Roger F. Gay
October 27, 1999

Ask a number of people who are engaged in development, use, and analysis of child support guidelines what child support "science" is and you are likely to get a variety of responses. I suspect the most common would be a nervous chuckle.

If you ask a number of scientists what science is, you would also get a variety of responses. Many years ago, I looked at a book that was intended to help young people interested in science understand what science is. In the opening section of the opening chapter, it explained that science is what scientists do. Since then, I've wondered how someone knowledgeable about science could ever feel satisfied with such a circular definition. Aren't scientists people who do "science?"

Today, the questions "what is child support science" and even whether there is or can be a science of child support are not being rigorously debated. In fact, I haven't noticed much discussion on the question at all. So, why is it important?

There is confusion as to whether current child support guidelines give reasonable results and whether their foundation is valid. Some courts have given up on reason and declared the very concept of child support to be "subjective." Others have a desire to be very objective in the development of child support guidelines. When presenting number tables developed using statistical tricks, some guideline developers claim that what they have built is the result of "science." Others claim that the numbers that have been delivered are pure nonsense.

The controversy exists even though it has not yet been widely recognized and discussed. Differing views are held and widespread confusion is likely to remain so long as we do not develop a deeper collaborative understanding of the science of child support. Without discussion and consensus at a basic level, we can expect the variety of specialists (lawyers, economists, engineers, collection entrepreneurs, administrators and judges, etc.) to remain confused about the importance and validity of concepts, results, theories and hypotheses, conjecture and speculation that they are not themselves thoroughly familiar with.

It might not have mattered so much. Traditional child support law allowed courts greater discretion to determine just and appropriate awards. Statues were constructed with necessary and sufficient rules and definitions. They did not go so far as to dictate a presumptively correct award. With more than a century of experience, careful reasoning, and recorded precedents, traditional child support law had through practice, developed a rather scientific conceptual understanding of the child support question and the details of making an appropriate award.

The fact that courts were allowed great discretion in deciding what a just and appropriate child support award is, also allowed the judiciary to engage in basic reasoning about the question. More than that, our constitutional traditions of due process and protection against arbitrary and inappropriate orders forced a search for answers that could be regarded as correct. No matter how much their process resembles science however, judges did not consider themselves to be part of a scientific establishment. Their precedent setting decisions were not counted as natural law but only as judgments in the context of man-made statutes.

In principle then, the results of long intellectual experience were not considered to be basic knowledge but only temporary rules subject to change as the result of political decisions. The decision to change the basic rules came about in a series of odd political events in the 1970s and 1980s.

The national Office of Child Support Enforcement was established in the 1970s. Since its jurisdiction included all child support orders in the United States, this development brought the federal government fully into domestic relations law for the first time in history. By the mid-1980s, OCSE had established its own political influence and was on its way to becoming the vast bureaucratic empire that it is today.

In 1984, the federal government passed the Child Support Enforcement Amendments to the Social Security Act, which required every state (as a condition for receiving federal funds) to develop a formula for calculating child support awards (for both welfare and non-welfare cases) and to distribute the "guideline" state-wide. The Office of Child Support Enforcement was given the task of developing a report that would provide states with "technical assistance" in developing their guidelines.

Based on a decision by the National Center for State Courts, the Office of Child Support Enforcement published a report written from Policy Studies, Inc., (PSI) the parent company of a child support collection agency. The recommendations of the company's owner, Robert G. Williams, became the basis of child support formulae throughout the country. Those states that did not simply copy his model modified their formulae to bring results closer to his.

The federal government did not mandate changes to the concepts and definitions already established in the states as the basis for making child support awards. Nonetheless, Mr. Williams' report presented a model with the explicit goal of increasing the amount of child support awarded by 250 percent. The inappropriateness of basing his "technical assistance" on this arbitrary and contrary goal was paralleled by the bizarre nature of his data and analysis.

To make matters worse, the process included an advisory group that recommended a set of principles for child support. Not only did the PSI model fail to correspond to established child support law, it did not correspond to the recommendations of the advisory panel either.

In 1988, Congress passed the Family Support Act, which required that state child support "guidelines" be used in making all child support decisions. In addition, the result of calculation would be presumed to be the correct amount of child support to be awarded.

The PSI model has never been validated. Why have states accepted the PSI model? Williams and his supporters claimed that his model gave not only credible but correct results because "they are based on real data." Supporters of the PSI model pretended to have "established" its creditability in a way that many regarded as superior to the experience of the judiciary. It was regarded as something based on "science."

As a practical matter then, the questions are already on the table. Is it "science?" What is the science of child support? Does it exist? Can it ever exist? And it is important to analyze and respond to the questions. Is the PSI model valid? Can it be validated or invalidated? Is it impossible to validate a child support decision formula, rendering the view of at least one federal court -- that child support decisions are purely "subjective" (arbitrary) -- correct? If child support decisions can be validated scientifically, what is required?

Discuss these questions in the PICSLT Discussion Forum