maximizing the amount of child support awarded is not an appropriate goal for all situations. In non-welfare cases, when the parents are able to support themselves and their children on their own, the situation was not as simple as getting as much as you can. Awards were based on the "actual and necessary needs of children" and the ability of each parent to pay were considered among other things. The Court went to some effort in Smith to explain what was meant by children's "needs".
Although the Oregon Supreme Court did not require the use of an alternative mathematical formula in non-welfare cases, they cited work on a child support model presented by Maurice R. Franks as coming closer to established non-welfare child support law(2). Franks' child support model, and those like it, have often been called Cost Sharing models because legal experts often referred to parental spending on children using the term "cost"(3). Those who read Franks' paper will be impressed with the fullness of his legal citations in support of his model. This is not to say however, that child support decision modeling had made sufficient progress to substantially replace judicial discretion in the application of child support law.
Franks' accepted the traditional approach of determining costs for raising children from family spending history, and cross credited expenses in both households to account for shared custody and visitation / parenting time. Franks' model is significantly different from the Income-Shares model developed by Robert Williams that is so widely used today. In contrast to Franks' model, the Incomes Share model is not based on real household expenses on children, and arbitrarily under-accounts for shared parenting time.
One similarity shared by both approaches is the notion that the parent's share of the "child support obligation" is taken to be in proportion to their share of the parents' combined income. This produces a major failing in both models. Neither considers either parent's actual "ability to pay", which is addressed below. Neither provides a solution to the problem of adjusting the standard of living in the custodial parent household. Franks' doesn't include it and Williams simply raises the numeric tables arbitrarily producing results so high that they often overshoot "child support" to include alimony(4).
Prior to the federal requirement for development of state-wide child support guidelines in the Child Support Enforcement Amendments of 1984, "guidelines" had already come into use in state courts. Many child support guidelines had been developed by judges and local bar associations. Some were simply used informally by judges and attorneys while others were used as county-wide guidelines. Robert Williams' 1987 report provides a small sample of such guidelines without probing into the details of their correspondence with established law or the subtleties of their use(5).
It is important to note that judicious use of a simple formula and / or numeric table as a "guideline" does not contradict the decision of the Oregon Supreme Court cited above. There is a fundamental and significant difference between using such tools as a guideline to assist in decision making and the presumptive use of a child support formula.
One argument given by advocates of presumptive child support formulae was that award levels appeared to be non-uniform. Two divorced fathers could be living on the same street and going to the same job with the same pay each day, and pay different amounts of child support. The problem with this complaint is that it suggests that the two fathers should be paying the same amount regardless of other mitigating factors. In other words, it presumes a simple percent-of-income formula rather than applying established child support law. But such an approach does not relate the child support award amount to the needs of children and the relative ability to pay. Its relationship to family circumstances is random, and therefore it cannot meet the federal requirement that the application of a guideline result in a "just and appropriate award in each case". It's also reasonably obvious that without a rational relationship between an award and family circumstances the method would fail any honest constitutional test for "due process" and "equal protection".
After the federal reforms took effect, approximately one third of the states are actually using a simple Percentage-of-Income formula. The method is so obviously over-simplified and unrelated to sincere policy modeling efforts however that it is often left out of general discussion. It has its origin in old Soviet Russian law and there is no reasonable argument for using it in the United States. A slightly reformed version still exists as Article 81 of The Russian Family Code, adopted in 1995. Its use was promoted in the United States by Irwin Garfinkel as part of a suite of Soviet Russian policy that has become known to us as "The Wisconsin Model". The Wisconsin Model then became a center-piece for the national child support and welfare reform movement.
Another item lobbied for by women's groups in the 1980's that ultimately affected child support guideline development was a push by women's groups to increase alimony. Failing sufficient political support for increased alimony, attention turned to increasing the amount paid under the rubric of "child support". According to Williams' 1987 report, the intent of his recommendations was to increase "child support" awards dramatically above what they had been according to established state child support laws. The detailed discussion and recommendations by women's advocates on increasing alimony is the only logical forerunner to Williams' child support recommendations that has to date been found.
In discussion on child support reforms, advocates of increasing support have typically pointed to differences in income between men and women as the basis for the need to increase child support. This argument is direct in its appeal to include alimony or spousal support in child support awards.
The common claim of inadequacy of child support awards (referred to as the "adequacy gap" in child support awards) typically centered on the fact that the average of child support awards was not sufficient to support both the custodial parent and child(ren), or at the very least that it wasn't sufficient to cover 100% of child(ren)'s needs - the latter ignoring the obligations of custodial parents to contribute and substantial direct contributions by fathers during visitation for example.The general belief among legal experts is that child support should never be used to equalize income between the two households. Such a policy is understood to result intentionally in the inclusion of spousal support in child support awards. Spousal support should not be awarded as part of a child support award because spousal support can be awarded separately when appropriate and more fundamentally, spousal support is not child support.
in Child Support Approaches / Improvements
One of the problems with Percent-of-Income, Income-Shares, and early Cost Sharing models is their simple use of income for assigning, or dividing the support obligation into a proportionate share for each parent. None of these approaches addresses the parents' ability to pay support for their children or the consequences to the children or either parent when meeting the obligation of support.
Consider a custodial mother who makes $8,000 per year take home pay and a non-custodial father who takes home $16,000 per year. For the purpose of this illustration, please assume that exactly $8,000 per year is required for basic support for one adult living alone. An Income-Shares type formula would distribute the obligation in proportion to each parent's income. The custodial mother in this example, who earns just enough to care for herself, would have an obligation to provide one third of the additional money needed for support of their children. If that amount is say, $3,000 per year, then she is left with only $7,000 per year for her own support, which is not adequate. This would likely lead to inadequate support of their children.
A simple but much more rational formula for "ability to pay" was proposed by Melson (Judge who developed the late-80s Delaware model), Cassetty (well known contributor to discussions on child support calculations) and others. Their basic model of "ability to pay" is net income minus an adult "self-support reserve". The self-support reserve is deducted immediately and becomes untouchable for both parents. The result in the case above, is that the mother's total income would be hers - the amount she needs to support herself - plus the additional amount needed in her household for care of the children. That is exactly the result that is needed.
The effect of income disparity, when one parent is able to pay significantly more than the other, is most obvious when one of the incomes is near or below subsistence level. A better specification of "ability to pay" would pass all reasonable logical tests when applied to all cases. It has the effect of shifting income as needed in a non-arbitrary way. It would thus allow replacement of the arbitrarily high numeric tables currently in use with tables that are realistically related to the circumstances of children and parents. The best vision is that of allowing consideration of a wider range of mitigating factors in determining ability to pay. Doing so would provide one part of the fine-tuning needed to meet the federal requirement for a "just and appropriate award in each case".
Franks' model is more advanced than Williams' in dealing with shared parenting and visitation. Franks presented a method known as "cross crediting". There has since been discussion on whether the cross crediting formula should be applied strictly according to the amount of time children spend in each household when one or both parents' income is low. This question is resolvable and the cross crediting concept provides a very solid theoretical basis for dealing with the sharing of direct expenditures by parents.
Williams' simply filters out all credits for visitation and shared parenting arrangements, and assume zero time-share. The tenacity with which he has fought to eliminate visitation credits is one of the reasons that he, rather than just his work, has often been characterized as extremely biased against non-custodial parents. Even when state representatives and state courts protest, opting for at least some token reduction in support as visitation credit, Mr. Williams has fought to minimize it. With the elimination of proper visitation credits, the income of many non-custodial parents can be reduced beyond the point where they are able to afford normal visitation. The objective evidence is telling us that much of the reason for lack of regular visitation is because the non-custodial parent cannot afford to support the children during visitation. This problem increases dramatically without proper credit for visitation in the formulation of child support awards.
Under the present child support award system, those who can still afford visitation are usually not convinced that they should be forced to pay many expenses twice; once directly during visits and again in the form of a child support payment calculated as if the children never visit. After the introduction of Williams' recommendations no state is cross-crediting.
As an example of the effect of improper cost crediting and sharing (that occurs in Oklahoma and elsewhere because of existing policy and law); if expenditure on a child is $3000 per year and the parents have equal net income and a 50/50 joint parenting arrangement (we assume in this example equal direct financial obligations), the non-primary custodial parent would pay $1500 in child support to the primary primary parent. The paying parent would still have $1500 in costs for the child in their home for a total expenditure of $3000 for the child. In contrast the primary custodial parent would spend $1500 but that full amount would be reimbursed by $1500 in child support leaving a total financial contribution by the primary parent of nothing.A long-standing problem with all models has been the lack of a theoretical solution to the question of adjusting the standard of living in the custodial parent household. This issue arises because traditional child support law intended to provide some reasonable protection for children against the decrease in living standard that often accompanies divorce when the mother does not remarry1. Franks doesn't deal with the question. Cassety made a very direct and pointed issue of the problem but did not solve it. In developing the late-80s version of the Delaware formula, Judge Melson went to a great deal of effort to create a basis for a reasonable judgment. He decided to add 5% of remaining income after deducting the self-support reserve and a basic amount of child support. The adjustment was later changed by the state of Delaware to more closely match the arbitrary increases in awards recommended by Williams. A theoretical solution to the standard of living adjustment problem was not presented until 1994(6).
Fault in Williams' Model
Child support guidelines in use in the U.S. are based primarily on the opinions and assumptions made by two people, Robert Williams and Irwin Garfinkel. Approximately two-thirds of all states use the Income Shares approach that has been questioned continuously since the publication of Robert Williams' report in 1987.
According to the Child Support Enforcement Amendments of 1984, Robert Williams technical report7 was supposed to "provide technical assistance to the states" in developing their child support guidelines. It was not under that legislation, and still is not today, an acceptable role for the federal government to decide each state's policy for making a child support award. Yet, each of us who have carefully studied Robert Williams' report recognizes that something is amiss. Robert Williams' recommendations are not based on the established state laws of the time. His methods are not flexible enough to adapt to policy choices states are entitled to make. Underlying his technical recommendations are Robert Williams' policy choices. (Something he admitted in a deposition in a federal case.) And those choices are not clearly specified so as to facilitate open discussion and debate.
The explicit argument given in favor of Williams' model rests on rather crude statistical methods which Robert Williams typically refers to as "economic studies". There is no appropriate economic data at the heart of these "studies". The data that Williams, Betson, and a few others have related to their "studies" is so off target that there is very little reason to refer to their studies as being "statistical".
The data used in these "estimates" is the national data on family expenditure that comes from the Consumer Expenditure Survey. Nearly all of the data is on what families spend as opposed to what is spent on individual family members. In other words, it does not provide a statistical view of what is spent on children and adults in households but what is spent in total on different items of expenditure. One can get a reasonable view of what households spend on housing and transportation for example. But no amount of witchcraft can transform that cost into how much is attributable to children, or differences due to cost of living variation amongst states.
The data used has very little effect on the numbers produced in the so-called "economic studies". The numbers are primarily the result of the arbitrary choices the modelers make in selecting an estimating method. The modelers choose the portion of family income "distributed" to adults and to children. That information does not exist in national data on family spending and there is no way to divine it from that source. Here is what two competent researchers said about such "studies".
. . . the presumption that underlies the focus of much of the empirical research and policy debate on income distribution [within households] seems born of ignorance and is supported by neither theory nor fact.(7)
In all fairness, some of those studies were carried out with some degree of competence and good intent. Although failing to provide much solid useful information, the most honest directly pointed out the weaknesses of the methods and uncertainty of results and provided the most open exposé of the arbitrary choices made. Even so, not a single of the so-called "cost of raising children studies" was done for the purpose of assisting in the development of child support law. One researcher, working directly on the problem of developing child support guidelines said the following about such studies.
… it is possible that achieving confidence in the data base through use of a simple methodology which explicitly relies on "user opinion" will be more effective in moving practices more uniformly toward a fair standard than does reliance on opaque and highly derivative expert interpretations of existing but fundamentally off-target primary economic data.(8)
It should be especially noted here that David Betson's study(9), although done to fulfill requirements in the Family Support Act, is no exception. No new techniques were developed for divining information more relevant to a child support award decision. Betson's task was simply to update old information using information from a newer Consumer Expenditure Survey collected between 1980-1986. His report contains a wide range of "cost of raising children estimates" and no solid scientific basis for preferring one over another. Betson's display of such a wide range of possible estimates from the same data should have been seen as yet another of the several very loud and obvious signals indicating that something is amiss.
This point is so important that it's worth a simple example to be sure non-technical readers understand what is meant by all this. Here's a simple example. Let's say that on average families with two adults and one child and an after tax disposable income of $35,000 per year actually spend $30,000 on "family spending" that's used in a "cost of raising children estimate". From that information you are asked to estimate how much of that $30,000 is spent on the child. One estimate starts by making the arbitrary choice that one third of all family spending is for that child. Dividing $30,000 by 3 gives $10,000.
Following the logic of Betson / Williams we would say that the estimate of $10,000 in spending on children came from the data - $30,000. But the data had little to do with the outcome. If we had wanted $7,500 as the result instead of $10,000, we simply would have selected the estimating method "divide by 4" instead of dividing by 3. The answer is manipulated by the method that is used for estimating what is spent on children. If we apply the same "divide by 3 method" to total family expenditure of $60,000 per year we get $20,000 per year. It's as simple as that. By manipulating the method the analyst can get any answers they choose to get.
Nothing in the data indicates which is the right or even the best choice. There is, as a matter of scientific fact, no method sophisticated enough to derive information from data that doesn't contain the information you're looking for. Whether in obscure or obvious form the information sought must exist in order for it to be found. The holy grail, so to speak, how much families spend on children is not information that is found in the data from the Consumer Expenditure Survey and it cannot be derived from that source. It definitely can't and even if it could be derived from some source, that still wouldn't answer the question that needs to be answered - how much should each child support award be?
Williams has defended his use of child cost estimates primarily by pointing out that someone else did them. From the beginning, he's based his model on the work of Thomas Espenshade. ("Updated" by Betson in reference to a more recent Consumer Expenditure Survey.) Espenshade's work did not emerge within the context of mainstream scientific process and was only a recent addition to commentary in the form of a book published by the Urban Institute Press(10). Espenshade's estimates of the cost of raising children had not and still have not been through any validation process or even substantial argumentation in the normal course of formal scientific discussions.
Moreover, Espenshade's "cost" estimates were in no way related to the question of making child support awards. Political use has been made of the fact that Espenshade's estimates are not related to the question. For several years, many analysts have been pointing out that intact family spending has no relationship to post-divorce circumstances. Use of such data can only lead to awards which are randomly related to post-divorce family circumstances and children's needs. Espenshade had not made estimates of single parent spending because his study did not concern the question of making a child support award.
Ultimately, Hewitt's suggestion is a very reasonable one. Adjustments to the numeric data are best provided by "user opinion" from those who can directly observe the results of the use of the guidelines and judge them according to their appropriateness in dealing with individual situations in the light of the principles of child support law. This is exactly the way many child support guidelines were evolving before the federal government stepped in and began manipulating the process. Williams' model, on the other hand, is estimated to produce awards on average 250-300% higher than those that were being awarded according to established state laws. That's obviously a case of someone being very generous with other people's money.
One report by the OCSE actually does point to a problem in review of Williams' based guidelines. The following specific comments can be found in the executive summary of volume II of the report(11).
Surprisingly, few States reviewed their core guideline model or methodology. Rather, guideline reviews focused on issues relating to income, adjustments to income, adjustments to the guideline amount, and deviations from the guideline amount.
It is surprising that the OCSE reporters find this result surprising. Let's take a step by step approach. The policy discussion has been focused on junk "economic studies" which in themselves mean nothing. Even if proper data existed so that a worthwhile statistical study could be carried out, there would still be issues to confront on the structure and purpose of those studies.
Even if we did have an appropriate data set there would still be a fundamental problem. Economic studies, even good ones, can at the very best provide only one element in putting together a final guideline and using it properly. Even if we had the perfect data set, there is no statistical formula for extracting child support policy from spending data. That is, the state must first decide what its child support policy will be. Yet, the distraction caused by the focus on so-called "economic studies" has prevented reasonable and much needed discussion about child support policy.Robert Williams' involvement in child support issues coincides with the formation of his company, Policy Studies Inc. in 1984. We find no record of his involvement in family questions, no history of academic achievement in the field or even evidence that he's qualified to deal with complicated policy / design issues. He came from nowhere in the mid-1980s as the Office of Child Support Enforcement's choice to provide technical assistance to the states in developing child support guidelines and was able to provide nothing except extreme policy views. Without having any legal authority, or a logical or scientific basis for his recommendations, he has to a very great extent dictated child support policy in all states ever since. Most disturbing of all is that his business operations include a collection company that takes a percentage of the amount of child support paid. Mr. Williams therefore has a direct financial interest in increasing award amounts
Rather than forcing compliance with a poorly designed formula by mandating acceptance of its flaws, the honest course is to expose the mechanics constantly to every legal test and by that approach force improvement of a formula and its use.
There is no example of a state statute that defines child support independently of the calculations used to determine a standard award. (Confirmed by OCSE in 1994.) This is a very serious situation. There is absolutely no way that states can be in compliance with the requirements of federal law without having a clear statutory definition of child support(12).
The law requires that awards determined by the application of child support guidelines be rebuttable. The federal law specifies: "A written finding or specific finding on the record that the application of the guidelines would be unjust or inappropriate in a particular case as determined under criteria established by the State, shall be sufficient to rebut the presumption in that case." It further specifies that guidelines "shall be reviewed at least once every 4 years to ensure that their application results in the determination of appropriate child support award amounts".
It is impossible to meet either of these requirements without having a legally established definition. A specific, meaningful, statutory definition is the only mechanism that can provide litigants in every case with criteria to determine whether the presumptive award is just and appropriate. It is also impossible to determine whether the application of guidelines generally result in appropriate award amounts without having a definition as a basis for the determination.
The arbitrary nature of current child support policy has its roots in the tendency of federally designated analysts to look for answers in ways in which answers cannot be found. Above, Williams' approach rests on data that does not contain the information he seeks. We've noticed a similar problem in the more general discussion on what a "just and appropriate" child support award might be. We've seen this issue shift to the general question of "fairness" and then "fairness" defined as an unresolvable conflict. Specifically, we've repeatedly heard "fairness" discussed as the condition that everyone gets everything they want. Given that the fundamental purpose of the court in many child support cases is to settle disagreement, it's quite clear that this view will not yield a general solution.
We have not heard of any credible argument, nor do we ever expect to, that fairness does not first require defining what is being done. No judgment can be made about what is fair unless there is a specific understanding of what a child support award is supposed to be. We can see in state laws that existed prior to the federal government's encouragement to use presumptive formulae that this principle of legal construction was well understood. It is also apparent that the law must also make a relational statement about the obligations of the parents and provide the courts with the proper authority to consider all relevant factors before making a final judgment.
In a 1993 conference paper, a step-by-step list was provided for development of a well integrated child support policy. "Well integrated" policy begins with a child support policy put in place by a state legislature. Child support guidelines are then developed to correspond to the state's legally established policy. The overall process is one in which guideline engineering is integrated with the well established traditional process of legal construction(13).
A reasonably broad survey of state child support statutes was made. Of necessity, the laws surveyed where those which were in place prior to the date the federal mandate for use of child support formulae took effect. What was needed was the essentials of well established definitions, relevant doctrine, and an understanding of the legally established considerations in child support award decision making. The survey included review of some important case law.
The model child support statute given below is based directly on the Oregon child support statutes and contains much of the original language. It is typical of many state child support statutes that were in place prior to 1990. The work was facilitated by the Smith case1. This brought a great deal of detailed understanding and clarity to the established law that would not have existed simply from reading the statute and reviewing a few less comprehensive decisions. The question in Smith was in fact the appropriate use of child support formulae, making it the perfect case study, especially since the judges chose the occasion to provide their most extensive discussion on child support law. The inclusion of the presumptive use of a child support guideline (rebuttable) explicitly brings the statute into perfect compliance with federal requirements.
In order to make the transformation from traditional legal principles to the process of formulating a mathematical model, a set of concrete statements was extracted and organized in a way that is convenient for a logistician / mathematician. Writing as logistician, the basic elements of any valid child support law / formula are described below as the "fundamental laws of child support".
Fundamental laws of child support
1. Child support is for the care and maintenance of children.
2. Both parents have an equal duty to support their children.
3 All relevant circumstantial information may affect the amount of the award.
These "fundamental laws" are typically found in traditional child support statutes. The "first law" seems almost trivial, but it is essential to build upon a basic statement of purpose. Without such a basic defining statement, all else is arbitrary. The "second law" was originally found in a separate statute(14). Logically, once one decides what child support is, one must also decide who is responsible for paying it. We've not found any reason to doubt the wisdom of those commentators who insist that "equal duty" is Constitutionally mandated(15).
The necessity of the "third law" can easily be explained from consideration of the second. Although both parents have an "equal duty" to support their children, it has never been held that each parent must pay an equal amount toward support. How much each parent should contribute is determined by the careful consideration of the circumstances of each parent. This third law determines each parent's ability to pay, what the children have already or need to be supplied in both households for their physical support, amongst other things. There is no way to produce results conforming to the "second law" without application of the "third law". In the end, the best decision can only come from reasonable consideration of the circumstances of each parent and the needs of their children(16).
It is not the purpose of this article to document a full detailed expansion of the modern mathematics of child support. In the traditional process, interpretation and detailed expansion of the rule of law was, of practical necessity, left to the courts. By providing a statute that resembles traditional state law, the legislature offers the state courts the benefit of the decades of legal development that preceded the Family Support Act of 1988. As shown by the model statute below, it is a rather simple matter to modify the statute to comply with current federal requirements.
But the discussion above does carry with it the intent to argue that this is the only proper way to construct child support law. The three "fundamental laws" are essential to any valid child support statute and to any valid child support formula as well. It is our opinion that Constitutionally acceptable child support law cannot be constructed without the central inclusion of the three "fundamental laws" given above. State child support policy must consist of these three "fundamental laws".
Proper implementation of federal law, requiring procedures that result in a "just and appropriate award in each case" as well as requiring periodic review to assure that the use of child support guidelines results in a "just and appropriate award in each case", should lead states to improve their guidelines. Each case provides an opportunity to learn about the weaknesses in the design of a child support guideline. At the very least, each required periodic review offers another opportunity to improve the design of the guidelines in light of what has been learned from experience.If this procedure is followed, it is reasonable to expect that the need for deviation from guidelines will be reduced over time so long as the need for deviation is reviewed and that information is used to improve guideline design. Currently the incidence of deviations is reduced simply by ignoring flaws in the guidelines and inappropriate results. Instead, it is recommended that case experience be used to direct pressure toward improving the quality of child support guidelines and thus reduce the need for deviation by improving the quality of the results they produce.
Above, it has been pointed out that there is at present no national data base which provides sufficient information on parental expenditures on children. Yet, we know that expected expenditure on children is one of the key questions in making an award. Traditionally, the courts would attempt to determine what had historically spent and in effect attempt to predict spending on children in the future. This process of course, led courts and bar association groups to develop tables from which one could quickly and consistently determine a "reasonable amount" in order to achieve better results more efficiently in dealing with this specific part of the child support award question.
In the section above which explains one of the major faults of the popular Williams' approach, William Hewitt, a researcher in Washington State, is quoted as pointing out that "user opinion" is likely to provide the best improvements to the numeric table. It seems apparent that those who are experienced in the direct application of guidelines can best contribute to their improvement. Nonetheless, there are important conclusions that basic research can provide.
Spending on children in split households has a random relationship to the combined income of the parents. The income of both parents can be appropriately considered in the award decision only if that consideration is consistent with the fact that the parents do not live together and therefore do not use their income jointly. The only approach that provides an appropriate outcome begins with consideration of the financial circumstances in the custodial parent home. The full effect of non-custodial income can properly be included in the detailed mathematical model, but not by a numeric table with values related to combined parental income.
Regardless of what a freshman economics textbook might say, "ability to pay" is not equal to income. Traditional statutes and case law provided that one of the important determining factors in the award of child support is the parents' relative ability to pay. Courts also concluded, on basic legal grounds, that so much could not be taken from the person ordered to pay support that they are unable to support themselves. Mathematical study has shown that there is no consistency of logic unless this rule is also applied to the income of the custodial parent. It is also apparent from this study that children are best protected against inadequate award levels when parental income is reduced by adult needs and the remainder is taken as "ability to pay". This view of "ability to pay" has been investigated by others as well, and was applied in the Melson formula used in several States.Unusual case circumstances (those which deviate from the circumstances presumed in developing the guideline) cannot be adequately considered unless the numeric table is categorically divided (food, clothing, shelter, transportation, entertainment, etc.) The State of Vermont tried categorical division with a presumptive child support formula. The experiment was tried early this decade when support for forcing an overly simple statistical consistency in awards was particularly high. The State quickly abandoned this feature when it produced a much higher number of deviations. This experience illustrated the poor quality of the design of their formula, which happened to have been a version of Williams' Income-Shares model. As stated above, the acceptable approach is to allow such problems to force improvement in the design of the guidelines.
It is apparent from the OCSE report mentioned above and from our own discussions with people around the country that most states have not carried out any meaningful structured review process. Most states are simply repeating the political process they began with. Supplementary to that, Robert Willliams has been making appearances to reassert his personal support for his own policy preferences.
One thing that would improve the review process tremendously would be to actually have a child support policy. In Fitzgerald, cited above, the Court characterizes the litigants view in trying to exercise the right of rebuttal to the presumption that the guideline amount is correct. Without an explicit and clear conceptual basis for the award a litigant attempting to rebut the presumptive amount on the basis that it is unjust or inappropriate must do so without knowing what just and appropriate means. (Obviously impossible, and thus unconstitutional.)
The same situation obviously exists in regard to state review of child support guidelines. Federal law requires reviews be conducted to assure that application of a guideline results in a just and appropriate award in each case. Without a credible child support statute, reviewers are in the same position as litigants (and judges). They have no basis for judgment. With a proper statute, including proper authorization for the courts to apply it (see model statute) the courts themselves will review the guidelines in the best and most comprehensive way - the way the Constitution intends.Child support committee members and others interested in child support need to clear themselves of the distraction of fundamentally invalid "economic studies" and discuss the definitions and logic involved in making an award. State legislatures need to do their job of deciding what the state's policy will be.
Child Support Statute
Model Child Support Statute
Based on OREGON REVISED STATUTE, ORS 107.105, 1989
Whenever the court grants a decree of marital annulment, dissolution or separation, it has power further to decree as follows;
For the recovery from the party not
care and custody of such children, or from either party or both parties
if joint custody is decreed, such amount of money, in gross or in
or both, as constitutes just and proper contribution toward the support
and welfare of such children. The court may at any time require an
from the custodial parent with reference to the use of the money
as child support. The court is not required to order support for any
child who has become self-supporting, emancipated or married, or who
ceased to attend school after becoming 18 years of age. In determining
the amount of the child support, the court shall consider the economic
needs of the children and determine payment by the parents in
to their respective ability to pay on the basis that each parent has an
equal duty to provide financial support for their children. There shall
be in any proceeding for determination of the child support award, a
that the [child support schedule] provides the proper award. Each
award is subject to review at the request of either party. The court
determine whether the presumptive award is just and appropriate under
terms of this statute and others in force. In all cases, the court
provide a written statement listing the relevant considerations and
facts related to its' decision. In making its' determination, the court
shall consider, but not limit itself to, the following factors:
(A) The financial resources of both parents;
(B) The ability of each parent to support themselves;
(C) The cost of day-care if the custodial parent works outside the home;
(D) The expenses attributable to the physical, emotional and educational needs of the child;
(E) The tax consequences to both parties resulting from spousal support awarded, if any, and the child support award, and determination of which parent will claim the child as a dependent;
(F) Expenses in the exercise of visitation;
(G) The existence of children of other relationships; and
(H) Expenses arising from other factors as the court may determine relevant in a particular case
International Influence and U.S. Welfare Reforms
Federal reform of the child support system has been the most significant part of welfare reform in the US over the past 15 years. The purpose of the reforms was to 1.) federalize the child support system, 2.) extend the welfare system's formulae and enforcement methods to non-welfare cases, and 3.) adapt to defined and as yet undefined international standards. The acceptance of reforms has been aided considerably by support from groups representing segments of the population which have profited directly from the initial increases in child support awards, primarily divorced, middle class mothers.
In 1973, The Hague Convention on Recognition and Enforcement established an international view of cooperation in the enforcement of child support orders. In 1974, apparently lacking any sense of coincidence, Senator Russell Long "perceived a connection" between "fathers who abandon their children" and a growth in AFDC spending. This led to the original federal child support and paternity legislation enacted in January 1975(17). Among other things, child support enforcement services were required for families receiving assistance under AFDC, FC, and Medicaid programs.(18)
The welfare community did not favor the legislation and only a few Senators spoke in favor of it. When it passed, it did so at least in part because it was tied to more popular social service amendments(19). When passing the legislation, President Ford contended that the provisions went "too far by injecting the Federal Government into domestic relations." He complained of "serious privacy and administrative issues," and promised to propose legislation to correct defects(20). During the Reagan years, our new addition to the federal bureaucracy, the Office of Child Support Enforcement, embarked on a national propaganda campaign. By the time the Child Support Enforcement Amendments were proposed in 1984, which began a dramatic expansion in the office's size, budget, and powers, most politicians were talking as if "deadbeat dads" were the nation's most serious problem(21).
At the Hague Conference on Private International Law in 1995(22), a U.S. delegate promised the international community that federal legislation would "provide for services at the federal level through a Central Authority to ensure an efficient, workable and uniformly implemented system in cooperation with the states and with the foreign countries which are willing to take part. In addition, the federal government is considering the possibility of the United States becoming a party to one or more of the existing conventions."
The delegate obviously went too far. The federal government has taken new steps to modify cooperative agreements among the states and between the states and foreign countries.(23) In general however, the situation is the same as it was before. The Constitution prevents the United States government from moving directly into territory reserved to the states and to the people by the ninth and tenth Amendments. The overall effect of federal tampering has been to make the system more bureaucratic, frustrating and even angering tens of millions of parents. In the end, it wasn't even important. There's simply not a significant difference between the law of a foreign country saying "all states within the United States"(24) instead of "the United States". Those feeling that it's too burdensome for foreign countries to have fifty addresses instead of one can think about a central office of communication rather than a "Central Authority". Additional government power was simply not needed to make cooperative efforts workable and efficient.
Reform efforts within the US have been driven by the same bias found in international forums, more so than just in the preference for bureaucratic, centrally controlled governmental systems. At a Hague conference on international enforcement of family / child support, one foreign delegate reported that the need for enforcement was clear due to the efforts of debtors to "do anything to avoid their responsibilities".(25) Promotion of reforms in the United States took that idea as its central theme, even though the United States has historically had one of if not the highest compliance with child support orders in the world.(26)
The Bureau of the Census in the US reported on child support payments in the spring of 1995.(27) According to that report, the so-called "deadbeat dads" are few and far between in the population with valid child support orders.(28) Comments on child support compliance often focus on the estimate that only about 66% of the child support that has been awarded is paid. This does not consider the fact that more than 14% of the amount under study had been recently awarded and was not yet due. Considering custodial parent reporting bias and adjusting for awards not yet due brings us closer in line with the information provided by Braver et al.(29) as well as information collected by commissioners in the states. Approximately 80% of the total amount of child support awarded in the U.S. has historically been paid each year. The compliance rate was not significantly effected by reforms.
While labor force participation by women has increased from 30 to 57% since 1950, participation by men has decreased from 82 to 74%.(30) Participation is still 23% higher for men than for women. Traditionally, child support has not often been awarded when the father is given custody. Increased participation by women in the work force as well as new child support laws are leading to change. The fact that custodial fathers received only about 44% of the amount awarded according to the Bureau of the Census study (compared to nearly 66% received by mothers) is largely, but not completely explained by the fact that a larger percentage of custodial fathers had very new awards and payments were not yet due.
This leaves us with about 20% of the amount fathers have been ordered to pay to custodial mothers unpaid each year. From that figure, we must deduct for fathers who have died, those who are unable to work due to incapacity and incarceration, and for orders that no longer require payment because the child has died or has become emancipated, the very serious problem of missing custodial parents and children,(31) and even for changes in custodial arrangements. As Braver et al. point out,(32) the remainder, as well as some lateness in payments (classified as partial payment) is primarily explained by un- and under- employment. These figures give us an understanding that the failure of the new child support system to reduce dependency on welfare was predictable.
Failure of federal child support reforms of the 1980s and 90s to reduce welfare dependency is easily explained. At the beginning of this paper, we mention the Oregon Supreme Court decision in Smith v. Smith (1981). The issue in this case was whether it was appropriate to apply child support formulae that were already being used in the welfare system to non-welfare cases. It should not be difficult to understand that extending use of these formulae to non-welfare cases would have no effect on welfare dependency.
Indiana State child support commissioner, Dr. David Garrod made a comparison between awards made prior to 1990 according to established child support law and application of a Williams type Income-Shares guideline used in his state.(33) He showed clearly that increases in award amounts went to custodial mothers with higher income. In fact, the higher the income of the custodial mother, the larger the increase. Mothers who then re-married and shared the income of a new spouse as well, tend to be much better off economically than they would have been if they had remained married to the father of their children.
One might expect that greater potential lies in the 46% of all custodial parents who have no support order. One can account for a significant number of welfare dependents as not receiving child support because paternity has not been established. Here again however, the reality is not as great as the expectation.
All the reasons for adjusting the expectation given above would also apply to this population. Some at least to a greater extent. Although we know of no study to date clearly focusing on the economic characteristics of this population, there is clear reason to believe that their average income is much lower and that they have a much greater problem with un- and under- employment. In addition, there are a large number of people who do not have child support orders simply because they are not interested in becoming involved in government programs and not in need of a court order for child support.
There should be some economic potential in establishing paternity and support orders in welfare cases. This is not a new revelation either. All states had paternity establishment programs prior to the reforms of the 80s and 90s. It may seem ironic that the federal government had to be coaxed during this period to re-focus its efforts from arbitrarily increasing child support awards in non-welfare cases to paternity establishment. But states were already experimenting with the idea of denying welfare benefits to women who did not cooperate in paternity establishment and had been reasonably successful in a program of locating so-called "absent fathers" in hospitals at mother's side when children were born. In addition, a more reliable DNA test was replacing a simple blood test, increasing accuracy of identification in contested cases.
We can go back to the Smith case to point out that a serious problem did exist. Why did Mrs. Smith feel there was reason to complain? There was a discontinuity in the treatment of child support cases at the boundary between welfare and non-welfare cases. The inequality of treatment on either side of the boundary was so significant that Mrs. Smith, apparently not a wealthy woman, proceeded through the judicial system to the State Supreme Court. Although we are confident that the court was correct in its determination that the welfare formula did not fit established non-welfare child support law, it doesn't alleviate the practical problems that arise from such starkly unequal treatment in similar circumstances.Mrs. Smith was caught between the moment the federal government adopted foreign methods for dealing with welfare cases and the moment the federal government manipulated states into extending those methods to non-welfare cases. In a political sense, she had been living in one country while on welfare and suddenly moved to the United States when her income became high enough to leave the welfare system. The two systems didn't match. What has happened since is that 10s of millions of Americans whose children are not dependent upon the welfare system have been transferred, sometimes kicking and screaming, into the non-American system. To date, there has been no government funded effort to adapt the new child support system to Constitutional requirements or those of the federal law mandating the change.
What should be regarded as the greatest mistake in the reform movement, as well as the greatest embarrassment to the United States is that the domestic political discussion has consisted almost exclusively of propaganda demonizing non-custodial parents. In the background, the American public has been aware of "The New World Order" in relation to the fall of the Soviet Union, and generally understand that new global trade agreements have been and are being forged. But not a hint of information has been fed to the general public on integration of or "cooperation" in an array of social programs or the impact of global integration on our domestic judicial system. Had the government made a greater effort at full disclosure, the American public would surely have responded with pressure to adapt newly proposed systems to Constitutional requirements.
The Milwaukee Journal Sentinel recently reported that an error of forty cents was made in withholding one man's final child support payments. The man believed he was finished with the child support enforcement agency and went on with his life. He wasn't told that he owed back support until one day he was informed that he faced contempt of court charges that could result in a 180 day jail sentence. By that time the agency said he owed $173.53 including interest and fees.(34)
The problems of the new system go well beyond inappropriate child support awards. The Los Angeles Times recently reported that in the process "designed to snag deadbeat dads and force them to pay up on their child support", one county had assigned paternity incorrectly to hundreds of men by default.(35) The bureaucratic idea of "due process" is that if the proper paperwork isn't submitted on time, the individual citizen they deem responsible for it faces the consequences. Those consequences included trouble with relationships and marriages and assignment of a child support obligation. The agency then refused to correct the problem and it has been necessary for innocent victims to go through the process of battling the bureaucracy in court.
It is time to talk about Americanizing our new child support decision system. Several states are actually using the Soviet Communist system promoted by Irwin Garfinkel, while all other states have been pushed in that direction by promotion of Williams' odd variation of the Income Shares model. Beyond the formulae for calculation of child support awards, there is the array of unrelated punishments such as the loss of drivers and professional licenses and the return of the United States to the ante-bellum standards of involuntary servitude and debtors prisons that had not been abandoned in the Soviet Union.
The fact is that the bureaucratic "efficiency" promoted by international integration isn't compatible with fundamental rights in the United States. "Efficiency" in this sense has nothing to do with doing the job correctly. It has had to do with a broadside attack on a large group of citizens, an over-controlling government, a federal government becoming too involved in the daily lives of individuals, and spending more money doing it. In the broad view, Americans who have become familiar with the system look at the billions of additional dollars spent on a system that provides no net benefit, a federally developed $3 billion dollar computer tracking system that doesn't improve payments but causes harm to innocent people, and they cannot equate the reforms with efficiency.
In 1995, the cost of operating the child support enforcement program was nearly 3 times the amount paid in reimbursement of welfare benefits.(36) The promotion of reform presented to the American people consisted of telling them that poverty in the US was largely the fault of "deadbeat dads" who had successfully transferred financial responsibility for their children to the American taxpayer. The expensive new child support enforcement system was called an "investment" aimed at forcing these fathers to live up to their responsibilities to the relief of the American taxpayer. Instead, it's added around $3 billion each year to their burden. The failure of the program to produce promised results was so predictable that one congressman said; "We're going to reform welfare in order to save money. As far as I can tell, it's going to be very expensive."
The most efficient thing to do is to eliminate the new system and focus on doing what's right the first time around. What this requires is exactly what Americans have always understood. Basic rights, including strong respect for due process of law, careful scrutiny to see that people are not treated unjustly, enforcing the necessary discipline against waste and error. The basic rules which most Americans believe to be fundamental to our Constitutional system should even protect groups of citizens who happen to find themselves at the cross-roads of bureaucratic interests against the ill feelings and biases created by intense propaganda campaigns.
The approach proposed in the sections above (based on Legal Construction) was taken in the Project for the Improvement of Child Support Litigation Technology (PICSLT) starting in 1989. A great deal was learned about tuning child support formulae to the American system. Generally, child support committee members have seen the application of mathematics to the child support question as something magical and transforming, as though once a report is produced alleging statistical support for one view, there is no longer any need for traditional legal processes.
But there most certainly is a need. The only way to properly apply mathematical decision models within the context of Constitutional justice is to fully disclose the nature of the mathematics, the underlying reasoning, and the assumptions in such a way as to make their review practical in comparison with the circumstances of each case. The only acceptable way to reduce the number of deviations from guideline amounts is to continue to improve the decision models so that they do a better job. In addition, making it easier to identify situations in which deviations are appropriate and developing simple ways to calculate deviations would improve the overall efficiency of the process.
As it turns out, developing a model within the discipline the United States Constitution provides an excellent body of theory useful to other countries in the world. Whereas other countries have developed models specifically fitting their economic / political structure and tuned to their current set of welfare state benefits, the Constitution simply tells us that we are required to "do the right thing" in each case. We are therefore under pressure to dig deeply into the question to determine what the right thing is in the greatest possible array of circumstances. This means developing the most general and complete theory.
In recent study within PICSLT, it has been found that models fitting a wide range of political / economic systems can be explained by the theory developed within the project. Beginning with the most complete model, we simply eliminate variables that do not apply in more controlled economies, and account for any array of welfare state benefits by methods which are already an integral part of the theory. This confirms the view given in the previous paragraph.
The scientific approach to developing child support science and technology should be understood as a parallel to the established judicial process, what we call "due process". Mathematical models are a precise, formal way of expressing concepts and relationships. "Precise" is not synonymous with "just" or "appropriate". To produce valid results ("just and appropriate" in the words of the Family Support Act) successful testing is required. In order to test, some set of independent criteria for deciding what "just and appropriate" is must be developed (i.e. a child support policy to which a presumptively correct "guideline" must conform).
It was disconcerting to see how quickly very simple child support formulae were accepted in the states. Even more so because their simplicity was given as one of their major selling points. Federal law (and the Constitution) require a just and appropriate award in each and every case. The goal is to construct guidelines that are sufficient to produce just and appropriate awards in every circumstance to which they are applied. It is required that judges can identify inappropriate and unjust results and that attorneys and parents can argue for deviation when a formula fails.No one should underestimate the task of adapting properly to the new requirements. The initial thrust, following OCSE / Williams recommendations for development of guidelines was a false start. In order to do the job properly, and in the quickest and most efficient manner, it will be necessary for legislatures and the courts to give the highest priority to the basic requirements of the Family Support Act ("just and appropriate award in each case") and the highest respect for Constitutional rights
3. Gay, Roger F., Child Support Guidelines: Resolving the Dilemma, A Summary Report on Design of Federally Mandated Child Support Schedules, Intelligent Systems Research Corporation; Special Report No. ISR-091490.01, Child Support Series Report No. 2, September 30, 1990.
4. Gay, Roger F. The Alimony Hidden in Child Support, New Scientific Proof that Many Child Support Awards are Too High, The Children's Advocate (NJCCR, Box 316, Pluckemin, NJ 07978-0316), January, 1995, Vol. 7 No. 5.
5. Williams, R. 1987. Development of Guidelines for Child Support Orders: Advisory Panel Recommendations and Final Report, II-68 to II-75. Washington, DC: U.S. Department of Health and Human Services, Office of Child Support Enforcement
6. New Equations for Calculating Child Support and Spousal Maintenance With Discussion on Child Support Guidelines, Final Report of the Project for Improvement of Child Support Litigation Technology, 1994.
9. Betson, David M. 1990. Alternative Estimates of the Cost of Children from the 1980-86 Consumer Expenditure Survey, U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation.
11. Final Report. Evaluation of Child Support Guidelines. Volume II: Findings of State Guideline Reviews, State Guideline Studies, and Unstructured Interviews. U.S. Department of health and Human Services, Administration for Children and Families, Office of Child Support Enforcement. March, 1996.
13. Gay, Roger F., Rational Basis is the Key Focus in Emerging 'Third Generation' Child Support Technology, in Proceedings of the Seventh Annual Conference of the Children's Rights Council, Holiday Inn, Bethesda, MD, April 28 - May 2, 1993. (This paper was provided to the Oklahoma House and Senate Judiciary Review Committee for child support guidelines prepared by Dr. Palumbo and provided last November to each member.)
21. During 1992 presidential debates, president George Bush said that he thought maybe the "deadbeat dad thing" was the problem Americans were most concerned about. To that, candidate Bill Clinton made the famous remark, "It's the economy, stupid." Politicians can apparently have short memories. President Bill Clinton continued to rely heavily on "deadbeat dad thing" through his 1996 re-election campaign.
22. "Parallel Unilateral Policy Declarations - Bilateral Arrangements as an Alternative to Conventions on the Enforcement of Support (Maintenance) Obligations", Hague Conference on Private International Law (13-17 November 1995), Working Document No. 2 submitted by the Delegation of the United States, Special Commission on maintenance obligations, 13 November 1995.
23. The Uniform Reciprocal Enforcement of Support Act (URESA) was first developed in 1950 by the National Conference of Commissioners on Uniform State Laws (NCCUSL), and was revised significantly in 1968 (RURESA). In August, 1992, an almost wholly new Act was completed to replace URESA/RURESA and was renamed the Uniform Interstate Family Support Act (UIFSA). UIFSA: "a foreign jurisdiction that has established procedures for the issuance and enforcement of support orders which are substantially similar to the procedures under this act." By this simple addition to the definition, the reach of the enforcement process of the states was greatly expanded.
24. Example; In 1994, a paragraph was added to Swedish law stating that recognition would be given to orders written in Australia, America's United States including Guam, The Virgin Islands, Puerto Rico and Samoa, with the exception of Alabama, District of Columbia, Mississippi and Nevada. (1976:108 with changes 1977:743 and 1994:1327)
25. Report made by the U.S. Delegation via the U.S. Embassy at the Hague to the Department of State in December 1995, Subject: International Enforcement of Family / Child Support: Meeting of Hague Conference Special Commission [November 13-17, 1995] to Examine the Functioning of Various International Conventions Concerned with Enforcement and Recognition. (Within the context of the Hague Conference on Private International Law.)
26. As discussed below, the compliance rate in the US is approx. 80% of what is ordered is paid. Generally, investigations by PICSLT indicate that overall compliance with child support orders is well correlated to the economy. It seems apparent that this relatively high compliance rate in the US is a result of the generally high level of wealth enjoyed by its citizens. Non-compliance is well correlated with inability to pay.
28. Although according to the data used in that report, child support had been awarded for only 56% of all separated custodial parents. Part of the lack of support orders however, can be explained by the death of an ex-spouse, agreement not requiring a court order, and other reasons. A significant part however is simply because paternity has not been established.
29. Braver, Sanford, Pamela J. Fitzpatrick, and R. Curtis Bay, 1988, Non-Custodial Parent's Report of Child Support Payments, presented at the Symposium "Adaptation of the Non-Custodial Parent: Patterns Over Time" at the American Psychological Association Convention, Atlanta, GA, August, 1988. Compared Bureau of Census custodial parents reports (approx. 70% received) with father survey (approx. 90% paid). Part of the payments goes directly to the welfare system as reimbursement rather than directly to the custodial parent.
31. It is estimated that approximately 300,000 have lost regular contact with their fathers due to visitation interference. It is unfortunately too often the case that custodial mothers who move out of state do so without making adequate arrangements for visitation, often do not notify the non-custodial parent of the move even if required by law, and may leave intermediate forwarding addresses and have unlisted phone numbers. National parent finder centers typically deny services to non-custodial parents as a matter of federal policy.