pageok
pageok
pageok
Ricci Revisited:

Adam Liptak has an interesting article on Ricci v. DeStefano in the NYT. An excerpt:

Almost everything about the case of Ricci v. DeStefano — from the number and length of the briefs to the size of the appellate record to the exceptionally long oral argument — suggested that it would produce an important appeals court decision about how the government may use race in decisions concerning hiring and promotion.

But in the end the decision from Judge Sotomayor and two other judges was an unsigned summary order that contained a single paragraph of reasoning that simply affirmed a lower court's decision dismissing the race discrimination claim brought by Frank Ricci and 17 other white firefighters, one of them Hispanic, who had done well on the test.

Particularly notable is Liptak's discussion of what went on behind the scenes leading up to the summary order deciding the case.

The appeals court's cursory treatment suggested that the case was routine and unworthy of careful scrutiny. Yet the case turned out to be important enough to warrant review by the Supreme Court, which heard arguments in April and is likely to issue a decision this month.

The result Judge Sotomayor endorsed, many legal scholars say, is perfectly defensible. The procedure the panel used, they say, is another matter.

There is evidence that the three judges in the case agreed to use a summary order rather than a full decision in an effort to find common ground. Allies of Judge Sotomayor, who was the junior judge on the panel of the United States Court of Appeals for the Second Circuit, correctly point out that the Second Circuit often decides even significant cases with summary orders that adopt the reasoning of the lower court. They add that the panel's decision reflected a respect for precedent, though it cited none. Judge Sotomayor certainly made no suggestion at the argument that she was constrained by precedent to rule for one party or the other. . . .

In the end, according to court personnel familiar with some of the internal discussions of the case, the three judges had difficulty finding consensus, with Judge Sack the most reluctant to join a decision affirming the district court. Judge Pooler, as the presiding judge, took the leading role in fashioning the compromise. The use of a summary order, which ordinarily cannot be cited as precedent, was part of that compromise.

I find this last bit particularly interesting. The unpublished order resolved the case without creating binding precedent for the Circuit. Perhaps this was an acceptable compromise because the panel eventually concluded that a written opinion based upon the particular facts in this case could create a problematic precedent; hard facts make bad law, etc. Yet this was not the end of the matter. Once it was clear other judges on the Circuit disagreed with the panel's disposition, a per curiam opinion was published, adopting the district court's reasoning as binding circuit precedent. Therefore, the considerations that likely led the panel to issue the original unpublished order were no longer applicable. Indeed, if Liptak's account is accurate, it makes the Second Circuit panel's conduct seem worse than I had presumed. The very fact that the panel had such difficulty uniting around a single rationale for the case in the first place is, in itself, evidence that summary affirmance and adoption of the district court's rationale as Circuit precedent was inappropriate (a point Judge Cabranes stressed in his dissent from denial of rehearing en banc). No doubt this is not the last we have heard about this case and how it was handled.

UPDATE: In an essay about another case in which Judge Sotomayor joined a panel issuing an unpublished opinion, Miller v. New York, Emily Bazelon suggests why Ricci was initially resolved with an unpublished order.

The 2nd Circuit may have more than its share of unpublished opinions in hard cases for the sake of preserving unanimity. This might help explain why Sotomayor and the other two judges who heard the New Haven firefighters' claim resorted to a short opinion stripped of analysis. Perhaps in that case, too, there was a fragile consensus that Sotomayor or another judge was trying to maintain or a difference of opinion about the reasoning behind the holding in New Haven's favor.
Yet as I noted above, insofar as the Ricci panel chose to issue an unpublished order instead of publishing an opinion with precedential effect, this cannot explain (let alone justify) the panel's subsequent decision to issue a brief, per curiam opinion adopting the district court's decision as binding law for the circuit.

drunkdriver:
Indeed, if you are an affirmative action defender this is a very hard case for you to justify, almost as bad a set of facts as you could write. Liptak did a good job on this one, even sourcing it up with comments from (we assume) clerks.

15 to 20 years ago, there was a case pending before the SCT that threatened to do some damage to affirmative action; Jesse Jackson led a coalition that put together enough money to induce the plaintiffs to settle the case. That might be the smart play here for the interest groups.
6.6.2009 6:55pm
Hans:
The result Sotomayor reached was NOT defensible, since the case was dismissed before trial, and there were triable issues of fact as to the city of New Haven's motivation -- even if one believes (contrary to the Supreme Court) that local government's should have broad discretion to engage in race-conscious employment decisions.

Ricci was plainly decided wrongly by Judge Sotomayor, but whether she was merely wrong about a couple of clear-cut but narrow issues that should have kept her from summarily throwing the case out before trial, or whether she was also wrong about larger constitutional and statutory issues, is open for debate.

Based on its own past precedents, the Supreme Court could either rule outright for the white firefighters, or merely revive their lawsuit on very narrow, uncontroversial grounds that merely enable them to have their cases heard at a trial, but don’t guarantee that they’ll win at trial.

Even the Obama Administration’s brief admits Sotomayor got the case wrong on some obvious, but narrow, grounds, in throwing the white firefighters’ case out at such an early stage.

Moreover, there is an additional ground for reversal, under the Supreme Court’s decisions in Shaw v. Hunt (1996), Mount Healthy Bd. of Educ. v. Doyle (1977), and Price Waterhouse v. Hopkins (1989), regardless of how one feels about affirmative action, or “quotas.”

There was evidence — at least enough to bar dismissal of the case before trial — that the City of New Haven was motivated not by a fear of minority firefighter lawsuits, but by simple racial favoritism and political motivations, in rescinding the white firefighters’ promotions.

The presence of such a motivation would make even an otherwise permissible instance of affirmative action invalid under Shaw v. Hunt (1996), where the Supreme Court said that using race is impermissible even if it would otherwise be justified to overcome the legacy of discrimination, if the actual motive for using race was not to overcome past discrimination, but something else, like political motivations. Under the Shaw ruling, a proper remedial rationale must “actually precipitate” the racial decision, not just provide an after-the-fact rationalization for them.

Moreover, even if the City of New Haven actually had a legitimate motive, if it also had illegitimate political or racial motives, that would allow a court to rule against it at trial under the Supreme Court’s mixed motives rulings in Mount Healthy City Bd. of Educ. v. Doyle (1977) and Price Waterhouse v. Hopkins (1989).

Even cities did enjoy broad leeway to use race to offset real or perceived disparate impact, and even if the City of New Haven were likely to win at trial, Judge Sotomayor thus plainly erred in upholding the dismissal of the firefighters’ claims before any opportunity for a trial. It is clear that she jumped the gun in her ruling.
6.6.2009 6:58pm
Cato The Elder (mail):
Here is my assessment with some of the problems in the Ricci case, a repost from an earlier thread of Adler's before I saw this one on the front page.

Sotomayor found against Ricci in the case despite the fact that Ricci's laywer, Karen Le Torre, had tried to patiently explain to her - while Sotomayor was busily interrupting counsel with inane questions - that the New Haven fire department had tossed aside the exams despite the fact that on previous test taking, it had seen it fit to allow those results even though the scores on most recent test taking showed the exact same statistical distribution of scores as one they deemed acceptable years before. Why the disparate treatment of results between those years? Because as it turns out, more slots were available for promotion in that particular year, and since politicians are largely innumerate, because then the acceptable number of black firefighters was promoted (2), that outcome was politically acceptable on its face despite the statistical insignificance between the two results. Did you listen to the oral arguments that the WSJ has handily provided?

The New Haven fire department had a chance to validate this test per the contract they negotiated with the fire-fighters, a measure designed to minimize the kind of institutional corruption and favoritism decried by the NAACP, and indeed the validation study had already been paid for in advance. The board responsible for that validation chose not to proceed with the statutory protocol because the Mayor exerted significant political pressure in light of his minority political base, and instead New Haven argued in court behind "expert" testimony that they could design a "better" test - even though the company that designed the test also testified in court that they had advised the fire department to validate. This is the standard for "facially neutral" behavior in the civil service?

The conclusion I draw? Either Judge Sotomayor is incredibly dense, and thus not fit to be on the Supreme Court, or she is deliberately malicious in her disregard for non-minority defendants.
6.6.2009 7:06pm
Cato The Elder (mail):
My personal transcript, please do not use this as definitive:


SS: "If you're charged, as a decision-maker, with, with not adversely affecting an interest, a group, unless there's a business necessity, doesn't the very definition of business necessity mean that you have to look at all the alternatives and see which one best, avoids the impact?"

KT: "That is done every single time there's a test, Your Honor, that's why they hired professional testing consultants...indeed, if you understand the, the standard protocol, that was always used, but which is suspiciously abandoned in this case, and cut off, prematurely, because they didn't want to listen to it... Chapman has submitted an affidavit, saying that as part of the protocol, when you get the results, you do the technical validation report, and one, one of the things you do as part of that report, was to ensure that the test and its use for celetiary[sp?] purposes is lawful under Title VII, and you explore alternatives as part of that process."
6.6.2009 7:12pm
Just an Observer:
No doubt this is not the last we have heard about this case and how it was handled.


You won't find anyone betting against that proposition.
6.6.2009 7:38pm
Constantin:
Some of Sotomayor's statements from the oral argument transcript are staggering. The worst one is where she openly refers to the government's position by asserting, "What we're saying is . . .".

I keep reading that all of this stuff with her is being taken out of context. Invariably, though, the context actually makes her look worse.
6.6.2009 7:46pm
AF:
Title VII prohibits employment actions that have a racially disparate impact. There was no dispute that the tests in Ricci had a prima facie disparate impact.

No case has ever held that abandoning an employment action that has a prima facie disparate impact is unconstitutional.

In all likelihood, the Supreme Court is going to do that now. But it will be the first court ever to do so.
6.6.2009 8:07pm
Tony Tutins (mail):

Once it was clear other judges on the Circuit disagreed with the panel's disposition, a per curiam opinion was published, adopting the district court's reasoning as binding circuit precedent... evidence that summary affirmance and adoption of the district court's rationale as Circuit precedent was inappropriate (a point Judge Cabranes stressed in his dissent from denial of rehearing en banc).



So why didn't Judge Calabresi support rehearing en banc? Is he racist too? Or is every judge out of step but Judge Cabranes?

Considering that rehearing en banc was denied, if Judge Sotomayor acted from a racist motive, one wonders how many anti-white judges does the Second Circuit have?
6.6.2009 8:11pm
kldi:
Tony, how many of those other 2nd Circuit judges repeated the 'wise latina' philosophy in speeches.
6.6.2009 8:14pm
M N Ralph:

The result Sotomayor reached was NOT defensible, . . . . Ricci was plainly decided wrongly by Judge Sotomayor, . . . . .


I have a difficult time taking anything you write seriously when you assert that this case was so clear cut. The US District Court judge ruled against the plaintiffs, all three Court of Appeals judges on the panel ruled against the plaintiffs, and a majority of judges on the 2CA thought the case should not be reheard en banc. The Supreme Court may rule for plaintiffs, but the notion that this is a slam dunk for them is hard for me to believe.
6.6.2009 8:14pm
Desiderius:
"So why didn't Judge Calabresi support rehearing en banc? Is he racist too?"

No, he's just another judge who imagines that he can slip his thumb on her scales (perhaps to right past wrongs) without throwing Social Justice completely off her game. He, and they, underestimate the weight of that thumb.
6.6.2009 8:21pm
tailgunner (mail):
The immediate issue is not necessarily the obvious injustice that Ricci et al suffered both at the hands of the city of New Haven and the Second Circuit; but the judicial competence and inclination to judicial activism, disguised as 'empathy' and claims to race and gender superiority of Judge Sotomayor.

Who nominates someone to the Supreme Court who has been overruled on average two out of three times by that same court...once by an 8-0 unanimous ruling?

It's almost as if the worst, dumbest, dirtiest and laziest soldier in the company suddenly gets assigned to your squad.
6.6.2009 8:31pm
Volokh Groupie:



Can we stop the 2 out of 3/60%/repeated claims of Judge Sotomayor being 'overruled' by teh supreme court? The breathless washington times article which propagated that bit of information buried away the fact it was 3 out of 5 cases she was overruled in and didn't contextualize how that compares to other judges at all.
6.6.2009 8:37pm
Cato The Elder (mail):
Here are a couple of examples of some typical questions from a sample firefighter's exam that I found online:

3. NFPA 704M is the standardized system for the identification of fire hazards associated with hazardous materials. This placard is a square (broken into quadrants) that is tilted up onto a corner (diamond). What does a blue quadrant represent?

A. Reactivity
B. Specific hazard (special information)
C. Health hazards
D. Flammability

7. At 212 degrees F water expands approximately how many times its original volume?

A. 1700
B. 1200
C. 1000
D. 800
6.6.2009 8:57pm
Psalm91 (mail):
What a surprise to see Hans Von Spakovsky on this blog, but not a surprise that he takes the anti-SS position. Indeed, SS is solely responsible for this case, and the actions of the USDC and the entire Second Circuit. If she is confirmed to SCOTUS, she will no doubt sway all of the other justices to such further obviously wrong results. Every day brings a new revelation of her power and venality!
6.6.2009 9:10pm
Psalm91 (mail):
Tailgunner:

And SS is now the "dumbest, dirtiest and laziest"? Do you like her hair style?
6.6.2009 9:12pm
DangerMouse:
Some of Sotomayor's statements from the oral argument transcript are staggering. The worst one is where she openly refers to the government's position by asserting, "What we're saying is . . .".

That is indeed jaw-dropping. What the hell is she doing on the court NOW? The heck with elevating her, she should be impeached.
6.6.2009 9:29pm
Cato The Elder (mail):
In my opinion, the coverage of this case is a clear indictment of the mainstream media's failure to treat any of Obama's political statements or nominations with anything more than kid gloves. If this is really one of those "5%" cases, one of the kind Obama alluded to within his statement justifying his opposition to Chief Justice Robert's nomination, why is the blogosphere and public at large still largely mislead on the actual facts of the case that would allow one to evaluate the quality of Sotomayor's jurisprudence, in one such controversial case, for themselves? Why am I the one typing an oral transcript for myself, if the public's only source to evaluate the panel's decision to resolve the case per curiam, is but a singular audio file?

How many times has it been mentioned in any mainstream newspaper that the city refused to validate the test, and that respondents could not actually give a legally valid reason for that action; indeed, within my transcript KLT describes the city upon further interrogation by the district court as replying, "that is a question for another day". How many people are still mislead to believe that test scoring was largely based on the written portion, rather than being scored as 60% written and 40% oral? How many times has it been mentioned that the city paid over $100,000 for testing and validation services, or discussed the many different hoops they jumped through to ensure a non-biased test - just one such example are the many out-of-state minority firefighters brought in to evaluate the oral examinations.

That is the true mechanism of media bias; it's not necessarily the overt yet mostly dis-regardable bias of connotation within the text of their articles, but more mendaciously, it's more the willful advocacy against those holding legitimate issue with Sotomayor's nomination by making abstract legal theories in dispute instead of her interpretation of the more damning and concrete facts at contention.
6.6.2009 9:44pm
geokstr (mail):

AF:
Title VII prohibits employment actions that have a racially disparate impact. There was no dispute that the tests in Ricci had a prima facie disparate impact.

No case has ever held that abandoning an employment action that has a prima facie disparate impact is unconstitutional.

In all likelihood, the Supreme Court is going to do that now. But it will be the first court ever to do so.

I certainly hope so. I want the best firefighters leading the way, and I don't care if they are all transgendered lesbians from Nairobi. If they happen to be all white with one Hispanic, that's OK with me too. But then again, common sense and the law seldom seem to overlap in this day and age of political correctness. And lawyers wonder why they rank all they way up there with used car salesmen in public trust.
6.6.2009 9:48pm
devoman:
I want the best firefighters leading the way

Who wouldn't. But you assume those with the best scores automatically make the best firefighters. There are likely many other factors.
6.6.2009 10:29pm
David M. Nieporent (www):
Title VII prohibits employment actions that have a racially disparate impact.
False. Your argument goes off the rails right there. Title VII bans employment practices that have a racially disparate impact if those practices are not job-related and consistent with business necessity.
6.6.2009 10:38pm
SPO:
Take a look at Second Circuit Local Rule 32.1.
6.6.2009 10:52pm
Volokh Groupie:
Indeed--perhaps firefighters who hail from diverse backgrounds can better empathize with the fire and reason it down to only a couple sparks.

Yes there are clearly other competencies that go into being a successful firefighter, but having superior knowledge on firefighting relevant procedure, policy and science as judged by an objective test certainly seems like a good initial criteria.
6.6.2009 11:00pm
pmorem (mail):
Who wouldn't. But you assume those with the best scores automatically make the best firefighters. There are likely many other factors.

Sotomayor has your answer:
“I would hope that a wise Latina woman with the richness of her experiences would, more often than not, reach a better conclusion.”
6.6.2009 11:18pm
Allan (mail):
The city was between a rock and a hard place, as many employers are.

On the one hand, if the city accepted the tests, it would have been subject to a disparate impact claim.

On the other, if it threw out the tests, it would have been subject to a discriminatory animus claim.

So, it had to decide which claim was most likely to succeed and drive on.

It is the problem with Title VII claims. Whatever the employer does, there is a good chance at litigation, especially if the plaintiffs can find outside funding for their suits.

There needs to be some mechanism for insuring fairness for employees while making the process less expensive for employers. When someone figures it out, please let me know.
6.6.2009 11:25pm
Laurence Dale (mail):
This news parody headline sums up the logical end game of the Ricci decision: "Lawyers Advise Fire Departments to Close Until Fires Destroy More Racially Proportionate Numbers of Homes." Full article at the April 19 entry on this blog: www.optoons.blogspot.com
6.6.2009 11:26pm
Just an Observer:
Allan: On the one hand, if the city accepted the tests, it would have been subject to a disparate impact claim.

On the other, if it threw out the tests, it would have been subject to a discriminatory animus claim.

So, it had to decide which claim was most likely to succeed and drive on.

It is the problem with Title VII claims. Whatever the employer does, there is a good chance at litigation, especially if the plaintiffs can find outside funding for their suits.


If this condition is true, unless there is a constitutional problem with Title VII, then it is Congress -- not activist courts -- that are responsible for enacting a schizoid policy, not a compelling reason to oppose Sotomayor's nomination. She was part of a unanimous panel whose result was effectively ratified by a divided vote in the full circuit. Hardly evidence of an activist, out-of-the-mainstream judge.

The conventioanl-wisdom handicapping seems to expect a reversal. Does the consensus predict that when the Supreme Court rules it will reach out and strike down as unconstituional the disparate-effects provision of Title VII as applied? Or, based on the record described by the NYT article, is it more likely that there will be a narrow remand with instructions to retry the facts of the case? That would be a more clear finding of procedural "error" by the court below, but would not implicate the big issues of affirmative action that are red meat in the political narrative.
6.6.2009 11:58pm
Desiderius:
"having superior knowledge on firefighting relevant procedure, policy and science as judged by an objective test certainly seems like a good initial criteria."

Good, meet perfect. Perfect, good. Can't we all just get along?
6.7.2009 12:51am
Volokh Groupie:
considering the commenter was responding to the snippet in this line 'I want the best firefighters leading the way, and I don't care if they are all transgendered lesbians from Nairobi. If they happen to be all white with one Hispanic, that's OK with me too.' I think my parsing of the word good and the specific language I used about objectivity and the fact that it would be an initial screen would have given pause to somebody trying to twist my words into some type of absolute criteria..but i guess i was wrong
6.7.2009 1:32am
JustBeingPedantic (mail):
Just An Observer: "Or...is it more likely that there will be a narrow remand with instructions to retry the facts of the case?"

I'm not sure the case has actually been to trial. The Second Circuit was reviewing the District Court's summary judgments, no?
6.7.2009 2:35am
Cato The Elder (mail):

The city was between a rock and a hard place, as many employers are.

On the one hand, if the city accepted the tests, it would have been subject to a disparate impact claim.

On the other, if it threw out the tests, it would have been subject to a discriminatory animus claim.

So, it had to decide which claim was most likely to succeed and drive on.


False. False. 100 Times False. I'll let Karen Lee Torre respond from the oral arguments in the Ricci case.

The bolded portions are my emphasis.


KLT: So adverse impact in civil service testing is not an anomaly, it's all but guaranteed! In adverse impact, to the extent, I-I noticed that you...

RP [interrupts]: But it had never happened before in New Haven.

KLT: It happens every time in New Haven, Judge Pooler, it's never not happened!

SS: But what the board was told, and you're disputing what they were told, was that, uhm, there were less adverse impact from prior tests than from this one...

KLT: -starts to speak-

SS[interrupts]: That's what they were told.

KLT: That's what they were told, and when I submitted as evidence to show that was a flat falsehood, a flat falsehood.

KLT: Uh, and statistical errors, was one of things that if I have time I wish to talk about...

KLT: One of the reasons, it was misleading, and one of the reasons why we really shouldn't even be here is because...

KLT: In 1999, the city sought to fill 42 Lieutenant's positions, the adverse impact, and the ratios in that test were no different, no different. They had one African-American in top 16 scorers...but they were filling 42 positions; so when you have that mass vacancy...you can reach, further and further down the list.

KLT: In 2003, they were only seeking to fill 7 vacancies. But the actual statistical breakdowns, the adverse impact ratios, the passing rates for Hispanics, for African-Americans, Whites, had been consistently the same, with almost every test. In fact, in the Captain's exam for 2003, there was improvement in the lot[???] of Latinos. In the 1998 Captain's exam, the exam produced one qualified Latino for Captain. In the 2003 exam, 3 Hispanics qualified for Captain, and all 3 had been victimized. They-they weren't intended victims, but they had been victims, because when [???] get the lists, you don't know how many vacancies are going to arise. So when city officials stood before the civil service board, and said, "UH OH, you can't promote these men, they're WHITE!", which is exactly what they said! - they looked at the top scorers, they looked at what they thought they were their vacancies, and they went on record as saying, "We can't promote them, they're White".


Once again, the board that was supposed to certify the New Haven firefighters' exam rejected a test whose validity or job-relatedness they did not dispute in district court, and also one where all the relevant statistical measures of disparate impact were exactly similar to the the results of a year that they had previously certified as acceptable. That's why KLT says, "we really shouldn't even be here", because it was obvious, to her at least, that any reasonable comparison of the two different actions of the board could only reveal that the board intentionally discriminated against the plaintiffs on the basis of their race. Indeed, they were even on record as saying so. Moreover, they could give no valid alternative to the test they refused to certify, for positions they filled with "acting" lieutenants, as the remainder of the transcript makes clear.


KLT: They were mistaken; over the two years, vacancies arose, so what happened is...

RP[interrupts]: Of course, what they were thinking at the time they made the decision is what's relevant, not what happened later.

KLT: I'm perfectly happy to allow the case to turn on what they said at the time or what they knew at the time, because what they did, was they because they admitted these exams were job-related, they admitted they were valid, they stood before Judge Atherton in an oral argument that I thought...involved, every single pointed question that I was hoping Judge Atherton would ask. When I reached that oral argument, I was hoping the Court would ask certain questions which is exactly what she did.

KLT: The first thing Judge Atherton asked was, of-of the defendants, was: is your position that these tests were not job-related or invalid? And what was the answer? "Your Honor, we have no basis on which to challenge or rebut the fairness and validities of these tests, or the job-relatedness of these tests."

KLT: The next question was, all right, what's your-what alternative do you have in mind? What's the alternative? And I'd like to give you the exact quote because I think it's meaningful, meaningful, for purposes of the Supreme Court's Equal Protection jurisprudence.

KLT: "While we don't know if we'll have a different result; the truth is that the record admittedly is not vetted out as to an alternative. There would have to be studies done, to show that those alternative methods would be substantially similarly valid, but that's not what we're dealing [with] here, because we're not at that stage."

KLT: And in response to pointed questioning from Judge Atherton, on the very linchpin for disparate impact liability under Title VII? The linchpin for disparate impact liability, if the test is job-related, is to show that there is an equally valid, available, equally valid, alternative with demonstrably less adverse impact.

KLT: And at oral argument, at the district court, when Judge Atherton, said , "What is your alternative?", what did they say - "Judge Atherton, that's a question for another day."

KLT: "Both counsel, and the record, and Ms. [????] admitted that Dr. Warner [respondents' "expert"] didn't even look at look at those examinations."

RP: But they don't have to have the alternative, all they have to do is know they have, an unfair test, and there were, alternatives that wouldn't have the same unfair results.

KLT: I disagree that that's the law, Judge Pooler. You-you're...

RP[interrupts]: Did they have to have an alternative? Are you telling me that they had to have an alternative, or they just had to look at this and say, "we have disparate impact".

KLT: You have disparate impact in almost every single civil service exam in every urban area in this country, Judge Pooler. This is not an anomaly.
6.7.2009 3:28am
ReaderY:
I think the case is less clear than many on this forum appear to suggest. It's a hard case. The consequences of ruling either way could be thickets of essentially needless litigation. If we reach a point where there are prima facia grounds to sue no matter what an employer does, the courts will move closer to becoming the nation's de facto personnel department, and each employment decision will become more likely to involve a lawsuit -- moving us significantly closer to the law adding little more value than a simple tax with the money simply spent on judges and lawyers without the device of lawsuits. The sphecter is a serious one and not to be taken lightly.

That said, precisely because this was not an easy case, the 2nd Circuit, and Judge Sotomayer as a panel member, was obligated to provide an opinion with a reasoned explanation for how they ruled.
6.7.2009 3:36am
devoman:
Re: Volokh Groupie:

Yes there are clearly other competencies that go into being a successful firefighter, but having superior knowledge on firefighting relevant procedure, policy and science as judged by an objective test certainly seems like a good initial criteria.

Yes, I agree. However, I was making the point that given applicants that all surpass the minimum requirements, small differences in scores do not necessarily correlate with on-the-job performance.

I don't think anything in the Ricci case indicates that the City of New Haven was promoting incompetent, or even marginally competent, firefighters. If I am wrong about this, I am open to be corrected.
6.7.2009 8:16am
Just an Observer:
JustBeingPedantic: I'm not sure the case has actually been to trial. The Second Circuit was reviewing the District Court's summary judgments, no?

Yes, I think that's correct. I should have been more precise. As I understand it, one possible outcome now is that the justices could find that it was an error to decide the case on summary judgment, and send it back without prejudice for a full trial in District Court.
6.7.2009 8:37am
Applekeys:
The unpublished order resolved the case without creating binding precedent for the Circuit. Perhaps this was an acceptable compromise because the panel eventually concluded that a written opinion based upon the particular facts in this case could create a problematic precedent; hard facts make bad law, etc. Yet this was not the end of the matter. Once it was clear other judges on the Circuit disagreed with the panel's disposition, a per curiam opinion was published, adopting the district court's reasoning as binding circuit precedent. Therefore, the considerations that likely led the panel to issue the original unpublished order were no longer applicable. Indeed, if Liptak's account is accurate, it makes the Second Circuit panel's conduct seem worse than I had presumed. The very fact that the panel had such difficulty uniting around a single rationale for the case in the first place is, in itself, evidence that summary affirmance and adoption of the district court's rationale as Circuit precedent was inappropriate (a point Judge Cabranes stressed in his dissent from denial of rehearing en banc).


I'm confused on what your view of how judges make decisions is. It sounds like from the article that Sack had issues to some extent with the judgment but mostly with throwing out language for future cases; resolving the case with an unsigned order is a compromise in the sense that equity can be done (what remedy is requested here?) without having to issue some kind of blockbuster. If you didn't realize, these kinds of things happen all the time. You are located within the Sixth Circuit right? Try reading their unpublished opinions sometime. There's tons of these in criminal cases and discrimination cases. The panels often reach agreement where you'd be surprised to find it. (I have argued several of these cases disposed of this way.)

There are valid arguments against this procedure, but they are arguments against unsigned/published opinions, not anything this panel did. As it stands, they are a judicial tool; judges are busy; and it's not always easy to reach agreement. Maybe the entire concept of unpublished opinions is bad news (Judge Kozinski has written a defense of them) in that it harkens back to the days of Courts of Equity, but bottom line this is no worse than what goes on in any number of cases, largely because of judicial workload.

Again, I'm not saying this was perfectly defensible, but what you cannot do is pick this case out as some kind of incredible outlier.
6.7.2009 9:25am
SamW:
A little perspective is in order.

U.S. District Court Judge Janet Arterton issued a lengthy opinion in which she dismissed the firefighters reverse discrimination claims and found for the city of New Haven. Judge Arterton is no stranger to employment law cases, having represented mainly employees in private practice before getting appointed to the bench. The decision is well worth the read. Reasonable people can disagree with the outcome, but Judge Arterton's decision hardly lacks logic or thorough reasoning.

The firefighters appealed and the case went up to the Second Circuit. Judge Sotomayor as a junior judge, was one of three Second Circuit judges selected to serve on a panel to hear the case. She and two other judges decided to affirm the district court's decision, in a decision by Judge Pooler.

Judges Pooler and Sack, and Judge Sotomayor all agreed that they did not have anything to add to Judge Arternon's decision so they issued a "per curiam" opinion which, in essence, adopted the lower court's reasoning. (Summary orders are quite commonly used in the Second Circuit). W

Here was the essence of the the Second Circuit's decision:

We affirm, for the reasons stated in the thorough, thoughtful, and well-reasoned opinion of the court below. Ricci v. DeStefano, 2006 U.S. Dist. LEXIS 73277, 2006 WL 2828419 (D.Conn., Sept. 28, 2006). In this case, the Civil Service Board found itself in the unfortunate position of having no good alternatives. We are not unsympathetic to the plaintiffs’ expression of frustration. Mr. Ricci, for example, who is dyslexic, made intensive efforts that appear to have resulted in his scoring highly on one of the exams, only to have it invalidated. But it simply does not follow that he has a viable Title VII claim. To the contrary, because the Board, in refusing to validate the exams, was simply trying to fulfill its obligations under Title VII when confronted with test results that had a disproportionate racial impact, its actions were protected.

The firefighters appealed to the Second Circuit again, asking the entire court to hear the case en banc (meaning that all 13 judges would hear the case). That request was rejected.
6.7.2009 9:38am
jrose:
Cato,

The fact the city previously felt the test was valid when more candidates were to be promoted, does not necessarily imply that the city was motivated by race - rather than avoiding a Title VII lawsuit - when it concluded the current test should be discarded (although I agree the city's motivation is a factual matter that should go to trial).

As far as the law goes, we are in new waters with a preventative measure to avoid a Title VII lawsuit (assuming New Haven prevails on the above factual finding), that in turn provokes a Title VII lawsuit. It is not at all clear cut that Arterton's view of the law is wrong, and she was the one whose pointed questioning Le Torre and you think makes this case clear cut.
6.7.2009 9:39am
Stevie Miller (mail):
if Liptak's account is accurate, it makes the Second Circuit panel's conduct seem worse than I had presumed. The very fact that the panel had such difficulty uniting around a single rationale for the case in the first place is, in itself, evidence that summary affirmance and adoption of the district court's rationale as Circuit precedent was inappropriate (a point Judge Cabranes stressed in his dissent from denial of rehearing en banc).



Presuming is almost as bad as assuming, Jonathan.

But don't feel bad; you're new to this analysis stuff and a young pup at that.

Stick with the story through the hearings, and don't be so quick to fall for the "criticize her, you're a redneck racist" argument.

Never too late to admit you erred, and should have given Ricci a much, much closer look in your earlier posts beginning to assess the Sotomayor nomination.
6.7.2009 9:49am
Desiderius:
Volokh Groupie,

I was agreeing with you.

Those who support the throwing out of the test are letting the perfect be the enemy of the good.
6.7.2009 9:57am
Tony Tutins (mail):
In 1999, the city sought to fill 42 Lieutenant's positions, the adverse impact, and the ratios in that test were no different, no different. They had one African-American in top 16 scorers...but they were filling 42 positions; so when you have that mass vacancy...you can reach, further and further down the list.

Well, there's your solution right there: Wait till there are enough vacancies before you promote anyone.

Now that their own oxen have a few gore marks, conservatives are flip-flopping from their historic positions.

Conservatives advocate judicial activism this time: They want a federal court to order a government agency to promote people.

And conservatives want a Court of Appeals to make policy: Not satisfied with a disposition in this case, conservatives want an opinion with precedential effect in the Second Circuit.

(I considered but dismisse the idea that conservatives wanted the Second Circuit to rule that all employment tests shall be presumed valid if white people score sufficiently better than minorities.)
6.7.2009 10:19am
Tony Tutins (mail):
KLT: And in response to pointed questioning from Judge Atherton, on the very linchpin for disparate impact liability under Title VII?

Repeatedly getting the District Judge's name wrong might be one reason Pooler's panel might have affirmed Judge Arterton's decision.

SamW, your reasonable explanation of what happened does not take into account the conservatives' theory that Judges Calabresi, Pooler, Sack, Arterton, and others who voted against en banc rehearing, are all self-hating race traitors.
6.7.2009 10:31am
Cato The Elder (mail):


The fact the city previously felt the test was valid when more candidates were to be promoted, does not necessarily imply that the city was motivated by race - rather than avoiding a Title VII lawsuit - when it concluded the current test should be discarded (although I agree the city's motivation is a factual matter that should go to trial).



This is the most pedantic and ridiculous thing I've read in a long time. How else did the city determine the examinations had disparate impact over some well-defined threshold if all their analyses could not distinguish statistically between these unacceptable examinations from the previously acceptable examinations? Did they sniff the tests, and say, "Boy, those tests sure do smell like they do a bad job evaluating minority applicants!" Am I supposed accept this tendentious logic knowing the dynamics of urban politics in New Haven, when there was evidence and record that Mayor DeStefano had made threats towards the board members not to certify the examinations since a prominent black pastor and ally of the mayor was strongly opposed to the certification of the results?

No, I regard that kind of stupidity is deliberately evil.

In any case, the facts of the case before Sotomayor are being reported incorrectly. Many people in the media are discussing her actions in the case under the pretext that there was no way for the Judge(s) on the panel to be convinced one way or the other whether or not the tests were valid or job-related, and ALSO that there was no way to divine a malicious intent SEPARATE from the Title VII issues in the case. If there's a full-throated legal defense of the ruling, I want to see that defense, right in front of me, not mere assertions that "this is not a hard case". Everyone should be defending Sotomayor with the full facts at hand.
6.7.2009 10:56am
pluribus:
So Judge Sotomayor joins two other judges on her panel in a unanimous decision, later supported by a denial of an en banc hearing by the full circuit, to affirm a decision of the district court that affirmed a decision by the New Haven Fire Department. For this, she is condemned as an activist judge with a racist agenda. Is this thread developing as an honest effort to understand a decision in which Sotomayor reached the same conclusion as most of the other decision-makers who considered the case, or as an effort to fuel opposition to her nomination? Forgive me if I suspect the latter.
6.7.2009 11:04am
pluribus:
Cato The Elder:

This is the most pedantic and ridiculous thing I've read in a long time. . . . I regard that kind of stupidity is deliberately evil.

This kind of hyperbolic rhetoric is really counter-productive.
6.7.2009 11:09am
Tony Tutins (mail):

For this, she is condemned as an activist judge with a racist agenda.

Don't forget, that Hispanics suffered as well as whites when the fire department threw out the test results. So, along with being prejudiced in favor of Latinas, Sotomayor is a self-hating race traitor as well.

Logic and consistency would suggest that Sotomayor is a wise Negrita. Or maybe she just hates men.
6.7.2009 11:09am
Adam B. (www):
Since when do an advocate's unsworn offerings during oral argument qualify as evidence of truth of the matters asserted?

How about making the entire transcript available rather than your carefully pruned excerpts?
6.7.2009 11:10am
Cato The Elder (mail):

This kind of hyperbolic rhetoric is really counter-productive.


I don't need your unwanted advice, thank you.


Since when do an advocate's unsworn offerings during oral argument qualify as evidence of truth of the matters asserted?

How about making the entire transcript available rather than your carefully pruned excerpts?


That was all one long excerpt, nothing was cut out at all from it, it is continuous throughout its length. Moreover, I was addressing a specific question, and nothing requires to be "fair" in this forum because tedious leftists wish it to be so. But, of course, I might have the defendant's transcript ready as soon as I gather the energy to proofread my garble. Perhaps you'd like to transcribe it yourself?
6.7.2009 11:19am
SamW:
It appears Cato that you are arguing that the Board's reason was a pretext, for racial animus against the firefighters. That is, that the Board did not in fact believe that certifying these results, opened the city to a title VII claim, and its real reason was that it did not like whites.

To my mind, it is no wonder that there is reluctance to open the Board's rationle to such second guessing because of the costs to businesses of such endless litigation and the realities that a jury would not likely find that the board did not like white people.
6.7.2009 11:21am
AF:
David Nieporent: If my point is wrong, you shouldn't have any trouble citing a case where an employer violated Title VII or the Constitution by abandoning an employment test with a prima facie disparate impact.
6.7.2009 12:01pm
jrose:
This [it does not not necessarily follow New Haven was motivated by race, rather than avoiding a Title VII lawsuit, even though the city previously accepted earlier results from the same test] is the most pedantic and ridiculous thing I've read in a long time.

Then, I guess Arterton, seven memebrs of the Second Circuit, and likely 4 SCOTUS justices are pedantic and ridiculous.

How else did the city determine the examinations had disparate impact over some well-defined threshold if all their analyses could not distinguish statistically between these unacceptable examinations from the previously acceptable examinations.

By merely looking at the completely non-controversial result that the 1999 test did not produce a disparate impact and the 2003 test did. Maybe the city should have done due diligence after the 1999 test to avoid what eventually happened when fewer promotions were available, and maybe that ought to be the legal standard. Or maybe, the city is guilty of a pre-text. But, maybe not in both cases. Thus, your adamant conclusion that alternate views should be ridiculed as obviously in error is not persuasive.

Many people in the media are discussing her actions in the case under the pretext that there was no way for the Judge(s) on the panel to be convinced one way or the other whether or not the tests were valid or job-related, and ALSO that there was no way to divine a malicious intent SEPARATE from the Title VII issues in the case.

Citations for the "many people in the media" taking this view?
6.7.2009 12:16pm
pluribus:
Cato The Elder:

I don't need your unwanted advice, thank you.

Au contraire, you do need it, but you don't know it. If you can't brook criticism, then don't post on a board where others are free to post contrary views.
6.7.2009 12:27pm
jrose:
AF,

You lost me on how the lack of a case (which as of the end of June might not be true anymore) where an employer lost a lawsuit after abandoning a test that resulted in a disparate impact proves your claim that "Title VII [always] prohibits employment actions that have a racially disparate impact."
6.7.2009 12:34pm
DiverDan (mail):
Tony Tutins says -


Well, there's your solution right there: Wait till there are enough vacancies before you promote anyone.


Sounds like a great solution to me! What could possibly go wrong if a Fire Department is short by 12 Lieutenants and 6 Captains? I think it works for the Supreme Court, Too! Don't give Obama any nominations to Supreme Court Vacancies until there are 4 -- after all, it only takes 5 Justices to decide a case!
6.7.2009 12:39pm
frankcross (mail):
Cato, I don't know if you need my advice, but I would advise:

1. The assertions of an attorney for a client in a case made in support of that client, even when boldfaced, should not be taken at face value. Perhaps true, but only judged so after further investigation

2. Reference to those who disagree with you as stupid and evil is not a persuasive rhetorical tactic, even when boldfaced. Indeed, it makes you look closed minded and therefore undermines your credibility.
6.7.2009 12:39pm
AF:
You lost me on how the lack of a case (which as of the end of June might not be true anymore) where an employer lost a lawsuit after abandoning a test that resulted in a disparate impact proves your claim that "Title VII [always] prohibits employment actions that have a racially disparate impact."

That wasn't my claim, with or without your [in]accurate bracketed addition. My claim was if the Supreme Court reverses the Second Circuit, it will be an unprecedented decision.
6.7.2009 12:53pm
jrose:
AF,

Of course if SCOTUS reverses, it will be unprecedented (has there been any such case at any level before). But, you did make the assertion, "Title VII prohibits employment actions that have a racially disparate impact."

Not sure why you made that assertion to back-up your point about Ricci being unprecedented.
6.7.2009 12:59pm
drunkdriver:
pluribus wrote: So Judge Sotomayor joins two other judges on her panel in a unanimous decision, later supported by a denial of an en banc hearing by the full circuit, to affirm a decision of the district court that affirmed a decision by the New Haven Fire Department. For this, she is condemned as an activist judge with a racist agenda. Is this thread developing as an honest effort to understand a decision in which Sotomayor reached the same conclusion as most of the other decision-makers who considered the case, or as an effort to fuel opposition to her nomination? Forgive me if I suspect the latter.

While it's a respectable position to think Ricci should lose, I have a problem with how the panel handled this case, and share Judge Cabranes's suspicions that the panel was trying to bury a difficult-to-defend decision, and insulate it from Supreme Court review.
6.7.2009 1:14pm
AF:
jrose,

I'm not sure what you're confused about.

Title VII does prohibit employment actions that have a racially disparate impact -- unless a court finds that they are justified by business necessity and there is no alternative action with a lesser impact.

In order to comply with this law, employers are well-advised (as a practical necessity and according to EEOC guidelines) to consider the prima facie disparate impact of their employment actions before taking them.

While employers are entitled to move forward with employment actions that have a prima facie disparate impact if they can prove business necessity and there is no alternative, no court has every held that they are obligated to do so. Indeed, given the difficulty of predicting what courts will conclude as to business necessity and alternatives, the most practical solution is often to scrap the proposed practice and start again.

Until Ricci, it was pretty clear that this practical solution was perfectly legal.

The bottom line point is that Court of Appeals judges were under no obligation to anticipate that the Supreme Court would establish new precedent. Under existing precedent, the decision was straightforwardly correct.
6.7.2009 1:21pm
jrose:
Title VII does prohibit employment actions that have a racially disparate impact -- unless a court finds that they are justified by business necessity and there is no alternative action with a lesser impact.

That's what you should have said. Nieporent and I read it as a more sweeping statement.

Under existing precedent, the decision was straightforwardly correct.

I disagree for two reasons:

1) There is the factual matter of whether New Haven's purported motivation to avoid a Title VII lawsuit was a pre-text for a racial classification.

2) Assuming New Haven's motives were permissible, what precedent demands there is a straightforward conclusion? I thought preventative actions, particularly after a test is completed, was new territory.
6.7.2009 1:29pm
AF:
Nieporent and you need to work on your reading comprehension. If you thought I was saying disparate impact is "always" illegal, what did you think I meant by a "prima facie disparate impact"?
6.7.2009 1:33pm
Cato The Elder (mail):
It's clear that both the Obama DOJ and the Second Circuit Court of Appeals did not want the Ricci case to create a binding precedent, which is why the amicus brief filed by the former calls for the Second's judgment to be vacated by virtue of the panel's substantial procedural errors and the case to be remanded for trial in a lower court.

From the amicus brief, my bolded emphasis:

Neither the district court nor the court of appeals, however, adequately considered whether, viewing the evidence in the light most favorable to petitioners, a genuine issue of material fact remained whether respondents’ claimed purpose to comply with Title VII was a pretext for intentional racial
discrimination in violation of Title VII or the Equal Protection Clause.

6.7.2009 1:45pm
A. Zarkov (mail):
Of course this case couldn't come to trial. Then they would have to have actually looked at the facts, and that would have benefited Ricci.

Let's take a quick look at the Captain's test results, which you can find here. The following shows the rank order of the test scores by race. Ranking decreases from left to right. Twenty five white, 8 Hispanic, and 8 black applicants took the test. The test score is a weighted average of the oral and written components of the exam.

WWWWWHWHWWWWHWWBWWBWWBWWWWWHWBHWHWBHBHWBB

New Haven uses "the rule of 3" to fill captain vacancies. The first position gets filled from the top three and so on. Thus the 7 captain vacancies would be filled from the applicants with the 10 top scores. These--

WWWWWHWHWW

The top 10 scores came from 8 white and 2 Hispanics. City officials described this outcome as "shocking" and refused to certify the test. But is this outcome shocking? Not really. To understand why we must do a little analysis.

First we need to realize that the exam is not an aptitude test like the LSAT. It tests acquired knowledge pertinent to the job. A fire captain needs the skills, knowledge and abilities to make important decisions in fighting a fire because a wrong decision can cause death. Thus it's perfectly reasonable to test the applicants' level of pertinent knowledge and this must include some book learning. To do otherwise risks the lives of both firemen and the public. Since the test is not an aptitude test, hard and prolonged study will increase the applicant's score. Indeed one of the top scorers is dyslectic and had to put in a great deal of work. Thus to some degree the top 10 applicants simply studied more than others.

To see if the test outcome is "shocking" let's look at an important bounding case. I think most will agree that the most neutral selection procedure would be a lottery. Write each applicants name on a ball, throw them into a bin, mix and withdraw 10. This gives everyone an absolutely equal chance. If all the applicants had the same native ability and studied equally hard, then the lottery would approximate the exam.

Could a lottery produce "shocking" results? Yes. Obviously in this context, "shocking" means too many white or not enough black applicants are in the top 10. But how many is "too many?" The following are possible answers. Remember 61% of the test takers were white.

1. Any more than 6 white applicants. (proportional representation)
2. Eight of more white applicants. (Any more than what has occurred.)
3. No black applicants. (The actual outcome.)
4. One or fewer black applicants

Using some statistical theory we can calculate the chances of cases 1-4 are:

1. 20.1%

2. 1.2%

3. 8.3%

4. 35.8%

We see that only case 2 would be a rare event, and might qualify as "shocking." Most likely it's case 3 that caused the city to reject the test: no blacks. But we see that even under this most neutral system, (3) occurs frequently enough to be considered unexceptional. It seems to me that it's hard to argue that the results alone establish that the test is unfair.

New Haven's rejection of the test means minorities could game the system by deliberately scoring low so as to decertify the test. After all at least one of the test takers who actually failed is now an interim captain.
6.7.2009 2:37pm
Cato The Elder (mail):
A. Zarkov,

It seems to me that what some are arguing here is that the New Haven fire department needn't even use statistics at all to claim the fear of suit based on a Title VII claim! Apparently, to some, if a test is given to groups X and Y, a municipal fire department can simply note that for any threshold of X <= 0.8Y who score over the cutoff, then, without any statistical evidence or theory whatsoever, this board can repeatedly throw out these tests ad infinitum and those actions do not constitute a de facto quota or any sort of racial discrimination. New Haven could plausibly justify their actions in the Ricci case with theory from Marx's Das Kapital, perhaps saying that a discriminating test constitutes class exploitation and division of the working class. By the panel's interpretation, that facile theory provides legal protection from suit under an Equal Protection claim from the afflicted.
6.7.2009 3:05pm
troll_dc2 (mail):
There are some interesting issues lurking under the discussion of Ricci that are getting no air at all, and I wish I knew why. They include:

1. Do so-called reverse-discrimination plaintiffs have standing to challenge an employer's decision not to adopt a selection procedure that has produced an apparent racially disparate impact and that the employer reasonably could believe would be found illegal?

2. To obtain relief, do the plaintiffs have to establish that the employer would have prevailed if a lawsuit had been filed against it and if the employer had opposed the lawsuit? (This would resemble the trial within a trial that occurs in a legal malpractice case.)

3. If the answer to No. 2 is a yes, does Title VII really mean that an employer can be liable for not behaving in such a way that a lawsuit challenging the selection procedure will be brought against it?
6.7.2009 3:11pm
Tony Tutins (mail):

It tests acquired knowledge pertinent to the job.

From the respondents' brief, the exam also tests knowledge irrelevant or actually contrary to department procedures. Thus test performance corresponds poorly to mastery of the relevant job knowledge. Further, mastery of wrong and irrelevant knowledge suggests that the knowledge was acquired through cramming and rote memorization, not experience. Such knowledge allows a well-rested candidate to fill in the correct bubbles in a quiet hall, but can seldom be put to actual use during the stress of fighting actual fires. Further, what is rapidly memorized is often rapidly forgotten.

b. A number of firefighters testified that the exams
tested material that was irrelevant in New Haven
or contrary to New Haven firefighting policies. JA44-
48; JA67-70. The source material also contained contradictory information on the same subject matter
(CAJA786)—a problem that had been acknowledged
previously by IOS (JA19-21). Other firefighters (including
three petitioners), however, defended the tests,
contending that they should be certified. CAJA772-
773, 784-789, 1139-1142, 1145-1148.
6.7.2009 3:45pm
Tony Tutins (mail):

It seems to me that what some are arguing here is that the New Haven fire department needn't even use statistics at all to claim the fear of suit based on a Title VII claim!

Yes. There is no Statisticians' Full Employment Act. Likewise 12(b)(6) does not require plaintiffs to perform a statistical analysis before filing suit.
6.7.2009 3:50pm
ShelbyC:

Yes. There is no Statisticians' Full Employment Act. Likewise 12(b)(6) does not require plaintiffs to perform a statistical analysis before filing suit.


IANAL, so I don't know what 12(b)(6) is, (actually isn't it the requirement that a claim be stated), but don't you have to do a statistical ananysis before you can claim a "disparate impact"?
6.7.2009 4:11pm
A. Zarkov (mail):
Tony Tutins:

"From the respondents' brief, the exam also tests knowledge irrelevant or actually contrary to department procedures. Thus test performance corresponds poorly to mastery of the relevant job knowledge."

Then the city should ask the test designer for its money back. Moreover if the test is defective then then how come no one brought this up before? Didn't anyone review the test to make sure it related to the job before they used it?

I don't know if respondent's brief is the best source of information about the test. But my argument does not require that the test be relevant. I showed that even a lottery system (no test at all) could and would produce outcomes that the city would deem evidence of a defective test.
6.7.2009 4:22pm
klp85 (mail):

[D]on't you have to do a statistical ananysis before you can claim a "disparate impact"?

For the purposes of a disparate impact claim, there's what's called the four-fifths rule:

Specifically, the EEOC “four-fifths rule” provides that a selection tool that yields “[a] selection rate for any race, sex, or ethnic group which is less than four-fifths (4/5) (or eighty percent) of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact, while a greater than four-fifths rate will generally not be regarded by Federal enforcement agencies as evidence of adverse impact.” 29 C.F.R. § 1607.4(D).

Here, the evidence shows that on the 2003 Lieutenant's exam the pass rate for whites was 60.5%, for African-Americans 31.6% and Hispanics 20%. The four-fifths score would be 48%. In other words, African-Americans had a pass rate that was about half the pass rate for Caucasians, yielding an adverse impact ratio (“AIR”) of 0.59, significantly below the AIR of 0.80 that is presumed to not evidence adverse impact under the EEOC Guidelines. See Pl. L.R. 56(a) Stmt. ¶ 246; Def. L.R. 56(a) Stmt. ¶ 246. While the parties dispute the Captain's exam pass rate for African-Americans and Hispanics (see supra note 7), the pass rate for Caucasians was 88%, which is more than double that of minorities and thus by either party's statistic an AIR far below the four-fifths guideline is yielded.

Ricci v. DeStefano, 554 F.Supp.2d 142, 153-4 (D. Conn. 2006).

Since the pass rate of Black and Hispanic firefighters was less than 80% that of Whites, under EEOC rule there would have been a prima facie case of disparate (adverse) impact.

I have no idea whether the 80% rule is statistically sound, but according to the district court, that's the rule that the City confronted when the results came in.
6.7.2009 4:34pm
ShelbyC:

...there's what's called the four-fifths rule:


Interesting. I must have skimmed that part :-)

Unfortunate name, though.
6.7.2009 4:49pm
Sol:
A. Zarkov: Your argument appears to grant any fire fighter who takes the test and scores high enough a right to the job. If that's true, there is no need for the certfication procedure at all. More importantly, this is a heavy government intrusion into the hiring process by the creation of a new (property?) right in the test taker. I don't see any warrant in contract or law for the creation of such a right by the courts.
6.7.2009 5:06pm
A. Zarkov (mail):
klp85:

"Since the pass rate of Black and Hispanic firefighters was less than 80% that of Whites,..."

How is "pass rate" defined. To "pass" does an applicant need to reach a previously defined score or does "pass" mean to get into the top rankings?

The "four-fifths" rule sounds illiterate. As I have shown random fluctuations can cause outcomes to violate proportionality.
6.7.2009 5:07pm
A. Zarkov (mail):
Sol:

"Your argument appears to grant any fire fighter who takes the test and scores high enough a right to the job."

You need to read me more carefully. I am not proposing a lottery. I use the lottery example to show that even a neutral test could cause outcomes that could get attacked as disproportionate. In other words, New Haven needs to show why the test was biased against blacks beyond that the results of that particular test. For example-- provide a question that's biased.
6.7.2009 5:12pm
Desiderius:
ShelbyC,

"Unfortunate name, though."

Unfortunately accurate in what it says about the real game being played here, if that's what you mean.

Are there provisions in Title VII for applicant blood tests to confirm that they pass the one-drop rule as well?
6.7.2009 5:13pm
ShelbyC:

The "four-fifths" rule sounds illiterate. As I have shown random fluctuations can cause outcomes to violate proportionality.


Isn't the four-fifths rule just alpha=.2?

Assuming, of course, that the underlying skills are evenly distributed, which is probably a big assumption.
6.7.2009 5:14pm
Cato The Elder (mail):
It's interesting to note that even those who support the Second Circuit panel's judgment in this case agree that Title VII is perverse law and needs to be struck down, possibly on Equal Protection claims or not, which is directly contradictory to many of the amici briefs filed for the respondents in the Supreme Court's hearing of the case.
6.7.2009 5:24pm
ShelbyC:

Isn't the four-fifths rule just alpha=.2?


Nevermind, just thought it through. stats was awhile ago.
6.7.2009 5:26pm
jrose:
I use the lottery example to show that even a neutral test could cause outcomes that could get attacked as disproportionate. In other words, New Haven needs to show why the test was biased against blacks beyond that the results of that particular test. For example-- provide a question that's biased.

Is that the same standard a black firefighter would face in suing New Haven?
6.7.2009 5:36pm
jrose:
It's interesting to note that even those who support the Second Circuit panel's judgment in this case agree that Title VII is perverse law and needs to be struck down, possibly on Equal Protection claims or not

Citation?
6.7.2009 5:38pm
Cato The Elder (mail):

Citation?

I'm not going to do this again. Lot's of people have defended the city by creation of a so-called "Catch-22" situation with regards to Title VII claims.

It is the problem with Title VII claims. Whatever the employer does, there is a good chance at litigation, especially if the plaintiffs can find outside funding for their suits.

No case has ever held that abandoning an employment action that has a prima facie disparate impact is unconstitutional.

The bottom line point is that Court of Appeals judges were under no obligation to anticipate that the Supreme Court would establish new precedent. Under existing precedent, the decision was straightforwardly correct.

6.7.2009 5:45pm
troll_dc2 (mail):

It's interesting to note that even those who support the Second Circuit panel's judgment in this case agree that Title VII is perverse law and needs to be struck down, possibly on Equal Protection claims or not, which is directly contradictory to many of the amici briefs filed for the respondents in the Supreme Court's hearing of the case.



I'm not sure where I stand on the Second Circuit decision. I may not know until after I see what the Supreme Court does to it.

But the "problem" with Title VII is not that it is perverse law. Rather, there are at least two problems. The first, about which I could go on at great length but will not, is that the law allows anyone to go into court to claim anything instead of providing for a mandatory winnowing of claims. The EEOC's general counselor, unlike the NLRB's general counsel, has no authority to decide which complaints may proceed, and the Commission does not have the authority to issue binding decisions (in the non-federal sector), unlike the NLRB.

The second problem is that the issues like the methods of proof are still not entirely settled. Title VII has been around for well over 40 years, and we are STILL grappling with elementary questions.

The whole subject of discrimination law is fraught with uncertainty, as so much depends on which facts are considered controlling, and the employer cannot always know in advance.
6.7.2009 6:14pm
AF:
Cato -- Two of the quotes you cite are mine. I agree with you to the extent you are saying that New Haven's alleged constitutional violations are inherent to Title VII itself.

We part ways on the solution. You seem to think Title VII should be struck down. I think that the constitutional theories that lead to that conclusion are misguided and should be abandoned.
6.7.2009 6:26pm
jrose:
Cato,

Belief that there is a Catch-22 does not imply a belief that Title VII should be abandoned. To the contrary, the Second Circuit's ruling resolves the Catch-22 by favoring preventative action.
6.7.2009 6:29pm
jrose:
AF,

Maybe the Court will reverse without concluding, using existing theory, that abandoning an employment action that has a prima facie disparate impact is presumptively unconstitutional.
6.7.2009 6:43pm
Phil Josen (mail):
If most of you are law professors you are seriously undeworked to have the time to engage in this unproductive posts. Try working for a living :)
6.7.2009 6:55pm
Tony Tutins (mail):

Didn't anyone review the test to make sure it related to the job before they used it?

No one in the NHFD reviewed this most recent test, for fear that the answers would leak out. As I recall one active and one retired FD officer reviewed the exam. They were both from outside Connecticut, so small wonder they missed any NHFD-specific questions.

To me, this is part of a growing list of reasons that persuaded management that this test was flawed. If management reasonably believes their test was flawed, they should be free to throw it out, without having to perform a statistical analysis.
6.7.2009 7:25pm
A. Zarkov (mail):
The following article investigates Ricci case from a statistical standpoint. In particular note:
The opinion relied on the government’s four-fifths rule to compare the pass rates even though the government guidelines specifically state that when the sample sizes are small, differences in selection rates that fail the “rule” may not constitute an adverse impact when they are not statistically significant.
Thus New Haven did really have to worry about the four-fifths rule since it did not apply to this small sample outcome. Of course they might have had a lawsuit anyway, but the possibility of a lawsuit over something is always present.

Also note:
Statistical tests confirm that the differences among the pass rates of the three groups on the lieutenant exam were statistically significant but not on the captain exam.

... it will be seen that for the sample sizes involved in either exam, more than 60% of the time a fair test would be classified as having disparate impact if the four-fifths rule is applied to either of the two pass rates of the three race-ethnic groups.
I did not analyze the four-fifths rule, but this article in is using similar reasoning. Because of the small sample sizes, even a absolutely fair test would fail. This is why the law makes an exception for small samples, and why New Haven had no basis to refuse to certify the test.
6.7.2009 7:46pm
Tony Tutins (mail):

Because of the small sample sizes, even a absolutely fair test would fail. This is why the law makes an exception for small samples, and why New Haven had no basis to refuse to certify the test.

It's been a while since I did hypothesis testing, but I'm fairly certain that the risk of rejecting a good test is not the same as the risk of accepting a bad test. An employer reasonably wants to prevent accepting bad tests.
6.7.2009 9:00pm
Hans:
Yet another reason why Sotomayor was wrong: Small sample sizes defeat disparate impact claims, as a matter of law. See, e.g., Coe v. Yellow Freight (10th Cir. 1981).

Of course, Sotomayor was also wrong for several other reasons, such as the fact that the case was dismissed long before trial, despite triable issues of fact as to the City of New Haven's motivation (such as whether it even acted as it did based on a fear of a real or imaginary disparate impact claim).
6.7.2009 9:29pm
Repeal 16-17 (mail):
Let's put Judge Sotomayor aside for a moment. How many want (I'm asking for opinions, not prediction) the Supreme Court to affirm the Second Circuit opinion? How many want a reversal/vacating of that opinion?
6.7.2009 9:43pm
Desiderius:
Repeal 16-17,

"How many want a reversal/vacating of that opinion?"

Judging by the legal arguments advanced here, I risk being hoisted on my own "judicial activism" petard, in that it appears that it is the law (Title VII) that needs changing, and that judicial action with the aim of ameliorating the bad effects of that law could serve to impede that change.
6.7.2009 10:27pm
A. Zarkov (mail):
"... but I'm fairly certain that the risk of rejecting a good test is not the same as the risk of accepting a bad test. An employer reasonably wants to prevent accepting bad tests."

I think you are talking about a Type II error (accepting the null hypothesis when it's false) versus a Type I error (rejecting the null hypothesis when it's true. You are correct in that the error probabilities can differ. But I can't believe New Haven did a power calculation (1 - prob[type II]) and made a policy trade off between error rates. Obviously rejecting all tests drives the Type II error rate to zero, but that's hardly an option.

In any case, the power calculation is something one does before the experimental outcome is observed. After the data is in, you reject or don't reject the null hypothesis depending on the desired p-value (usually less than 5%). In other words, the Type II does not come into consideration once the data is observed. This is just one of the funny things about Neyman-Pearson hypothesis testing. I generally prefer the Bayesian approach.

On the captain's exam, rejecting the null hypothesis (no difference in pass rates) would have implied a p-level greater than 5%, which does not meet the generally accepted legal standard.

I suspect I've thoroughly confused you. The reasoning behind this stuff is fairly convoluted.
6.7.2009 10:45pm
Repeal 16-17 (mail):
I suspect I've thoroughly confused you. The reasoning behind this stuff is fairly convoluted.


That's the only part of what you said that I did understand. :)
6.7.2009 10:48pm
A. Zarkov (mail):
Interesting except from the New Haven Register as reported here.

Nov. 6, 2004: The New Haven Register reported:

"The state's anti-discrimination commission has thrown out complaints made by about 21 city firefighters who contend they weren't promoted because they are white. … The Commission on Human Rights and Opportunities [CHRO] recently ruled in favor of the city and dismissed the complaints, filed in June after the city threw out two promotional tests because too few minorities scored well. … Attorney Karen Torre, who represents the firefighters, said she was not surprised by the decision. She filed with the CHRO primarily because the government requires it before filing a civil rights action.

"She said she's filed a number of complaints with CHRO and lost every one, including one for two white New Haven police officers who were not promoted; they went on to win more than $800,000 from a federal jury.

"Torre said her experience has been that the CHRO, the state agency that hears discrimination cases, is comprised mainly of affirmative action advocates."
6.7.2009 10:50pm
Desiderius:
Zark,

"Obviously rejecting all tests drives the Type II error rate to zero, but that's hardly an option."

Indeed, it's becoming mandatory.
6.7.2009 11:18pm
Volokh Groupie:
@Devoman

fair enough regarding the selection criteria---i also can't comment on the spectrum of qualifications of those specifically at issue in Ricci because I definitely haven't kept up with the case well enough


@Desiderius

Yeah, when I read the post again today I realized that---unfortunately you got a response from dumb VG---which is pretty much just the regular VG with some beers in him (not to be confused with the surprisingly insightful drunk VG).
6.8.2009 12:03am
Tony Tutins (mail):

Small sample sizes defeat disparate impact claims, as a matter of law. See, e.g., Coe v. Yellow Freight (10th Cir. 1981).

Counselor, can you help me understand how you extracted this statisticolegal rule from Coe? The closest ruling I can find in Coe is

A claim of discrimination resulting from the mode of filling a single position does not give rise to a disparate impact.


But I can't believe New Haven did a power calculation (1 - prob[type II])

The lack of power calculation has never stopped anyone from doing statistical hypothesis testing. All you need is a chi square table.
6.8.2009 12:08am
A. Zarkov (mail):
"The lack of power calculation has never stopped anyone from doing statistical hypothesis testing. All you need is a chi square table."

If you use a test with little power, then failure to reject the null hypothesis indicates nothing. Let's put it in radar terms. With no power you never detect the target. So does it matter if you rarely call noise the target?

Sometimes the Chi-square table is useful, but its range of application is hardly useful.
6.8.2009 12:41am
A. Zarkov (mail):
"Counselor, can you help me understand how you extracted this statisticolegal rule from Coe?"

I read it too, and I don't see it either. But you don't need a case because small samples are excluded from the four-fifths rule.

I think four-fifths rule assumes very large sample sizes so certain approximations follow.
6.8.2009 12:46am
A. Zarkov (mail):
Look at Questions and Answers on the Uniform Guidelines on Employee Selection Procedures-- here. In particular at Q&As 20-25.

21. Q. Is evidence of adverse
impact sufficient to warrant a
validity study or an
enforcement action where the
numbers involved are so small
that it is more likely than not
that the difference could have
occurred by chance?

A. No. If the numbers of
persons and the difference in
selection rates are so small that it
is likely that the difference could
have occurred by chance, the
Federal agencies will not assume
the existence of adverse impact
,
in the absence of other evidence.
It seems to me that New Haven was ignoring the guidelines, and had other reasons for not certifying the test results, than fear of an adverse impact lawsuit. What's the evidence for adverse impact? I can see none, and it does appear New Haven had an anti-white agenda. This does not look good for Sotomayor, who at least on this case seems incompetent or biased.
6.8.2009 1:12am
Tony Tutins (mail):

I think four-fifths rule assumes very large sample sizes so certain approximations follow.

The sample size here was 118, which is enormous for personnel work. Are you going to interview 118 people? Blacks alone numbered twenty-seven.

Have you done any practical industrial work? Comparing the fraction of whites who passed the exam to blacks who passed the exam will yield robust results.
6.8.2009 1:43am
A. Zarkov (mail):
"The sample size here was 118, which is enormous for personnel work."

There are two exams with sample sizes 41 and 77. These have been treated separately as they are separate tests. In the captain's test there were only 8 blacks. These are not large samples. You don't have to take my word for it read the article I linked to for details. They reach the same conclusion about the captain's exam.

"Have you done any practical industrial work?"


This is a statistical matter. I'm not sure what you mean by "industrial work," but I think the answer is "yes." I have done much industrial work.

"Comparing the fraction of whites who passed the exam to blacks who passed the exam will yield robust results."


No. From the article:

For each test the specific overall pass rate was calculated (34/77=.4416 for lieutenant
and 22/41=.5367 for captain) and the corresponding cut-off value of a standard normal
distribution was obtained .35 For the lieutenant exam, samples of 43, 19 and 15 were
selected from the same standard normal distribution and the pass rates for each group
calculated. Similarly, the captain exam scores for applicants of each group were
simulated from samples of 25, 8 and 8. Ten thousand simulations were carried out for
each exam. Table 2 reports the fraction of times the “four-fifths” rule was violated by a
fair test. Column A refers to the first criterion for passing listed above (having a score of
at least 70). Notice that fair tests would fail to satisfy the “four-fifths” rule 70% (87%) of
the time for the lieutenant (captain)
position; so, it is almost inevitable that one of the
tests would be deemed to have a disparate impact
The "4/5 rule" does not apply for this data. The samples are too small.

You are just making statements with no analysis or references to back them up.
6.8.2009 3:49am
Tony Tutins (mail):

You are just making statements with no analysis or references to back them up.

The glaring error that leaped out at me was the authors' use of a normal distribution to model a process with only two outcomes: pass or fail. Such an approximation would only be valid for extremely large sample sizes. Because the authors are arguing the sample sizes are too small, their choice of the normal is suspicious at best.

Their conclusion is manifestly suspect as well. If the sample sizes are too small in this case, they will almost always be too small, making the 4/5 rule useless and irrelevant to proving disparate impact. Such a rule will always have some meaning at the cost of lower statistical confidence.

Then, the authors do not justify or even mention their decision to separate the results of the captain's and lieutenant's exams. Surely both exams test the same body of knowledge for leading and managing firefighters. What are the differences, and what percentage of the exam do they comprise?

Further, the authors took the wrong approach entirely. They are arguing that the test is fair. Putting themselves in the place of the Fire Department trying to justify a personnel selection test, the proper approach is to prove the test is not unfair -- a consumer's risk approach rather than a producer's risk.

The best solution would to have a real statistician -- with none of the authors' axes to grind -- do an independent analysis. This article reeks of trying to justify a predetermined outcome -- for me the mixture of politics and statistics reads like Bellesiles Arming America.
6.8.2009 11:49am
Cato The Elder (mail):

Surely both exams test the same body of knowledge for leading and managing firefighters. What are the differences, and what percentage of the exam do they comprise?



WTF are you talking about? They could load on different skills and knowledge sets, how do you reach this conclusion from any evidence at all? They are *different* jobs with different responsibilities, you know. What you're basically saying, is, they didn't make this simplifying assumption that would make allow them get the correct result I want, and the burden is on the authors to show that the Four-Fifths rule is not a good criterion to ascertain bias.
6.8.2009 1:19pm
Brett Bellmore:
Doesn't this get to the original point: That Ricci should have gone to trial, because who would win was dependent on the facts, and you're only supposed to throw the case out like that if the plaintifs would lose even if you assumed all the facts went their way?
6.8.2009 1:23pm
b-rob (mail):
One thing is clear from the commentary here: the people who think Ricci was wrongly decided know nothing about employment law.

Step back a second and let's look at what happened. A test was offered by a city with a history of losing race discrimination cases. The test was part of the officer selection process, but it was not the only criteria. The test produced a distinct disparate outcome. The city threw out the test and started over.

The Ricci plaintiff's had to prove that they were discriminated against because of their race and they were damaged by that discrimination. To defeat the claim, the city could establish that they had a legitimate non-discriminatory reason for making the decision.

In order for the Supreme Court to overturn Ricci, they will have to create a rule whereby an employer, in essence, can be sued for intentional discrimination if it rejects its own suspicious hiring criteria. In other words, if an employer gets wind that a hiring practice might have a disparate impact on a particular group, the employer must ignore that, lest it "discriminate" against the successful dominant group.

Exactly what, pray tell, would the conservatives aping the Ricci plaintiffs' line have the employer do? Apply the disparate test scores, make employment decisions based on what they acknowledge is disparate, and HOPE they can prevail in the subsequent litigation by the Hispanic and Black firefighters? And what if those Black and Hispanic firefighters point to the very internal wranglings over the disparity? What then, conservatives? Those minority firefighters can already point to past discrimination and support their argument that this was just another discriminatory hiring practice. The failure to rectify that disparity even though you acknowledge it exists . . . how do you think that will play out to a jury? And at what point does application of a disparate policy, in the face of acknowledged disparity, start to look like intentional discrimination . . . since you knew there was a problem with the test and you used it anyway?

Sotomayor made the entirely appropriate comment that courts of appeal make policy. If you want to give employers a disincentive for trying to eliminate their own discriminatory hiring practices without getting sued first, overturning Ricci is the way to go. Talk about full employment for lawyers!
6.8.2009 1:44pm
b-rob (mail):
And another thing: it really does not matter whether test actually created an actionable disparate impact. Ricci was not suing alleging that the test did not create a disparate impact. Rather, they had to prove an entitlement (a promotion) was taken away because of their race. For a number or reasons, they would not have won at trial because the test was not determinative. And second, the employer does not have to prove that there is a disparity; they only have to prove that their reason for rejecting the test had nothing to do with the Ricci plaintiffs' race. Several hundreds of thousands (if not millions) of dollars at risk defending a suspect test (even if it does not have an actionable disparate impact) is a pretty friggin good reason not to use those test scores. Which explains why the employers groups are sweating this case, big-time. A vote for the plaintiffs opens Pandora's box.

Lastly, please stop with the "reverse discrimination" b.s. There is no such thing. Either there is race discrimination, or there isn't.
6.8.2009 1:59pm
Cato The Elder (mail):

The test produced a distinct disparate outcome.

False. Everything you write sadly crumbles from there. It did not produce a "distinct" disparate outcome, because the city used no statistical measure that could have actually distinguished between 2003 and prior years. Repeat one more time after me: "Your Honor, we have no basis on which to challenge or rebut the fairness and validities of these tests, or the job-relatedness of these tests." That's because New Haven was motivated by the base political considerations of urban racial politics, and since coincidentally politicians are innumerate, their actions actually precluded the handy defense they could have found in the language of Title VII. Indeed, Obama's DOJ actually notes in their amici brief that both the district court and the court of appeals failed to rigorously examine these triable questions. Oh, and that history of losing race discrimination cases? Yes indeed, you're right that the New Haven Fire Department was forced to settle with WHITE fire fighters for $800,000 a few years back. It seems that Karen Lee Torre has been at this business for quite a long time, until finally the Ricci case, a case with a perfect factual history and an unimpeachable primary plaintiff, came along and has begun to shake up the cozy corruption that municipal fire departments have engendered for a long time under this terrible law.
6.8.2009 2:05pm
jrose:
b-rob: The test produced a distinct disparate outcome.

Cato: False. Everything you write sadly crumbles from there

If in this case Title VII was used as a pretext for racial discrimination, then the rest of what b-rob wrote about the law is not ripe for a court case. That doesn't mean his argument crumbles assuming an honest employer. And he asks a good question that deserves to be answered: what should an employer do?
6.8.2009 3:52pm
A. Zarkov (mail):
"The glaring error that leaped out at me was the authors' use of a normal distribution to model a process with only two outcomes: pass or fail."

That's not what they did, but first I'll give my results, which are similar and make no assumptions.

With (at the time) 7 captain positions open, promotions would be drawn from the applicants with top 9 scores from a total of 41 (25,8,8) test takers. I'm defining "pass" as getting into the top 9 group. There is another version of "pass," which means getting at least a 70 score on the combined oral-written exam. The outcomes are thus triples (w,h,b) where w+h+b = 9 and w = less 9, h = less 8 and b = less 8. So an outcome could be (6,2,1) meaning 6 whites, 2 Hispanics and 1 black made it into the top 9 group. We have a total of 55 possible outcomes some of these 55 will violate the "4/5 rule." I'm could say that the rule gets violated if anyone group has a pass rate greater than 80% of the smaller of the other 2 groups. In other words, should we include those cases where either a black of Hispanic pass rate exceeds 80% of the white pass rate? Strictly speaking, "yes," but I'll neglect those cases because they simply add to the overall rate of violation. Thus (4,4,1) violates the rule because 4/25 > .8 (1/8). Of course in this case it's the Hispanics violate the rule for both whites and blacks, but I'm only counting white violations in the interests of simplicity and because as a practical matter New Haven would never have failed to certified the test because whites scored too low. Indeed, there would have been a celebration party. Of the 55 possible outcomes, a total of 26 violate the rule.

If we compute the probability of each outcome triple (under a lottery) that violates the rule and sum we get 70%. The probabilities are computed from the multivariate hypergeometric distribution, which is the appropriate for a three outcome lottery of this kind.

In summary, for the captain test, even the maximally neutral lottery system would violate the "4/5 rule" more than 70% of the time.

Now back to the paper. The authors did not use the normal distribution to model a two-outcome process. They used it to generate the test scores, which are not two outcome and then counted the times those scores generated a rule violation. We could check to see if the scores are normally distributed but I won't bother.

"If the sample sizes are too small in this case, they will almost always be too small, making the 4/5 rule useless and irrelevant to proving disparate impact."

No. This case has a small number that get into the top 9. Usually in disparate impact cases, it's getting a job or promotion over some extended time period that generates a large sample size.

"Then, the authors do not justify or even mention their decision to separate the results of the captain's and lieutenant's exams."

No. We must treat the exams as separate cases. Pooling would be justified if everyone took the same test. But the captain's test is not the same as the lieutenant's. The "same body of knowledge" does not mean the two exams could not be different. If I measure a distance using a tape measure and a laser range finder, I cannot pool the measurements just because both measure the same thing. There errors are different and have to be analyzed separately.

"The best solution would to have a real statistician — with none of the authors' axes to grind — do an independent analysis."

What's a real statistician? Someone with a degree in statistics? That would have left out John Tukey who degree was chemistry. Statisticians are not licensed.
6.8.2009 3:59pm
Joseph Slater (mail):
Zarkov:

Assuming for the sake of argument that everything you say is true, remember, to defeat a disparate treatment discrimination claim by plaintiffs in Ricci, the city doesn't have to show that it really would have lost a disparate impact claim. It doesn't even need to show that it had an objectively reasonable belief that it would have lost a disparate impact claim. It only needs to show that the reason it acted the way it did was because it actually did fear losing a disparate impact claim.

That's because plaintiff only can win if plaintiff can show the real reason for the employer's act was intentional discrimination. Plaintiff doesn't win if plaintiff shows that the employer acted out of a motivation that was mistaken, dumb, or both, if it the reason was not discriminatory.

For example, suppose Employer fires Employee, and the Employer says the reason is that Employer thinks Employee is a Virgo and Employer hates Virgos. In a disparate treatment claim by employee, even if Employee shows that Employee is not a Virgo and that Employer had no good reason to think Employee was a Virgo, and that there's nothing wrong with Virgos, all that is not enough for Employee to win. Indeed, even given all that, the Employer can still win on summary judgment. See St. Mary's Honor Center v. Hicks, and the rule crafted by the (conservative) majority there.
6.8.2009 4:18pm
A. Zarkov (mail):
b-rob:

"Rather, they had to prove an entitlement (a promotion) was taken away because of their race."

That's why the case should have gone to trial so it could be heard on merits.

"For a number or reasons, they would not have won at trial because the test was not determinative."

Prior statement makes no sense.

"Exactly what, pray tell, would the conservatives aping the Ricci plaintiffs' line have the employer do?"

I don't know about what "conservatives" would have the employer do, but I can say what I think is reasonable. New Haven should have accepted the test results and gone ahead with the promotions on that basis. I see no evidence the test was defective. New Haven cannot prove that the "4/5 rule" applies because the sample size is too small, and applying the rule to results of the exams violates the very guidelines that govern the rule.

New Haven paid $100,000 for test which was designed to test necessary knowledge for the job. If the firemen who didn't make it into promotion group think the test was somehow biased against them then they should sue, and that case heard on the merits. At trial they could present experts who could point to questions that were not job related, and so fourth.
6.8.2009 4:25pm
Careless:

And he asks a good question that deserves to be answered: what should an employer do?

It's an obvious problem with the law. It's why everyone passes the test in Chicago: can't sue if you don't know why you weren't hired.
6.8.2009 4:25pm
Cato The Elder (mail):

Assuming for the sake of argument that everything you say is true, remember, to defeat a disparate treatment discrimination claim by plaintiffs in Ricci, the city doesn't have to show that it really would have lost a disparate impact claim. It doesn't even need to show that it had an objectively reasonable belief that it would have lost a disparate impact claim. It only needs to show that the reason it acted the way it did was because it actually did fear losing a disparate impact claim.

If this is actually true, and I'm not conceding that it is, then I feel that this is clearly a violation of the plaintiffs' Equal Protection Rights. There needs to be some sort of grounded and reasonable theory that underlies this fear to prevent employers from constructing de facto racial classifications via an iterated discarding of some necessary test, as one could easily construct an alternative hypothetical where a malicious employer illegally discriminates in exactly this manner.
6.8.2009 4:45pm
Joseph Slater (mail):
Cato:

It is absolutely true that, in disparate treatment (intentional discrimination) cases, employers will win if they can show that the employer took the action for any reason that is not discrimination -- including reasons that are facially stupid and nonsensical.

There is no question that, again, if I fired you because I thought you were a Virgo and I think all Virgos carry the virus for the plauge, I will win any discrimination case you bring against me, even if you show that you are not a Virgo, that Virgos do not carry the virus for the plague, and that I had no reasonable basis to believe either of those things. The point is, my actual motivation was not discrimination.

Now, you hypothesize a situation in which a "malicious employer" who really is trying to discriminate gets around these rules via subterfuge. Of course, if the plaintiff can prove that the employer's stated reason is a pretext for discrimination, plaintiff will generally win the case.

But as you can imagine, that's often not so easy to do. Which is one reason why employment discrimination plaintiffs lose a greater percentage of their cases than any other type of civil plaintiff. Glad to see you recognize the problem, though.
6.8.2009 5:12pm
A. Zarkov (mail):
Joseph Slater:

I agree that Ricci has an uphill battle, but not an impossible battle. New Haven's rejection of the test looks like invidious discrimination against the white applicants for political reasons. Their story about being afraid of a disparate impact suit just does not sound credible. Why should this case have not been heard on its merits at the trial level?

In any case, I'm opposed to the whole notion of disparate impact because statistical arguments in this context are weak arguments. I also believe an employer should be able to hire people intelligent people even for jobs that don't require high intelligence despite any impacts. This comes from my always recruiting the best and the brightest for my projects.
6.8.2009 7:19pm
Joseph Slater (mail):
Zarkov:

Let me be clear that I agree with you that Ricci does not face an "impossible battle." Indeed, I think this is a close and frankly fascinating case on which reasonable minds can differ. I wouldn't be shocked if the Supreme Court reversed. And while I don't think Sotomayor's ultimate position on the issue is wild, radical, racist, or even troubling, I do think that the Second Cir. probably should have written a full opinion/explanation.

I do disagree with you, apparently, on whether it could have been reasonable to fear a disparate impact suit. I wasn't there, so i don't know but disparate impact suits have been brought, and occasionally won, with shaky stats.

The idea of disparate impact theory, of course, is not to preclude employers from hiring intelligent people, but rather to try to make sure that employers don't put unnecessary obstacles in the path of groups who are traditionally discriminated against. If X requirement really does greatly disproportionately disadvantage one group, let's make sure that requirement really is business related. It probably also exists so employers can't find easy proxies for intentional discrimation that would be harder to prove. Don't want to hire women? Just say you believe sound minds = sound bodies and require all your employees (at your law firm) to be able to bench press their weight. . . .
6.8.2009 9:57pm
Desiderius:
Slater,

"The idea of disparate impact theory, of course, is not to preclude employers from hiring intelligent people, but rather to try to make sure that employers don't put unnecessary obstacles in the path of groups who are traditionally discriminated against"

So you either have a test that:

(a) screens out those obviously unqualified for the job, with the resultant pool of applicants available to be hired based on other factors

(b) provides an objective ranking of those most capable to execute the duties required of the position

Given the history of, say, civil service examinations, and what came before (the Tammany Hall-style spoils system), your confidence that the other factors in (a) will be just seems overly sanguine. Of course perfect objectivity is impossible, but (b) still seems a more promising prospect then relying on the tender mercies of political log-rollers.
6.9.2009 12:13am
A. Zarkov (mail):
Slater:

"I do disagree with you, apparently, on whether it could have been reasonable to fear a disparate impact suit."

Whenever you don't select anyone for a job or promotion, you face a possible lawsuit. Look at Coe v. Yellow Freight. He didn't get promoted or hired for various positions, and he sued and sued even though his case was weak. Just as you said, the plaintiff has an uphill battle. New Haven might have faced a lawsuit from one of the black applicants who didn't make the top group, but on what basis? Disparate impact? The statistics don't support that. I suspect the ranking was pretty much a function of the time and effort put in to prepare for the test. The blacks didn't study enough and that would come out at trial.

"The idea of disparate impact theory, of course, is not to preclude employers from hiring intelligent people, but rather to try to make sure that employers don't put unnecessary obstacles in the path of groups who are traditionally discriminated against."

Well that brings us back to Griggs v. Duke Power. Duke used aptitude tests in hiring and promotion and SCOTUS said that had a disparate impact on blacks. That does seem to preclude employers from hiring intelligent people unless they can show intelligence is a necessary requirement for the job. I think that almost always it's better to have a more intelligent rather than a less intelligent employee even for what might appear to be a menial task. Many of our captains of industry started out at the bottom of the company.

The decision in Griggs was great for the universities as employers now rely on them as the aptitude screen. Lot's of English and History majors with PhDs end up doing something completely different. This is why someone with a Phd in engineering can go to a tier 3 law school and still work for a large firm that would normally require a tier 1 graduate. The firms know the applicant has already proved himself by getting the engineering PhD which is much harder than law school. The dirty little secret is that law schools exist more for the benefit of the professors than anyone else. Keep this quiet.
6.9.2009 12:51am
Desiderius:
"The decision in Griggs was great for the universities as employers now rely on them as the aptitude screen."

Bingo. And its is the downtrodden whom "disparate impact" theory is designed to aid, but who cannot afford to forgo the income necessary to pursue the credential, leaving aside the cost of the credential itself, who ultimately bear the brunt of the present system.
6.9.2009 1:16am
Tony Tutins (mail):

We must treat the exams as separate cases. Pooling would be justified if everyone took the same test.

In industry, one has to make decisions all the time on which process to use, or if a new product is just as good as its predecessor. Samples (and sampling) cost money, so samples are aggregated as far as is reasonable. You can't assert that subgroup populations used are too small unless you've made a good faith effort to combine them.

To do so here might have undermined support for the authors' chosen outcome.

while I don't think Sotomayor's ultimate position on the issue is wild, radical, racist, or even troubling, I do think that the Second Cir. probably should have written a full opinion/explanation.

A terse thumb's up is not limited to the Second Circuit. One of the cases on which the plaintiffs rely to support their small sample size statistical argument is (one of the many) Stewart v. City of St. Louis. Here black firefighters were denied promotion even though tests made them eligible for promotion - the 2:1 discrepancy between white passers and black passers was explained away as based on small sample size. It was hard to find the argument, however, because the Eighth Circuit affirmed the District Court's reasoning in a brief, per curiam opinion. I'm not sure why a procedure commonly used to dispose of racial discrimination cases becomes somehow suspect when a wise Latina is nominated to the Supreme Court.

(Moreover, the DC opinion was apparently not published in F.Supp; I could find only a Lexis citation: 2007 U.S. Dist. Lexis 38473 or 38421. )
6.9.2009 2:54am
A. Zarkov (mail):
"In industry, one has to make decisions all the time on which process to use, or if a new product is just as good as its predecessor. Samples (and sampling) cost money, so samples are aggregated as far as is reasonable."

True enough. But this is not the usual industrial problem where being wrong is not that serious. Working on Ricci is work in a goldfish bowl, and anything you do will be highly scrutinized. If we are going to pool the data then we had better be able to justify doing so.

Suppose the captain exam and the lieutenant exams have biases in different directions. Then you wash out the effect you're trying to detect. It would help to have access to the questions. I will do a QQplot of the exam numerical scores to see if they how they might differ.
6.9.2009 3:46am
devil's advocate (mail):
TT




Conservatives advocate judicial activism this time: They want a federal court to order a government agency to promote people.



i disagree. conservatives might want the court to invalidate Title VII, at least with respect to these circumstances if not more broadly.

now this would be compelling govt. not to do something, albeit this is a semantic conundrum

The most anti-judicial activist conservatives -- think Lino Graglia -- would protest the courts disturbing the government action. So they would object to positive or negative interference from the courts at least that purported to invalidate the legislative underpinnings.

But if you construe judicial activism to be positive interference the semantic here is truly the opposite.

Now if the court were to, instead of invalidating Title VII, compel the promotion as an enforcement of Title VII, which is also a possibility I suppose, your argument carries more force. But the issue is still fraught because the government here claims to be enforcing Title VII so such an action is effectively a rollback of their enforcement of Title VII.

Semantics matter. I take your point but don't find it prima facia persuasive.
6.9.2009 9:50am
devil's advocate (mail):
Adam B.:


Since when do an advocate's unsworn offerings during oral argument qualify as evidence of truth of the matters asserted?


This isn't a precise answer but recalling that this is a summary judgment matter, the agreed circumstances/evidence is to be contrued in favor of the plaintiff, no?




How about making the entire transcript available rather than your carefully pruned excerpts?


Actually this was first asked by Sotomayer's doubters, see, e.g. Cato. His point was that given the importance of this case and the lack of explication from the 2nd circuit itself that access to the arguments before the 2nd circuit is critical and the lack of a transcript produced by the media or the court is unfortunate. (This would be additionally useful given allegations such as Rosen's regarding SS's conduct during arguments).

Noone has posted a link to the oral argument although there was a comment that said the WSJ provided it.

Brian
6.9.2009 10:02am
Joseph Slater (mail):
Desiderius:

I'm not "overly sanguine" about any form of testing process. I was merely describing the law as it is.

Zarkov:

Tony Tutins provided some helpful cites for a point I made: employers do sometimes lose disparate impact cases where the stats are shaky because of small sample sizes. So I continue to believe that fear of losing a suit could well have been the city's motivation.

But moreover, remember that if the city can show it acted out of a fear of losing a disparate impact case, the city does not have to show that its fear was objectively reasonable. It only has to show that the fear -- stupid and illogical though it may have been -- was the actual reason for the decision. That's disparate treatment law for you.

Desiderius and Zarkov:

As I also mentioned above, another point of disparate impact theory is to get at what likely was intentional discrimination that the employer is attempting to mask through "neutral" policies. Look at the history of the employer's behavior in Griggs, for example. It's pretty obvious the employer was trying to preserve segregated occupations and promotional lines.
6.9.2009 10:17am
devil's advocate (mail):
Shelby C



...there's what's called the four-fifths rule:





Interesting. I must have skimmed that part :-)

Unfortunate name, though.



Jeez, remind you of another fifth's rule, say Art. I Sec. 2 para 3.

The more things change . . .

Brian
6.9.2009 10:28am
devil's advocate (mail):
want to nominate

A. Zarkov and Joe Slater with honorable mention to Desiderius and Tony Tutins later contributions

for civil disputation honors on this thread that was headed toward the precipice of flame war and pulled back.

Brian
6.9.2009 11:42am
Joseph Slater (mail):
Thanks, Brian/devil's!
6.9.2009 1:35pm
troll_dc2 (mail):
This has turned into a great thread.

I have no idea what the justices are thinking, but for me the case has come down to one of two possibilities. The first is that the city should win if it can show that it simply feared that it would lose a lawsuit if one were brought challenging the exam-based promotions. But the city should lose if it is shown that it was simply unhappy with the exam results because they meant that no blacks or Hispanics would be promoted. Under this analysis, the case is less dramatic than it has been portrayed to date, which is that the city can be sued successfully no matter what its decision is.

The Supreme Court, I am confident, would never sanction such an outcome. The best way out would be to look for "actual" intent, whatever that is. (So we seem to have here a merger of disparate impact and disparate treatment.) The Court would be justified in remanding the case to explore this issue. Mixed-motive analysis might be relevant here (if it were found that city had both a lawful and an unlawful motive).

Can we all agree that the district court erred in failing to conduct a trial?
6.9.2009 2:31pm
Desiderius:
Slater,

"I'm not "overly sanguine" about any form of testing process. I was merely describing the law as it is."

Apologies for lumping you in with disparate impact advocates. Your cogent explication of the doctrine would be all the more impressive if you do not indeed count yourself among their number.

troll_dc2,

"The first is that the city should win if it can show that it simply feared that it would lose a lawsuit if one were brought challenging the exam-based promotions."

So what recourse does one have if one wishes for one's institutions to cease living in such trembling fear of hypothetical lawsuits or, and I find this more likely, using the plausibility of such fear to justify actions that would otherwise not pass muster - legal, moral, rational, or otherwise.

As a teacher, I'm thinking of such things as zero-tolerance policies and strip-searches in this category.
6.9.2009 4:57pm
devil's advocate (mail):
Desiderius



So what recourse does one have if one wishes for one's institutions to cease living in such trembling fear of hypothetical lawsuits . . .


Get elected mayor and stop acting that way on behalf of those institutions.

If these are close calls and tough cases, be willing to fight instead of fold.

Troll_dc2


Can we all agree that the district court erred in failing to conduct a trial?


I haven't read the decision under so I don't want to sound smug or be hasty in offering my affirmation of this consensus. But if I can take to the bank the proposition you stated earlier in your post, that this boils down to examining the evidence whether motive here was actually fear of lawsuit or political desire for a different outcome on the test, I would hope that the consensus is broad.

Brian
6.9.2009 5:09pm
troll_dc2 (mail):

So what recourse does one have if one wishes for one's institutions to cease living in such trembling fear of hypothetical lawsuits or, and I find this more likely, using the plausibility of such fear to justify actions that would otherwise not pass muster - legal, moral, rational, or otherwise.


I'm afraid that the law is such that fear cannot always be eliminated. I am sure that you would agree that there is a beneficial aspect to fear in that it can cause an employer to think about what it is doing or wants to do. But the fear has to be in good faith; otherwise it is a pretext.

Unfortunately, fear can be honest but unreasonable. You referred to "hypothetical lawsuits." It has been my experience that people sue when they are unhappy, and they let their contingent-fee attorney figure out how to raise the arguments to support the claim. For instance, I saw today a short article about a lawsuit against a restaurant that refused to hire "a woman who says she was denied a job as a bikini-clad barmaid because her speech was too 'ghetto.'" The restaurant said that it has a right to refuse to hire inarticulate speakers. But there are racial undertones, the plaintiff's attorney asserts. I do not know the case facts, of course, but I am quite skeptical of the plaintiff's claim.

As you know, lawsuits are expensive, and they can cloud reputations. So employers all the time try to avoid them, and, yes, they sometimes do awful things in the process (like disowning white supervisors who discipline minority-group employees for obvious shortcomings). Moreover, even when they think that they are doing the right thing and even when they present a lot of evidence that they did so, a jury can disagree nonetheless. So a prudent employer may find itself having to calculate which group of employees will be more expensive to deal with if there is a lawsuit.

Is this a legal, economic, and moral mess? You bet. I would be happy to replace the present system with something that resembles the NLRB system, with centralized decisionmaking both in terms of which claims can be litigated and who decides. In fact, that is what the EEOC originally was supposed to be all about until it became clear that taking away that power from it was the only way to get the bill through the Senate in 1964. A centralized system would provide some clarity for everybody, and there still could be court oversight to make sure that the EEOC was not too ideological in its decisionmaking.
6.9.2009 5:24pm
Joseph Slater (mail):
At risk of being deemed less commendable by Desiderius, I would like to paint a fuller, more complete picture of the general landscape of employment discrimination law (going beyone Ricci).

On the one hand, yes, some employers are overly sensitive about being sued under Title VII, and take actions that they wouldn't and in an ideal world shouldn't because of that.

On the other hand, we are still -- yes, today, still -- living in a world in which old fashioned discrimination against blacks and women is unfortunately quite common. For example, the study showing that those with "stereotypically black" names do much worse than those with "stereotypically white" names in getting interviews, and the study showing that testers identical in all but their race get many more job offers after interviews if they are white. I used to do employment discrimination cases, and yes, while some employers think relative trivia is "harassment" and over-react, real, gross, awful harassment -- not just "hostile work environment" but also old fashioned quid pro quo "sleep with me or I'll fire you" kind -- is all too common.

Also, as I mentioned upthread, plaintiffs in employment discrimination do worse than plaintiffs in pretty much any type of civil litigation, at pretty much every level. So while employers may fear litigation expenses, plaintiffs lose the actual litigation quite regularly. In part, that's because of legal rules that make it very, very difficult to prove disparate treatment (ironically, the problem plaintiffs in Ricci are facing).

I don't have a big solution for this -- I don't even have a solution for the Ricci case -- but that's the whole picture.
6.9.2009 6:02pm
troll_dc2 (mail):
Mr. Slater, there is one point that you left out, but that is understandable because there is no way to measure it.

There are many more EEOC and state FEP charges filed than there are lawsuits brought. Sometimes apparently valid claims are successfully conciliated and never go to court. Sometimes cases go to court and are settled early on. Sometimes they are settled after a motion to dismiss or some other preliminary motion is denied. Sometimes they settle just before trial. When we talk about wins and losses, the settled cases do not get counted (nor do the cases that are dismissed voluntarily).

So in terms of cases that go to judgment, the overwhelming majority are wins for employers. But in terms of the entire universe of cases, I suspect that the percentage is much lower. The system is set up to try to get settlements (some of which may not be justified in theoretical terms but are cheaper for the employer or give it more control over the scope of the remedy).

Your post otherwise is pretty much "there."
6.9.2009 6:36pm
Desiderius:
devil's advocate,

"If these are close calls and tough cases, be willing to fight instead of fold."

Me and what trust fund and/or minting authority?
6.9.2009 7:10pm
Desiderius:
Slater,

"Your post otherwise is pretty much "there.""

Agreed. I am very much an amateur among professionals in this discussion, and your contributions are invaluable to furthering my understanding, and, I would hope, many others reading along.

Slater,

"For example, the study showing that those with "stereotypically black" names do much worse than those with "stereotypically white" names in getting interviews, and the study showing that testers identical in all but their race get many more job offers after interviews if they are white."

Is it your estimation that the threat of "disparate impact" lawsuits and the like ameliorates or exacerbates this problem?
6.9.2009 7:18pm
troll_dc2 (mail):
I hope that this threead does not close down for a while. We have some Title VII specialists here to talk shop.
6.9.2009 7:22pm
devil's advocate (mail):
slight threadjack but it follows on troll_dc2 comment.

I believe the default comment periods are 5 days, so if you see at what time and what day the post was made you know how long it is going to go - of course once it gets a few days out or down to 100 plus comments you don't usually get a lot of new folks jumping on.

There has been some discussion of comments, e.g., on another contemporaneous thread regarding how cowardly it is of certain conspirators to occasionally not enable comments (or in some cases they offer a much shorter comment window but usually mention it in the post).

According to MN Ralph:


It's almost always a dead giveaway that the contributor knows his post is superficial or slanted, or he just wants to get in a cheap shot at a person or cause without facing criticism.


I beg to differ. Randy Barnett had several threads devoted to his own decision not to enable comments on any of his posts because he felt compelled to read them a bit maybe respond and it would be his responsibility to act as the politeness police if they get out of hand. It seems that those explanations could readily extend to other conspirators and, indeed, the rancor here is, no surprise, directed at David Bernstein.

While I suppose one cannot entirely discount MN Ralph's analysis with regard to David Bernstein, because he enables comments on some posts and not on others, I see a degree of judiciousness in his conduct. While it may protect controversial assertions of his own, it has the obvious purpose of chilling the most out of hand flame wars that can result from the issues on which he focuses, Israel/US relations in particular.

I find these conversations a privilege, afforded to me by the participants and, of course, by the conspiracy. If DB wants to kill comments on a couple of posts get over it.

Occasionally I wish that a thread would continue past the 5 days and I once begged Orin Kerr to add a comment to a thread that had closed because I started writing it, in typically long-winded fashion, while the 5 day period was open, but by the time I was done it was closed. I don't think he obliged me, and it was kind of a cheesy thing to ask anyway, but it was some point I cared about. As if anyone is ever going to read the 134th comment on some expired volokh thread.

And if it is a hot topic, there will always be another thread, witness this topic, where I first mentioned the Leiter business as like the 250th comment or something on Adler's orginal thread. Of course it went no where, but then I got satisfaction here, so I can't even sing the Rolling Stones tonite, man I miss the Sunday Song Lyric.

Best,

Brian
6.9.2009 8:52pm
A. Zarkov (mail):
Tony Tutins:

I looked at the test data some more, and I found that the scores on the two different exams do not have the same distribution. That makes pooling more difficult if not impossible. We could do a transformation to make one set match the other, but that introduces new statistical complications and opens up the analysis to criticism. If the results for court critically depend on the pooling, and you can't really justify it, you might trash your case. Even if you do something valid you will need to explain it in an understandable fashion and that can be very hard to do for a jury. Their eyes will glaze over-- they won't understand a word and will ignore the evidence.

If you have any good ideas on pooling that have proved successful for you, I'm anxious to learn about it.
6.9.2009 9:21pm
A. Zarkov (mail):
Joseph Slater:

"... the study showing that those with "stereotypically black" names do much worse than those with "stereotypically white" names in getting interviews,..."

I have heard about that study, and I'd like to get the details. I am ever the skeptic, and until I can go over it myself, I'm not convinced. I can tell you that I have "broken" many a study-- especially those coming out of government. Remember the New Jersey Turnpike study that showed blacks are stopped more often for speeding? It came apart big time, and DOJ tried to bury the new report, but the Bergen Record leaked it.

I'm not saying I necessarily disbelieve the result, I just want more than the newspapers tell you.

There is something called the "file drawer effect." This is where studies that don't give the desired result never get published. The study gets done again and again until the desired result comes up. It can be very hard to find out when this happened because the study looks ok absent the unavailable knowledge.
6.9.2009 9:30pm
troll_dc2 (mail):
In today's Washington Post, Ruth Marcus has a column that is relevant to this thread. For Sotomayor, a Fine Line in New Haven

She makes an observation that cuts to the heart of the problem:


The paradox at the heart of Ricci v. DeStefano is that federal anti-discrimination law requires employers to examine the racial implications of their hiring and promotion procedures. Yet the law and the Constitution could also expose them to a claim of intentional discrimination if they view results through a racial lens and take action accordingly.



She further observes: "Minorities are far more likely to exercise the levers of political power, which is why I think the panel should have ruled differently and sent the case back for further review, as the Obama Justice Department has urged. There is some evidence that New Haven's claimed fear of a lawsuit was mere pretext and that what was really going on here was racial politics -- city officials protecting themselves with minority constituents."

I cannot wait for the Court to issue its opinion.
6.10.2009 3:57pm
A. Zarkov (mail):
troll_dc2:

Thanks for the pointer. This has been a very good thread indeed with people making substantive and informative comments. Now we are pretty well prepared to read the SCOTUS decision which should be out soon.
6.11.2009 1:42am

Post as: [Register] [Log In]

Account:
Password:
Remember info?

If you have a comment about spelling, typos, or format errors, please e-mail the poster directly rather than posting a comment.

Comment Policy: We reserve the right to edit or delete comments, and in extreme cases to ban commenters, at our discretion. Comments must be relevant and civil (and, especially, free of name-calling). We think of comment threads like dinner parties at our homes. If you make the party unpleasant for us or for others, we'd rather you went elsewhere. We're happy to see a wide range of viewpoints, but we want all of them to be expressed as politely as possible.

We realize that such a comment policy can never be evenly enforced, because we can't possibly monitor every comment equally well. Hundreds of comments are posted every day here, and we don't read them all. Those we read, we read with different degrees of attention, and in different moods. We try to be fair, but we make no promises.

And remember, it's a big Internet. If you think we were mistaken in removing your post (or, in extreme cases, in removing you) -- or if you prefer a more free-for-all approach -- there are surely plenty of ways you can still get your views out.