« Infield flies and taking a knee | Main | JOTWELL: Pfander on Burbank, Plager, and Ablavsky on judicial departure »

Sunday, November 18, 2012

The Class of One at the Supreme Court II: The Trouble with Engquist

In my last post I talked about how the Court’s analysis of the class of one issue in Village of Willowbrook v. Olech prompted lower courts to start seeking ways to limit what otherwise would be the doctrine’s extremely broad scope. I explained, base on some past scholarship I’ve done, that lower courts responded to Olech by circling back to the animus requirement, despite Olech’s insistence that animus was not a necessary part of a successful class of one claim. I also noted that the Court in Engquist v. Oregon Department of Agriculture cut back on the class of one theory by making that theory inapplicable to government employee-defendants, and using an analysis that potentially stretched beyond those cases.

In my own view the Court’s analysis was somewhat flawed. The Court explained that employment decisions were what Chief Justice Roberts called “discretionary” judgments featuring “subjective” and individualized” considerations. He analogized those decisions to decisions of a police officer stationed on a freeway where many people speed and the officer can’t, as a practical matter, stop all of them. In both situations, he concluded, the subjective and individualized nature of the decision rendered them unsuitable for a class of one claim when the official chooses an unlucky target.

Two thoughts about this analysis. First, it seems to me that the employer-officer analogy is flawed. Perhaps the officer’s ticketing decision cannot be challenged on a class of one theory, but (assuming he makes a random decision who to ticket) that’s because there are literally no factors governing his decision. Thus, in a very real way, he hasn’t singled anyone out with an intent to burden them because of their identity. As I argue in this working paper, to me this appears roughly analogous to a failure to prove discriminatory intent in a standard equal protection case.

By contrast, employment decisions are at the opposite end of the spectrum. While I’m no employment law expert, it seems to me that if employment decisions are difficult candidates for class of one challenges, it’s not because the employer chose the unlucky person for no reason (as the officer did), but because he had an entire collage of reasons (work quality, absenteeism, etc.). In turn that collage makes every worker quite likely unique for purposes of her fitness for continued employment. So understood, many employment class of one claims should fail, but not because they’re analogous to the police officer example. Indeed, the examples stand at polar opposites in terms of the reasons for the official’s conduct. Instead, employment claims, if they should fail, should fail because the plaintiff will find it very hard to establish a truly similar comparator. 

So even if the Court’s analysis is a little off, doesn’t it get to the right place, by cutting off liability when lawsuits are highly unlikely to present valid equal protection claims? I don’t think so, for reasons I’ll explain in the next post.

Posted by Bill Araiza on November 18, 2012 at 02:24 PM in Constitutional thoughts | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference The Class of One at the Supreme Court II: The Trouble with Engquist:


Sam: Thanks for your comments, both on the site, and the privately-communicated ones. A couple of thoughts, which refer to your comments and to all three of my posts on the site.

1. I take your point about the traffic officer to be that there's never a situation where government makes a purely random decision. I supposed it's probably true that an officer will never just say, in response to the question "why did you stop this speeder rather than that speeder," "well, just because" or "well, I had no reason." In other words, the officer may be able to cite the ever-present "rich mix of factors" as support for any decision. But I wonder. I've heard, for example, that NYC police bag inspections in the subways are ostensibly based on numbers: every 49th person gets inspected, or something of that sort. (Maybe that's not what they really do, but I've heard that that's one of the sorting mechanisms they use.) That example seems close to Engquist's traffic officer example: there's nothing "real" about stopping every 49th speeder except that it allows the officer to make a decision -- he could have just as easily sung a song and the first person he saw after finishing would be searched. The point is, maybe some government action is in fact based on that sort of randomness.

2. If so, I'm not sure, contra your statement, that that in itself is a problem. When everyone's equally guilty and there are no real reasons to pick one person as opposed to another to ticket (or search, or whatever) I'm not sure why arbitrary choice is so problematic as a culling instrument.

3. With regard to employment, though, I completely agree with what I think your point is: that employment is different. As I say in the post, it's not like there are no relevant factors in employment decisions, thus necessitating random action: instead, I would think in most cases there are so many that no two employees are ever likely to be exactly alike. For that reason I agree that there's no reason to make government employment an "Olech-free" zone. I think Engquist is just wrong on this.

4. Finally, in your email to me you suggest that animus should not be required as an element of class of one claims, given the other obstacles that stand in the way of successful Olech claims and the difficulty of proving animus. Again, a fair point, especially from someone, like you, who actually does a lot of trial level litigation. My own fear is that those other obstacles don't stop enough such claims, especially if they're tacked on to other claims that may survive some of the hurdles you mention. I also believe, as I explained in the third post, my view that the rule against animus is a fundamental constitutional principle. Given that, if there does have to be an obstacle to the supposed floodgates of class of one litigation, I think animus is a very appropriate one -- because of its fundamental status, and also because of its proof difficulty. But that doesn't mean I favor a per se rule excluding certain types of claims (such as employment claims). I just think that animus is the better cutting tool. But I take your point that there are others that are just as effective and already operate in real-life litigation.

Posted by: Bill Araiza | Nov 21, 2012 5:29:07 PM

The most significant consideration in a "class of one" claim is te nature of the act. A ministerial act is one that is absolute, certain and imperative, involving merely the execution of a specific duty arising from fixed and designated facts. In contrast, a duty is discretionary if it involves more individual professional judgment that necessarily reflects the professional goal and factors of a situation. If the employment decision is in defiance of a mandatory policy, the act is ministerial and the claim is strong. An example would be guaranteed promotion based upon completion of specific designated tasks. If the employment decision rests upon the judgment of a superior such as promotion, transfer, and/or raise, then the act is discretionary and the claim has little chance of success

Posted by: Larry Watson | Nov 20, 2012 6:48:37 PM

The speeding ticket hypothetical is not useful, in reasoning about Olech/Engquist, I believe.

First, it is a departure from reality: I think it is very unlikely that any officer would say, "I don't pull over everybody I see who's going over 72; I just choose randomly." The officer would instead say something like, "I make a judgment based on the particular situation, including the flow of traffic and whether I have something more compelling that I need to do instead of pulling the person over." Non-arbitrary, and so no Olech problem. No need to "solve" a non-existent Olech problem by positing that "intent" means "animus" or even something close.

Second, if an officer did say, "I just choose randomly who, among all the people I see driving over 72, I pull over" - then that is problematic. No way for a government to run. If an Olech rule would deter that, then good for Olech!

Employment matters are similar. If there is a non-arbitrary reason for firing Mr. X and not Mr. Y, when both have missed the same number of days of work, then let's hear it. If it's non-arbitrary, government will win. If it's arbitrary, government should lose. No need to make a categorical "no Olech zone" around employment cases.

Posted by: Sam | Nov 19, 2012 4:28:43 PM

The comments to this entry are closed.