Thursday, July 9, 2020

Gough: New Empirical Work on Employment Arbitration

GoughMark Gough (Penn St. School of Labor & Employment Relations) has posted on the ILR Review website his article A Tale of Two Forums: Employment Discrimination Outcomes in Arbitration and Litigation, Industrial & Labor Relations Rev. (forthcoming 2020). Rather than just posting his abstract, I'll post instead a summary I asked Mark to draft for me, which helps situate this empirical work among other empirical work on similar topics:

Most of the empirical literature comparing outcomes between forums uses relatively crude descriptive statistics to show stark differences in employee win rates and monetary award amounts within the populations cases disposed of in arbitration and litigation. Indeed, scholars have provided robust evidence on the resolution of employment disputes within individual forums such as:

  • The American Arbitration Association (AAA) – see, e.g., Alexander J.S. Colvin, An Empirical Study of Employment Arbitration: Case Outcomes and Processes, 8 J. Empirical Legal Studies 1 (2011); Lisa B. Bingham, On Repeat Players, Adhesive Contracts, and the Use of Statistics in Judicial Review of Employment Arbitration Awards, 29 McGeorge L. Rev. 223 (1998).
  • The Financial Industry Regulatory Authority (FINRA) – see, e.g., J. Ryan Lamare & David B. Lipsky, Employment Arbitration in the Securities Industry: Lessons Drawn from Recent Empirical Research, 35 Berkeley J. Employ. & Labor L. 113 (2014); J. R. Lamare, & D. B. Lipsky, Resolving Discrimination Complaints in Employment Arbitration: An Analysis of the Experience in the Securities Industry, Industrial & Labor Relations Rev. (2018).
  • Federal court – see, e.g., Kevin M. Clermont & Stewart J. Schwab, How Employment Discrimination Plaintiffs Fare in Federal Court, 1 J. Empirical Legal Studies 429 (2004).
  • State court – see, e.g., Theodore Eisenberg & Elizabeth Hill, Arbitration and Litigation of Employment Claims: An Empirical Comparison, 58 Dispute Resolution J. 44 (2003).

These studies often are used to support the perceptions of arbitration as an employee-unfriendly forum. See, e.g., Mark Gough, How Do Organizational Environments and Mandatory Arbitration Shape Employment Case Selection? Evidence From an Experimental Vignette, 57 Industrial Relations 541 (2018); Mark Gough, Employment Lawyers and Mandatory Employment Arbitration: Facilitating or Forestalling Access to Justice, 16 Advances in Industrial Relations 133 (2016). And while informative, a limitation of this literature is it provides minimal controls to account for systematic variation between forums. It is clear that the average monetary award and employee success rates at trial are lower in arbitration than litigation, but are employee claimants genuinely at a disadvantage in arbitration? Or does systematic variation exist within the underlying merits of cases, presence or quality of counsel, party resources, or other case characteristics which account for differences in outcomes between arbitration and litigation? In short, one must be careful to compare “apples to apples” when drawing evaluative conclusions about arbitration’s effect(s) on access to justice.

In a 2020 empirical study, Mark Gough attempts such apples-to-apples comparisons by surveying 1,256 employment plaintiff attorneys about their most recent cases adjudicated in arbitration, state court, or federal court. Even while accounting for claim, plaintiff, defendant, and attorney characteristics, Gough finds employment discrimination plaintiffs in arbitration are less likely to receive a judgment in their favor and smaller awards compared to similar cases disposed in state and federal court. Specifically, he reports, “compared to arbitration, employees' odds of winning increase by 70.7 percent in a federal jury trial, 183.7 percent in a state judge-only bench trial, and 146.0 percent in a state jury trial…[and] relative to arbitration, monetary damages awarded to success


Arbitration, Employment Common Law, Employment Discrimination, Scholarship | Permalink


A commenter who wishes to remain anonymous writes:

I haven't read Gough's article and am too tied up with other matters to do so right now, but one big question jumps out from the summary you posted. The article compares results of "similar cases" in federal and state courts and in arbitration. The author's conclusions would be strong if, in addition to the statistical factors used to determine similarity, a similar percentage of cases filed get to decision in those three forums.

For example, if there were 100 cases filed in each of the forums and in each 10 of them went to decision, then comparing the results in the 30 decisions would be very useful. The results wouldn't be perfect because they wouldn't account for the 90 cases in each forum that didn't get to trial. Moreover, we can't just assume that the 10 in each forum are representative of the whole 100 in each; perhaps cases heading to court trial settle at a higher or lower rate than those going to arbitration, or perhaps different types of claims settle at varying rates. If the 10 that survive to decision are not representative of the whole 100, we'd need more information, like the number of cases withdrawn without settlement and with settlement, and the amount of settlement dollars. Still, comparisons between the forum decisions would be helpful.

But even that limited utility breaks down if the percentages of cases going to decision vary between the forums. To use the same example, imagine that 20 arbitration claims go to awards while only 5 federal and 5 state claims go to trial. (Or worse, if only 1% of the court cases go to trial.) Obviously I don't have actual figures, but we do know that very, very few employment cases actually go through trial. Overwhelmingly, those cases fall out because of settlements and withdrawals (and, in court cases, dismissals or summary judgments). It's reasonable to assume (although I don't have information at hand to back up that assumption) that a higher percentage of employment arbitration claims go to hearing, not only because of monetary factors but also because claimants can more easily do arbitrations pro se. Like many arbitrators I've had more pro se employment claims than I would have expected.

So I come back to the question that popped up: are the cases that go to decision representative of whole and similar enough cases in the other forums to make a strong comparison?

Posted by: Rick Bales | Jul 10, 2020 6:20:10 AM

Post a comment