Yesterday, I highlighted a study and asked readers to examine it. The study authored by Amanda Farahany and Tanya McAdams, both of Barrett & Farahany, LLP, entitled “Analysis of Employment Discrimination Claims for Cases in Which an Order was Issued on Defendant’s Motion for Summary Judgment in 2011 and 2012 in the U.S. District Court for the Northern District of Georgia” makes an important contribution to the empirical literature on the subject of resolving employment cases in the federal courts by summary judgment rather than trial. In this post, I will provide a brief summary of the study, and then add several observations.
A Caveat and Then a Brief Summary of the Study
First the caveat and then the brief summary. The following is not intended as an extended analysis of the statistics generated by the study. I have omitted a lot of information in the interest of brevity and for that reason only. Indeed, that is why I requested that readers of this blog examine the study themselves. That said, here is my brief overview:
*In 2011 and 2012, of the 181 employment discrimination cases in which plaintiffs were represented by counsel and summary judgment motions were in play, 147 (81.21% rounded to 82%) cases were dismissed entirely, 170 (94%) cases were dismissed in whole or in part and 11 (6%) cases survived entirely intact.
*In 2011 and 2012, in race discrimination cases where the claimant was represented by counsel, out of 71 cases, 10 cases (14%) survived summary judgment.
*In 2011 and 2012, in race discrimination cases where the claimant was represented by counsel, black claimants survived summary judgment in 4 (7%) of the 56 cases filed by blacks. During that same time period, in race discrimination cases where the claimant was represented by counsel, white claimants survived summary judgment in 3 (30%) of the 10 cases filed by whites. The remaining 5 cases were comprised of other races or ethnic groups or were not identified by race or ethnicity. Regarding gender, and for that same time period and regarding race discrimination cases, black female claimants survived summary judgment in 1 out of 29 cases or 3% of such cases whereas black male claimants survived summary judgment in 3 out of 25 cases or 12% of such cases. White male claimants survived summary judgment in 3 out 5 cases or 60% of such cases, whereas white female claimants survived summary judgment in none (0%) of the 4 cases they brought.
*In 2011 and 2012 and for sex discrimination cases involving hostile work environment, sexual harassment and disparate treatment claims (lumped together by me) where the claimant was represented by counsel, out of 55 cases, 6 (11%) survived summary judgment.
*In 2011 and 2012, in sex discrimination cases where the claimant was represented by counsel, I could not find a break down by the race of the claimant. Regarding gender, and for that same time period regarding sex discrimination cases, female claimants survived summary judgment in 6 out of 48 cases or 12.5% of such cases whereas male claimants survived summary judgment in none (0%) of the 4 cases they brought. In “multi-female” class cases, none (0%) survived summary judgment in the 3 cases they brought.
*In 2011 and 2012, the great majority of judges in the Northern District of Georgia granted dismissals in 75% percent or more of all race and sex discrimination cases.
Kopf’s Preliminary Observations
*In a comment to this blog, there is a veiled suggestion from one of the authors of the study that an observer might read into the statistics a discriminatory animus on the part of the Georgia federal judges based on the race or sex of the claimant. See Amanda A. Farahany, October 21, 2013 at 11:53 am, here. I categorically reject that suggestion, if indeed that is what the author intended to imply. The numbers are just too small to make such a claim. For example, only 10 cases of race discrimination involved whites, and only 3 of those survived summary judgment. To use those tiny numbers to suggest judges favored whites because the white dismissal rate was only 70% whereas the black dismissal rate was 93% is weak support indeed for a claim of racial animus. To illustrate further, if you want to use tiny numbers, black females had a 100% greater chance of prevailing than white female claimants of race discrimination, but the absolute difference is 1 case as no white female claimants survived summary judgment while 1 out of 29 black female claimants survived summary judgment. Moreover, given that the great majority of judges dismissed these cases at rate equal to 75% or more of the time, it is impossible to believe that most of the judges in the Northern District of Georgia act because of a racial or gender animus.
*According to the Atlanta Journal Constitution, “The U.S. District Court headquartered in Atlanta currently has eight active judges, five of whom were appointed by Democratic presidents and three of whom were appointed by Republican presidents. Title VII cases are also heard by the court’s eight senior judges; six of these judges were appointed by Democratic presidents, two by Republicans. The magistrates are appointed by district court judges.” As a result, it is impossible to believe that the results of the study represent the actions of judges with the same ideological bent.
*In a comment to this blog, see Amanda A. Farahany, October 21, 2013 at 11:53 am, here, one of the authors of this study has written: “Having heard that the problem is the law of our circuit, we expanded the study to the Northern District of Alabama, a state with no state laws (like Georgia) and within our circuit. With essentially the same sample size, rather than dismissing 82%, 66% were dismissed in full. Rather than letting 6% past summary judgment unscathed, Alabama let 12%. Statistician Dr. Michael Freeman has already concluded that this a statistically significant difference.” I have several reactions to this comment and address them to the authors. A. Please publish the complete statistical analysis for the Northern District of Alabama so that it can be reviewed and tested by others. (Please give due attention to the statistician’s definition of “statistical significance” and the statistical method employed to arrive at such significance.) B. In absolute numbers, the difference between 6% (11 cases) in Georgia and 12% (22 cases (rounded)) in Alabama is only 11 cases. As a result, I would not be willing to conclude that something is wrong with the judges in the Northern District of Georgia on the basis of 11 cases spread among 16 active and senior district judges over a 2-year period even if the results are “statistically significant.” C. Finally, is the number of “unscathed” cases in the Northern District of Alabama about right, too high or too low? What is the statistic that federal trial judges should shoot for and how do you derive that number?
*My experience leads me to believe that most race and sex discrimination claims against most employers are legally baseless. By this, I do not mean to suggest that most employees or job-seekers are dishonest or that most employees or job-seekers may not have legitimate (but not legally cognizable) grievances. However, the federal law (whether you agree with it or not) does not make employers responsible for work place disputes unless the employer can reasonably be expected to have known and approved of conduct prohibited by the federal law. In short, these are hard cases to get to a jury because derivative liability (you pay me because somebody who worked for you hurt me) is manifestly not the standard in federal employment law.
*Despite my significant reservations about this study, it is a serious attempt to get at a serious subject. Civil jury trials in the federal courts are rare indeed and that is particularly true in the employment context. This study shows us how rare they are in one large metropolitan federal court and presumably throughout the country. I sincerely thank Vince Powers and Barrett & Farahany, LLP for giving me access to the study and for their commitment to the rule of law.
*The reader is reminded that these are my preliminary observations. I am willing to learn more. If I have erred, I am also willing to be corrected.