Introduction
A competent decision in a child support case involves the complex consideration of a wide variety of factors. In order to make such a decision, attorneys and judges must apply their skills to extract essential and sufficient information from litigants and understand the proper application of the information they collect. That will be true regardless of the technology used to calculate an award. Modern attempts to change the decision process using crude statistical models have reduced complexity for the sake of an odd sort of consistency. Although child support guidelines themselves show an impressive self-consistency, there is no longer any concrete relationship between an award and the wide variety of factors that are important in making a reasonable decision.

The Project for the Improvement of Child Support Litigation Technology has demonstrated that the logic of traditional child support decision-making can, to a large extent, be transformed into a concrete science. The application of such a science in the training of judges and attorneys can result in a more desirable sort of consistency. Each award should be just and appropriate, given consideration of the individual circumstances in each case. Just and appropriate results require an understanding of the detailed logic used in reaching a competent child support award decision. With that understanding, it should be possible to produce a similar award in different courts on different days in consideration of a similar set of facts.

In this article, I would like to summarize the work of Project for the Improvement of Child Support Litigation Technology (PICSLT) toward developing better child support guidelines, and provide an overview of the PICSLT model. Detailed theoretical development in the PICSLT work has led to a solid definition for the boundary between child support and alimony. This development provides scientific proof that many child support awards contain a hidden margin of alimony, and are thus too high in a common, legal sense.

History of the Project:
Project for the Improvement of Child Support Litigation Technology began in 1989. It was at this time that the United States was on the brink of a major change in the way child support award decisions are made. A federal law, known as the Family Support Act of 1988, required each state to base every award decision on formulae known as "child support guidelines." As a result of this legislation, award amounts have increased dramatically from those awarded according to the legal principles that had been established in the states.

The most fundamental problem resulting from the federal legislation is that rigid mathematical formulae have replaced the rational principles upon which child support decisions had been made. In traditional child support statutes, a definition of child support was given along with a set of principles for guiding complicated decisions. A traditional definition would state that child support is an amount paid by a non-custodial parent for his or her share of the actual and necessary needs of children. Additional guiding principles could include a reasonable consideration for sheltering children from the standard of living loss that accompanies divorce, and that both parents have an equal duty to support their children. Statutes could also explicitly include such considerations as the need for each parent to support themselves, time the non-custodial parent spends with their children, and travel expenses involved with visitation.

New child support statutes typically do not provide an alternative definition for child support. They are based on arbitrary analysis of national data on family spending and do not correspond to any set of rational principles for making an award decision. Therefore, it has been necessary to rely on traditional principles for research on developing better child support guidelines. It was discovered early in the project that there are two concepts that are fundamentally important to traditional thinking. The first is the "equal duty principle" (both parents have an equal duty to support their children). The second is "ability to pay." One deals with the issue of fairness, the other with practicality. Both are needed as the basis of a good judgment.

It is simplest to describe the equal duty principle by first saying what it does not mean. It does not mean that both parents should pay the same amount toward support of their children. The award decision takes into consideration other important factors, including each parent's ability to pay. When all is considered, "ability to pay" is not equivalent to income, as it appears to be in the Income-Shares (example: New Jersey guidelines) and Percent-of-Income (example: Wisconsin guidelines) formulae. (These are the most commonly used types.) There was an established prohibition against taking from a parent for support of children, so much that a parent is no longer capable of self-support.

Many of the portions of the current PICSLT model can be found in previous work. Anyone wishing to delve deeply into the detailed mathematics of child support, should begin by reading "How to Calculate Child Support," by Maurice Franks (Case & Comment, January-February, 1981). Franks provided the most complete Income-Shares model ever published, which included detailed mathematics for accounting for children's time with each parent, how to deal with "extraordinary" expenses (expenses that are not included in the standard table), and joint custody. Oddly enough, more recent authors of Income-Shares models have claimed an inability to perform these simple operations.

In the limited way of the Income-Shares approach, Franks dealt very logically with the equal duty principle. The first step toward improvement upon Franks' model was to replace income with a more sophisticated view of ability to pay. Some newer models deduct a standard amount required for support of one adult from each parent's net income before performing the child support calculation. The remaining income represents "ability to pay."

There are reasons to believe that this approach is too simple. If a parent uses part of his remaining income to purchase tools necessary for work, for example, that amount is additionally needed for self-support. Ability to pay a "standard" amount of child support is also changed by "extraordinary" expenses. When a parent must pay an extraordinary amount of medical expenses, for example, their ability to pay for standard expenses can be significantly reduced. The PICSLT model uses a few simple mathematical operations to account for a wide range of circumstances that effect actual "ability to pay."

Several new equations were developed for the first version PICSLT model that were later replaced by a completely new basic formula for child support. Of the technology that remains in the most recent version, there is only one more major area to discuss--the numeric table. This important component had also received attention in previous work by others. A most intriguing view was expressed in a report for the Washington State Association of Superior Court Judges (1982). The author of the report, William Hewitt, noted that data and methods used to estimate the amounts families spend on children were woefully inadequate. Yet most numeric tables used in guidelines today are a direct product of such estimates.

The better approach is to build a standard table giving separate costs related to different spending categories; housing, transportation, food, etc. The amounts listed in each category should be determined from court case experience in which many individual cases have been decided in full view of all relevant facts. Another advantage identified in the PICSLT work is that the categorical approach provides a way of comparing individual case circumstances with the amounts listed in the standard table. If for example, a divorce settlement provides a custodial parent with a house that is paid for, then understanding the details of the number listed under "housing" can greatly assist a judge in adjusting the award. A detailed look at modern cost analysis is given by Robert Braid (see The Children's Advocate, November 1994, Vol. 7, No. 3).

PICSLT has not studied a large number of real cases to develop a table. In order to estimate appropriate values, another approach described by Hewitt has been used. It begins with estimates of spending on children from national data, separated into spending categories. PICSLT selected estimates of spending on children by single parents developed by the USDA. The USDA work is exceptionally clear in explaining precisely how the numbers in each category are derived. It is better to start with a clear understanding of what one has, so that it can clearly be compared with what we would like to have. The numbers are then reduced in proportion to the marginal cost of children, with each category receiving an independent appraisal. We then have a table with values that can each be understood in the context of a rational child support decision process and compared with evidence provided in each case.

The first version PICSLT model can be seen largely as arising from the integration of selected portions of other models. Reasonable modifications were made to complete the integration. Special adjustments were needed to assure that a low income mother is able to maintain a household, for example, even after a support payment is reduced for visitation time. An additional formula was created to adjust child support payments to show what could be done to reduce welfare dependency. Although joint and sole custody cases could be handled with the same model, special additions were needed when either parent remarried or had other children to support.

The Heart of the New Model
It was immediately apparent that it would be nicer to rebuild the model such that it was less of a "house that jack built." But there were two compelling reasons to think that integration of the best of the existing best was not good enough.

First, there is the issue of sheltering children from the standard of living loss that accompanies divorce. This means increasing the award by some amount that is considered "reasonable" in light of the payer's ability to pay. It remained an arbitrary decision, and its arbitrary nature is the root of the current political problem with child support. How much is enough? How much is too much? These questions were unanswered.

Second is the political reason. The popular political philosophy is that more child support is always better. Awards granted in state courts have risen dramatically as a result of new child support formulae. Lobbying groups, operating on behalf of single mothers, have been pushing for further increases. The Federal Government has been considering the creation of a national child support guideline. Many politicians have supported the notion that a national guideline should result in awards even higher than those currently forced upon the states. Should the courts decide, on a constitutional basis, that there are limits to the amount of child support that can be awarded? What factual or scientific information is needed to make such a decision?

Over the years, many people have viewed the new increased child support amounts as containing a hidden margin of alimony. This would seem apparent, but how does one find the boundary between the two?

In the PICSLT research, the boundary was defined as being the point at which any additional payment would be primarily for the benefit of the custodial parent rather than the children.

This definition follows from case law: "the money is for the support and welfare of the children, not for the enrichment of the custodial parent" (Oregon Court of Appeals, 1987).

The two important questions became: "Does such a boundary exist?" And if so, "How do you find it?" The boundary between child support and alimony is a solid one, if we rely on custodial parent spending patterns to find it. Take the basic child support amount to be the non-custodial parent's share of what the custodial parent would spend on her own. Payment of child support increases the standard of living (i.e. income) of the custodial parent household. This in turn, increases the amount we would expect to be spent on their children by the custodial parent. The increase in spending justifies increasing the amount of child support awarded. We can use this argument over and over, to continue to justify a greater and greater increase from the basic child support amount, and thus continue to increase the standard of living in the custodial household. But there is a limit.

To illustrate that this limit exists, let us say that a custodial mother has a take home pay of $18,000, a non-custodial father has a take home pay of $25,000, and the mother spends 20 percent of her income on a child. In other words, she spends $3,600 if she receives no child support from the father. If $100 dollars a year is paid in child support, we can easily see that it is 100 percent child support. $100 is just a small portion of the $3,600 the mother is spending. In fact, applying the "equal duty principle," it is just as obvious that a payment of a much larger portion of that $3,600 (in proportion to the father's relative ability to pay) is still 100 percent child support. This fact is easily seen, because it is just his share of what the mother actually spends on their child. (Visitation and joint custody are outside this example.)

The payment of child support provides more income for the custodial parent, which we expect to increase spending on the child by 20 percent of the payment in this example. If we use $2,000 in child support for our example, we should expect an increase in spending on the child of $400. If the father pays an additional amount, equal to his share of this $400, the additional amount is also 100 percent child support. It is still nothing more than a part of his share of what is actually spent on the child. As you can see however, the add-on is far less than the original $2,000. Each iteration results in an add-on amount that is less than the last until the limit is reached. Any additional dollar in excess of the limit in this example provides only 20 cents in child support instead of a dollar. The remaining 80 cents is alimony.

In a more complete and detailed examination of this example, the result from a typical Income-Shares model was compared to the limit. The Income-Shares model produced a result of $5,237.40, which was $2,309.48 above the limit. Just for the sake of simplicity, the example did not consider visitation time. Since most Income-Shares formulae do not account for typical visitation periods, and the PICSLT model does, the amount of alimony found in a real case at these income levels would probably be higher.

Alimony in the Balance
Custodial parents with low income have an additional need to maintain a household for children and for self-support. This is an example of when an award of spousal maintenance may be appropriate. It is a reasonable interpretation of government intent to believe that the government does not wish to support mothers or children when the father is capable of doing so. Given a mathematical definition of the boundary between child support and alimony, it was possible for the PICSLT work to go one step further.

Given a target standard of living for children, it is possible to calculate the right balance between child support and spousal maintenance to provide an additional increase in the standard of living of the custodial household. The new PICSLT model includes the equations for calculating both in the proper proportion to each other. The most general difference that would result from application of the PICSLT model can be stated as follows. The hidden margin of alimony in current child support awards has been identified, and proportionate amounts of child support and spousal maintenance can be awarded separately when appropriate. When spousal maintenance is not appropriate in a particular case, it need not be awarded as a hidden part of the child support award.

This article was first published in The Children's Advocate.

Roger F. Gay