IV. THE EEOC'S SYSTEMIC LITIGATION
The three case studies discussed above illustrate the variety of types of injunctive relief obtained by the EEOC during our study period. One could find some support for aspects of both the gladiator and collaboration theories in the Dial and McKesson cases, while PJAX seemed largely consonant with managerialist theories. But what does the EEOC's systemic docket as a whole reveal about the agency's injunctive practices? In this Part, we undertake systematic analysis of a large sample of the EEOC's systemic cases.
The first step in a systematic analysis is identifying which of the EEOC's cases are "systemic" cases. Unfortunately, during the period of our study--cases filed from October 1997 through September 2006--the EEOC did not itself clearly identify which of its cases it viewed as systemic. (211) So in order to capture the cases most likely aimed at structural reform, we used seven criteria. Any case that met any one of these criteria was screened into the set of cases we call "systemic." The first two criteria are legal theories that suggest a collective element--allegations of a pattern or practice of discrimination, (212) or a disparate impact claim. (213) The next two criteria ask whether twenty or more individuals were potentially affected by the suit. (214) The remaining three criteria focus on the breadth of the remedy obtained--namely, whether twenty or more complainants received monetary relief, whether the monetary relief totaled $ 1 million or more in real (2007) dollars, and whether the relief included an affirmative action remedy. The first two criteria show that broad relief for a workforce was likely obtained, while the presence of an affirmative action remedy again indicates a collective element to the suit.
Using these criteria, we identified a set of 281, which we refer to as the EEOC's "systemic cases." (215) Because our initial dataset was a stratified random sample of cases, we estimate that the total number of systemic cases brought by the EEOC over the ten-year period of our study was about 307, representing approximately 9% of the EEOC's litigation caseload during that period. (216) Table 1 lists the number of systemic cases in our sample by year and the percentage of those cases that satisfied each of our inclusion criteria.
Our criteria cannot precisely identify those cases and only those cases targeting systemic discrimination; nevertheless, we believe the criteria sufficiently capture the cases we are interested in--those aimed at structural reform of a targeted workplace. To the extent that the EEOC pursued structural reform in its cases, we are most likely to see evidence of it in this subset of cases.
What are these systemic cases about? Figure A reports the proportion alleging different types of discrimination. As it illustrates, the most frequent basis of suit is sex (including pregnancy) discrimination, asserted in over half the cases. Race, national origin, or color discrimination claims, grouped together as "race" in the figure, are included in over a third of the systemic docket, as are retaliation claims. Age discrimination is less commonly alleged. And as might be expected, disability and religious discrimination--claims that are more often individual, rather than collective in nature--appear more rarely (and notably less frequently in the systemic docket than the non-systemic). (217)
Figure B reports the proportion of systemic cases raising different types of employment issues. As is true of the EEOC's docket as a whole, the cases most often deal with allegations of harassment and discharge. Next most frequent, but far less common, are claims alleging failure to hire or discriminatory working conditions, pay or promotion.
With this brief summary as context, we turn now to a systematic analysis of these cases. In addition to examining features of the litigation, we look at the terms of the injunctive remedies obtained by the EEOC in these cases, whether through settlement or contested court order. (218)
Upon examination of the litigation characteristics of these cases and the type of injunctive terms obtained, we find little evidence that the EEOC's systemic cases fit the gladiator theory's depiction of structural reform litigation as hard-fought contests over liability with injunctive remedies requiring intense judicial engagement. Instead, these cases appear to involve fairly modest stakes, low-intensity litigation and, in most cases, minimal judicial oversight over decree implementation. Similarly, the cases do not match the collaborative theorists' vision of contextually- sensitive, problem-solving collaborations. The injunctive relief obtained in these suits impose a fairly standardized set of remedies, most of which are peripheral to the firms' core operations and fail to establish meaningful systems of accountability. Rather than seeking to fundamentally transform defendants' operations, the remedies imposed reflect routinized, bureaucratic solutions--the kinds of "best practices" endorsed by human resources professionals and embraced by firms as a rational (if not necessarily effective) response to anti-discrimiantion mandates.
Even when the EEOC appears to be pursuing systemic forms of discrimination, its cases were moderate in size. Tables 2 and 3 profile the EEOC's systemic docket, by year, in terms of the number of persons compensated and the monetary awards obtained.
Whether viewed in terms of the number of people benefitted or by the amount of money changing hands, the EEOC's systemic cases are fairly modest. Total damages are not tiny, but neither are these bet-the-company cases. And while these cases are clearly about more than individual grievances, they do not generally appear to entail thorough-going reform of large-scale institutions--at least as measured by the number of employees benefitted.
From this summary picture, it is difficult to know what explains the relatively modest size of these cases. It is possible that, given the changing nature of discrimination, more subtle forms of bias are less likely to generate blockbuster cases worth millions. Alternatively, private counsel specializing in employment discrimination class actions might be filing the big money class actions before the EEOC has the chance to act. The agency does have the power to intervene in privately filed employment discrimination suits, but it does not do so often, perhaps because it chooses instead to devote its resources to unrepresented parties. Whatever the explanation, the vast majority of the EEOC's cases--even those that might be characterized as systemic--are quite modest in scope.
In contrast to the early depiction of structural reform cases as hard-fought contests, the EEOC's systemic cases overwhelmingly involve low-intensity litigation. As seen in Table 4, the vast majority of the systemic cases we examined--more than 87% of the resolved cases--ended by settlement. (219) A mere handful--8% of resolved cases--ended through some sort of litigated judgment. (220)
Of course a case can be the site of very intensive litigation and nonetheless end by settlement. That is hardly ever the case in this docket, however. Most of the EEOC's systemic cases show little evidence of any rigorous contestation of liability. One hundred sixty-one, or 70.3% of the systemic cases that settled, were resolved without a single substantive motion being filed, (221) and forty-three, or 19% of settled cases, were resolved before the defendant even filed an answer. Discovery motions (222) were somewhat more common than substantive motions, as seen in Table 5. Even so, more than half the cases resolved without a discovery motion being filed. Judicial involvement in the typical cases did not appear to be particularly intense either. As seen in Table 5, the number of discovery and substantive motions actually ruled on by a judge before the settlement was quite modest across most of the cases. Only a very small handful of the settled cases appeared to entail the kind of intense, prolonged litigation battle predicted by the gladiator model. In the vast run of cases, resolution might have been preceded by a scheduling conference or two, and less commonly, a judicial ruling on a discovery motion or two.
Other measures of litigation intensity, reported in Table 6, similarly suggest that the bulk of the systemic cases entailed low-intensity litigation:
The first row of Table 6 sets out the length of the pre-resolution litigation, which is often very modest. In fact, in about 6.5% of the systemic cases, resolution is reached in the first month after filing, often with joint resolutions proposed for court approval simultaneously with the court complaint. In such a situation, the court serves as a recorder and potential enforcer of the settlement, rather than a forum for dispute resolution. More typically, the litigation lasted between one and three years. The dockets do not show particularly intense conflict during that time, however, as Table 5 shows--an average of three motions are filed (two discovery and one substantive).
In any event, resolution having been reached, the decrees that result are not the behemoths predicted by the gladiator theory. Rather, as Table 6's row 2 sets out, they tend to be fairly short--sixteen pages is the mean, and 75% have fewer than twenty pages. And their length of time is also quite short. The vast majority of them impose remedial terms for a defined period of time--a term of months specified at the outset of the decree stage. Nearly 72% of the decrees specified a term of 2 to 3 years.
Of course, litigation does not necessarily end with the entry of a judgment. (223) In structural reform cases, the implementation phase may entail vigorous contestation. And even when a time limit is specified in a civil rights injunctive case, such a limit might be extended if the...