Empirical Legal Studies

Civil Settlements

In A Comprehensive Theory of Civil Settlement, J.J. Prescott (Mich.) and Kathryn Spier (Harvard) dig deeply into the world of settlements and push-back against some conventional wisdom. In particular, they challenge the “'trial versus settlement' trope," and "describe and defend a much richer concept of settlement, amounting in effect to a continuum of possible agreements between litigants along many dimensions."

In addition, the authors "study partial settlements and how they interact with each other in real-world adjudication using new and unique data from New York’s summary jury trial program. Patterns in the data are consistent with parties using partial settlement terms both as substitutes and as complements for other terms, depending on the context, and suggest that entering into a partial settlement can reduce the attractiveness of full settlement."

Patent Litigation, Remedies, and the eBay Decision

In the patent world, the Court's 2006 eBay decision is widely regarded as a pivotal moment, particularly as it relates to the likelihood of a prevailing party receiving a permanent injunction against an infringer. In eBay, the Court, rejecting the prevailing “general rule” that a prevailing patentee is entitled to an injunction, articulated a four-part test that lower courts must apply before granting such relief. Despite this critical legal change, eBay's effects have remained comparatively understudied.

In Permanent Injunctions in Patent Litigation After eBay: An Empirical Study, Christopher Seaman (Wash & Lee) presents results from a study of more than seven years of contested permanent injunction decisions in district courts after eBay. An excerpted abstract, summarizing the key findings, follows.

"... eBay has effectively created a bifurcated regime for patent remedies, where operating companies who compete against an infringer still obtain permanent injunctions in the vast majority of cases that are successfully litigated to judgment. In contrast, non-practicing entities almost always are denied injunctive relief. These findings are robust even after controlling for the field of patented technology and the particular court that decided the injunction request. It also finds that permanent injunction rates vary significantly based on patented technology and forum."

Defining "Statistical Significance"

While we can all agree that trying to reduce the concept of statistical significance to a written definition poses challenges, a recent effort to do so by the U.S. Dept. of Health & Human Services (HHS) (here) looks like it does far more harm than good. Candidly, it resembles a definition produced by a committee of bureaucrats. Andrew Gelman's devastating critique is worth reading. I certainly hope that HHS didn't squander too many of our federal tax dollars in their flawed effort.

Data on the "Teaching v. Research" Trade-Off Debate

A very interesting paper from Tom Ginsburg and Tom Miles, both at Chicago, explores "teaching v. research" debate. In The Teaching/Research Trade-Off in Law: Data From the Right Tail, the authors test the (long-assumed) "trade-off" hypothesis with the benefit of 10 years of data from the University of Chicago Law School. A summary of the paper's main take-away follows.

“There has been much debate about whether the performance of American law schools is failing, according to various metrics. Critics argue that U.S. law schools subsidize useless research, taking resources away from the core activity of educating students. Our own school is alleged to be a part of the problem, as it is part of a set of elite schools that ‘‘set the market’’ for the legal professorate. A key assumption of these critics is that a trade-off exists between teaching and scholarship: That the time and energy spent on research impair the quality of teaching or that professors proficient at scholarship (or scholarship that is not directly related to legal practice) are poor teachers of law…. The results of our study cast doubt on this assumption, at least at one elite school. Our analysis of 10 years of data encompassing nearly 50 faculty members over almost 500 courses reveals no strongly negative relationship between the volume of scholarship and the amount or quality of teaching.”

Identifying/Rooting Out "Duplicate" Cases in a Data Set

In general, a surprising number of researchers fail to develop and execute a strategy for handling "duplicate" entries in a data set. For example, in assessing criminal outcomes researchers need to decide ex ante whether their dependent variable of interest is the "crime" or the "criminal." If it's the latter, due to recidivism, researchers then need to think through how to handle the possibility that a single individual may commit multiple, independent crimes. If so, a single criminal may appear more than once in a data set and, in so doing, raise potential "double count" issues. (Obviously, one may plausibly treat multiple crimes committed by the same individual as separate events and intentionally "double count.") In the event that the researcher wants to avoid "double counting," however, identifying and rooting out duplicate cases in large data sets can prove difficult. To this end, a nice (albeit abbreviated) discussion of how to use Stata's "duplicate" command is found here.

Statistical Add-Ins for Excel Users

With an array of statistical software products on the market (e.g., Stata, SPSS, SAS, R, etc.), I limit my use of Excel to data management tasks. That said, I know that others exploit Excel's full capacity and use it for some limited analytical and statistical tasks. To this end, I pass along information on a recent release of an add-in, RegressIt, which is designed to enrich (and ease) Excel's regression capabilities. Developed by a pair of B-School professors, Robert Nau (Duke) and John Butler (Texas), the product is available for free. A brief description follows; click here for more technical information.

“RegressIt © is a PC Excel add-in that performs multivariate descriptive data analysis and multiple linear regression analysis with presentation-quality output in native Excel format…. It is offered free of charge as a public service in order to teach and support good analytical practices, especially data and model visualization, tests of model assumptions, appropriate use of transformed variables in linear models,  intelligent formatting of tables and charts, and keeping a detailed and well-organized audit trail.

The RegressIt program file is an Excel macro (xlam) file that is less than 500K in size.  It runs on PC’s in Excel 2007, 2010 and 2013, and it can perform analyses and fit models with up to 50 variables at one time and up to 32766 rows of data.  You can use it on a Mac if you run it in the Windows version of Excel using VMware Fusion or other PC virtualization software.”

ELS and the Supreme Court

Empirical legal studies’ impact was on display last week in the Supreme Court’s death penalty decision, Glossip v. Gross. In particular, there is a lively back and forth among Breyer (joined by Ginsburg) in dissent, and the concurrences of Scalia and Thomas with no fewer than three JELS articles cited.

Suffice to say, Breyer and Ginsburg find the empirical evidence of problematic capital punishment application compelling, Thomas counters with questions about methods, and Scalia summarizes much of the scholarship as “abolitionist.” The link to the opinion is here. Cited article links follow.

“An Empirical Evaluation of the Connecticut Death Penalty System Since 1973, Are There Unlawful Racial, Gender, and Geographic Disparities?”  by John J. Donohue III in JELS 11:4

“Frequency and Predictors of False Conviction: Why We Know So Little, and New Data on Capital Cases” by Samuel R. Gross and Barbara O'Brien in JELS 5:4

“A Broken System: The Persistent Patterns of Reversals of Death Sentences in the United States” by Andrew Gelman, James S. Liebman, Valerie West, and Alexander Kiss in JELS 1:2 

Collapsing Variable Categories

A common, hands-on task that can be accomplished in various ways (some more parsimonious than others). An example (with coding illustrations and suggestions) recently emerged on the Stata list (here) might help those new to empirical methods and/or data management.