Nihilism by the Numbers
John Goodman in "The Big Lebowski"
The Bankruptcy by the Numbers column and its authors recently received a singular distinction. Prof. Jay Westbrook, writing in the Texas Law Review, cited some of our work as an example, of all things, of nihilism.1 Based on comments in one of our columns,2 Prof. Westbrook placed us among a group of commentators on consumer bankruptcy who, in his view, are misguided about the relationship between bankruptcy facts and bankruptcy policy. He leaves for us alone, however, a position "beyond" the others, "a kind of post-modernist nihilism," from which, in his view, "[we] seem to conclude that data should not influence policy."3
We were honored to be listed among a group of scholars that included Margaret Howard, Douglas Baird and James White, whose work we admire. But we were puzzled that we would qualify for membership in a fancy philosophical club, and we had no idea what membership might require (dues? initiation rites? a post-modernist nihilist's handshake?). We weren't quite sure what it meant to be a post-modernist nihilist. So we did our homework. The result was clear: Though flattered by the nomination, we must decline membership. No matter how we try, there is no way that our position on the relationship between fact and policy in bankruptcy fits into nihilism, post-modernist or otherwise.4
This matter would not be worth pausing over, except that it provides an opportunity to discuss briefly what is important: the uses of quantitative analysis for developing and evaluating bankruptcy policy. These are very practical matters. They are the topics of Prof. Westbrook's article, and they stimulate the work that is published in this column.
Here we emphasize just one point. It is the same point that we made in the earlier article that led to Prof. Westbrook's effort to find a philosophical home for us. The point is this: The results of quantitative research in bankruptcy will never, by themselves, be sufficient to conclude a policy debate that has an action step (e.g., legislation, rule-making, administrative change) as a goal. Without the addition of another ingredient, the quantitative results will be relatively inert and subject to interminable criticisms by the parties who are contending in the policy or legislative arena.
No matter how we try, there is no way that our position on the relationship between fact and policy in bankruptcy fits into nihilism, post-modernist or otherwise.
What is this additional ingredient, the yeast that lets the data rise? It is a value judgment: a norm, criterion, threshold, ceiling or other judgment that links the way things are, as demonstrated by the research results, with the way things should be, as advocated by the proponents or opponents of the policy or legislation at issue. It is in the very nature of things that value judgments have their roots outside of the empirical domain of the research results. They are normative or moral positions that must be translated into terms that apply to the research results. Value judgments give the empirical work its meaning in policy and law.
We believed that this was an uncontroversial and very practical point. Apparently, there are those who believe we were wrong. So here, we demonstrate how practical it is, by providing two examples that arise from our own research experiences.
Example: The Consequences of Consumer Means-testing under Proposed Bankruptcy Reform Legislation
Between 1997 and 2000, as the arguments about means-testing of consumer debtors took center stage, four separate sets of research results were offered to provide estimates of how much money would be returned to unsecured creditors by chapter 13 debtors who, but for the (hypothesized) passage of the means-testing legislation, would have received chapter 7 discharges in no-asset cases. Bankruptcy by the Numbers reported on some aspects this work.5
The results of the four studies varied remarkably, ranging in estimated annual returns from $5 billion or more down to less than $1 billion. Much of the commentary about the studies focused on differences between research methodologies in the studies and whether some of the researchers, whose work had been paid for by commercial unsecured creditors, should have been more forthcoming with the details of their data.6 When the dust settled, however, the data had been left behind as the legislative debates moved to other contentious issues.
The lack of agreement among the studies may have been one reason their results did not have more impact in later legislative debates or editorial discussions. But even if the studies had produced very similar results, they would not have figured largely in the subsequent discussions unless the missing ingredient had been supplied. In this instance, the ingredient would have raised and debated the value question: How much does the imposition of means-testing have to return to unsecured creditors in order to warrant making the legislative change? Without addressing this question directly, research estimates of how much the program will return to creditors do not have much "punch" in the policy debate.
Putting the question this way may seem to reduce the question of means-testing to an oversimplified cost-benefit analysis. But this appearance, if it exists, would be misleading. To begin with, it is unlikely that anyone would argue that the means-testing program should be instituted if it would cost more than $100 in additional means-testing administrative overhead to return $100 to unsecured creditors. So there is an obvious cost-benefit aspect of the value judgment. But beyond that, some might argue that so long as the costs of the program were reasonable (or some other term of art), then the quantification of benefit as against cost should not be determinative, because debtors who can repay their debts should repay their debts, and the bankruptcy system should be designed to ensure that they do. This position takes the argument out of the simple cost-benefit framework, in part by leaving the argument about what "reasonable" means for a later day or forum.
The point is that the quantitative research results were not invigorated by such debates, which would have led to value judgments about the results. So advocates on both sides of the means-testing debate were free to ignore the results they didn't like. The numbers had not been displayed in a context that would have made them impossible to ignore.
In the next example, by contrast, the appropriate value question was asked and answered, and on that basis the research results became extremely useful.
Example: The Bankruptcy Court Time Study
In 1988, the Committee on the Administration of the Bankruptcy System of the Judicial Conference of the United States asked the Federal Judicial Center (FJC) to measure the workloads of all active bankruptcy judges. The purpose of the study was to provide an objective foundation for making and justifying requests for new bankruptcy judgeships. During 10-week periods between October 1988 and October 1989, 272 judges filled out extensive time records of all of their work-related activity, both case-specific and administrative. The results allowed accurate accounts to be made of the time judges spent on cases of varying sizes in each bankruptcy chapter as well as on different categories of adversary proceedings irrespective of chapter. Summing up these times over all the categories allowed an accurate estimate to be made of how many case-related and administrative hours judges in each court worked on average over a 12-month period.7
The Federal Judicial Center presented the quantitative findings to the Bankruptcy Committee, and the committee provided the missing ingredient. In this instance, the committee answered the value question: What is the threshold number of case-related hours bankruptcy judges in a given court should work annually before the Judicial Conference will consider requesting one or more additional judgeships for that court? The committee made that decision on the basis of many factors that included, but were not limited to, the results of the study. Once they made that value judgment, the study results could be put into use, after transfer to the Administrative Office of the U.S. Courts, for periodic review of the weighted caseloads of each bankruptcy court in relation to the number of judges serving there. The formula for weighted caseload, animated by the value judgment made by the Bankruptcy Committee, created a robust administrative system that has remained in use for more than a decade.
Without the informed value judgment made by the committee, the research results would have been subject to interminable, largely fruitless controversy. This is because, as noted in the publication of the results, "[i]t is not necessarily true that the amount of time judges should spend on cases is measured by how much time they are spending on them."8 The FJC's time study could only assess the actual expenditures; it remained for the Bankruptcy Committee to anchor those expenditures, i.e., the way things were, to the value judgment, i.e., the way things should be.
These two examples could be supplemented with others, within the bankruptcy world and in other policy research areas, to show many variations on our theme. The fundamental point would remain the same, just as we said in our earlier article: "This is why studies about the causes of bankruptcy provide ambiguous or insufficient guidance for answering bankruptcy policy questions. The data always require interpretations that include a set of assumptions that go beyond the numbers themselves."9 This is not nihilism, nor is it pessimism. It is realism.
Our position affirms what nihilism denies. Contrary to nihilism, we acknowledge and avow the central role of value in the construction and implementation of sound public policy. We emphasize that no matter how you collect and then slice the numbers, you will never find the requisite value judgments that are required to use empirical information in practical affairs. The purposes of "bankruptcy by the numbers" are several, as Prof. Westbrook has emphasized and exemplified in his own well-known work over the years as a contributor to the Consumer Bankruptcy Project. But they do not, and cannot, include creating policy judgments without the establishment of norms, criteria or other value judgments that originate from outside the empirical information itself.
1 Westbrook, Jay Lawrence, "Empirical Research in Consumer Bankruptcy," 80 Texas L. Rev. 2123 (2002). Return to article
4 Within philosophy, nihilism is "extreme skepticism; the denial of all real existence or the possibility of an objective basis for truth." (Random House Dictionary of the English Language, Second Edition Unabridged). But the academic elaborations on nihilism are as thick and daunting as the topic itself. One readily accessible and informative resource is The Internet Encyclopedia of Philosophy (www.utm.edu/ research/iep/n/nihilism.htm). From this source, we learn that the post-modernist version of nihilism is associated with the so-called Antifoundationalists. (Imagine putting that on your business card.) Antifoundationalists are either "mildly annoyed" or "upbeat" about having to accept as meaninglessness the nature of things. As demonstrated by the current article, we don't accept as meaninglessness the nature of things. But we are usually upbeat, and we become mildly annoyed on occasion. Does this mean we could be junior members of the post-modernist nihilists' club? Return to article
5 Flynn, Ed and Bermant, Gordon, "Estimating Means-tested Chapter 13 Case Yields from Current Chapter 13 Performance," 19 Amer. Bnkry. Inst. J. 22 (June, 2000). The most complete and balanced presentation of our work on this subject, as well as of the other three studies, was made by the General Accounting Office in its report, Personal Bankruptcy: Analysis of Four Reports on Chapter 7 Debtors' Ability to Pay. GAO/GGD-99-103 (June 1999). Return to article
6 See, e.g., Westbrook, supra n. 1 at 2142. For a review of how various contributors to these debates argued about (or without) data, see Howard, Margaret, "Bankruptcy Empiricism: Lighthouse Still No Good," 17 Bankr. Dev. J. 425 (2001). Return to article
7 Bermant, Gordon, Lombard, Patricia A. and Wiggins, Elizabeth C., "A Day in the Life: The Federal Judicial Center's 1988-1989 Bankruptcy Court Time Study." 65 Amer. Bnkry. L.J. 491(1991). Return to article