May 7, 2009
Ninth Circuit En Banc Finds GERA Valid under Section 5 of the Fourteenth Amendment
The Ninth Circuit sitting en banc issued Alaska v. EEOC, last week, reversing the panel decision (original decision here, petition for rehearing here, and blogged here) and finding that Congress had validly abrogated state Eleventh Amendment immunity in the Government Employee Rights Act of 1991 (Title III of the Civil Rights Act of 1991). The court found that Congress explicitly made states amenable to suit and that it had the power to do so under the Fourteenth Amendment at least insofar as these plaintiffs' claims were concerned. Judge O'Scannlain concurred in part and dissented in part, while Judges Ikuta, Tallman, and Callahan dissented. The facts of the case are relatively straightforward. Two policymaking employees of the former Alaska Governor's office sued for sexual harassment, sex and race discrimination in pay, and retaliation for opposing harassment.
This is an odd case for a large number of reasons. First, it's not the usual case where an individual seeks to sue the state in federal court. GERA gave the EEOC jurisdiction to adjudicate discrimination claims brought by policymaking state employees (and the other government employees that GERA extended antidiscrimination protection to). Nonetheless, the majority opinion assumed that the Eleventh Amendment would apply to these agency proceedings because they were sufficiently analogous to adjudication in a federal court--technically, in fact, this is adjudication in a federal court and the exercise of federal judicial power; it's just an Article I court rather than an Article III court.
Proceeding, the majority noted that there were two steps to the abrogation analysis: 1. Did Congress make clear that it intended to abrogate state immunity from suit; and 2. Did it have the power to do so under Section 5 of the Fourteenth Amendment? The majority answered the first question, "yes," explaining that GERA amended Title VII to remove the exemption of policymaking employees from the extension of protections to state employees in the 1972 amendments to Title VII. This was basically where the dissenters disagreed. GERA is its own title to the Civil Rights Act of 1991, it's codified in a place different from where Title VII is codified, it doesn't explicitly incorporate Title VII's definition of employer, which includes the state, and doesn't itself say that the state can be a defendant. It does empower the EEOC to impose most of the remedies that a court could impose under Title VII, and it does in rather passive form declare that "[a]ll personnel actions affecting . . . State employees . . . shall be made free from any discrimination . . . ." Overall, I agree that Congress intended to subject the states to actions within the EEOC, but I also agree that the provisions were somewhat inartfully phrased.
The majority also agreed with step 2, that Congress had the power to abrogate Eleventh Amendment immunity. This too has a test. In order to validly enact legislation under Section 5 of the Fourteenth Amendment, Congress must enact legislation that is congruent and proportional to constitutional harms. The Supreme Court has applied this a number of times, first looking to see what the alleged injury was, whether it violated the constitution (I think this is the congruence part), and if not, whether providing this remedy was nonetheless proportional to a documented record of constitutional violations by the states in this context.
Here is where the initial panel had started its analysis. It went straight to the legislative history and found no suggestion of a pattern of unconstitutional discrimination by the states against policymaking employees. The 1991 record was totally silent, and the legislative history from the 1972 amendments, which extended protection to state employees but excluded policymaking employees, provided no record on those policymaking employees. The majority of the en banc panel had the sounder approach, though, mimicking what the Court has done in its most recent Eleventh Amendment cases.
The majority found that the claims alleged by the plaintiffs were violations of the constitution--pay discrimination and sexual harassment are Equal Protection violations, retaliation is a First Amendment violation. Thus, enforcing GERA for these claims was perfectly congruent to what the Fourteenth Amendment provided, and no legislative record was needed to justify the legislation as it would be applied here. I won't go into the constitutional analyses--they are interesting, though, and I might follow up with a post on those alone.
Judge O'Scannlain disagreed on one of the claims. The majority had found that the retaliation claim actually stated a claim under the First Amendment. The employee was fired after she held a press conference, protesting the way that the office was handling claims of sexual harassment. Because policymaking employees do not have the same level of First Amendment protection as low level employees, Judge O'Scannlain believed that this employee would not have a valid claim. The majority rejected this reasoning because the level of position is relevant only in in the balancing portion of the Pickering/Connick test in the need to require loyalty at that level. In that balancing it is the employer's burden to prove that its interests outweigh the employee's, and so the majority accepted just the prima facie case as sufficient to state a First Amendment claim.
No one argued it this way that I could tell, but there also might be an argument that if retaliation for protesting discrimination is itself discrimination (a la CBOCS West, Inc. v. Humphries and Jackson v. Birmingham Bd. of Educ.), then the retaliation for protesting discrimination under the Equal Protection Clause would be itself discrimination under the Equal Protection clause.
UPDATE: Thanks to Will Schendel at Alaska Employment Law, who provided his link to the parties' filings on rehearing. Counsel for plaintiffs (represented in part by Sam Bagenstos (Wash U.)) did, in fact, make this argument. It's on page 12 in footnote 2 of Ward's brief, if you're interested.
In any event, it's a fascinating case, and I recommend a close read.
TrackBack URL for this entry:
Listed below are links to weblogs that reference Ninth Circuit En Banc Finds GERA Valid under Section 5 of the Fourteenth Amendment:
The appellate briefs are available here:
Posted by: Will Schendel | May 8, 2009 11:20:44 AM
A fine post, and one that helpfully clarifies the First Amendment issue for those of us not as familiar with Pickering/Connick.
My understanding is that the "congruent" part is part of the facial analysis, and has to do with how closely the statutory requirements hew to preventing or deterring constitutional violations, while the "proportional" part compares the degree of "congruence" to the seriousness of the need for a remedy (based primarily on evidence from the legislative record and secondarily on the evidence of judicial precedents). The "actual constitutional violation" step was added on by U.S. v. Georgia (2006), based on the recognition that with an actual violation there is perfect congruence. This decision reinforces the importance of U.S. v. Georgia.
However, it also leaves open the door for other courts, and even the Ninth Circuit itself, to question the facial validity of GERA’s abrogation of immunity when an allegation does not rise to the level of a constitutional violation (for example, a retaliation claim not involving public whistleblowing). It also raises more general questions about whether and to what extent new statutes that are, in form or function, amendments to existing statutes should be subjected to a separate abrogation analysis.
I wonder whether Alaska will seek certiorari; it seems to have a greater than usual interest on both winning this case and preventing it from getting to the merits stage at all.
Posted by: Harper Jean Tobin | May 8, 2009 12:44:07 PM
Harper Jean, I think you're right that "congruence" at least originally referred to the facial validity of the statute, but I think that the "actual constitutional violation" is an extension of that, stated most clearly in U.S. v. Georgia, but used as well in Tennessee v. Lane. It's really the same test in my mind. To what extent does the statutory remedy mimic what a cause of action directly under the constitution would provide--either ever or in this case? With a statute like GERA, the protected classes all get heightened scrutiny, and so it seems that the facial analysis (at least for disparate treatment) is always going to come out the same way as would the as applied analysis. Retaliation may be the weak link, but the retaliation, if framed as an equal protection or association claim connected to those protected classes, will probably also be viewed as a constitutional violation. But you're right, again, that the door may be open for courts to question the facial validity of the statute.
Posted by: Marcia McCormick | May 9, 2009 7:03:20 AM