Wednesday, January 21, 2009

Judicial Pay Redux

David Stras

Yesterday's edition of the New York Times had an interesting article written by Adam Liptak discussing whether federal judges should receive a pay raise, an issue on which I have previously commented here. In the article, which is available here on the New York Times website, Liptak discusses two recent studies performed by Scott Baker and Mitu Gulati, Eric Posner, and Steven Choi, which both suggest that judicial quality would not improve with higher judicial pay. When I spoke to Adam Liptak last week to provide background for the story, my reaction to these studies was very similar to that of Frank Cross, who suggested that the "new studies went off the rails" because "'we don't have a good measure of judicial quality.'" In my opinion, empirical tools are relatively ineffective in studying highly subjective questions like the quality of judicial work. In part, the problem is that it is difficult to find two constitutional law professors who would agree on what constitutes quality work. I have had disagreements with colleagues, for instance, who argue that lengthy, scholarly opinions are of higher quality, while my view is that brevity and clarity should be the goals of judicial opinion-writing. Yet variables such as the number and length of published opinions and even how quickly the opinions are produced are often variables in these types of studies.

As I have stated before, it is not that I oppose a judicial pay raise (actually I don't), it is just that I think that the Administrative Office and the Chief Justice have failed to make the case to this point for anything greater than regular, annual cost-of-living increases. As Judge Posner pointed out in his book How Judges Think (which is excerpted in the Times article), only 12 federal judges resigned during the period from 2000-2005 out of a total of 1,200 active and senior judges working during that period. That is only a 1% resignation rate, which I suspect is quite a bit lower than in the vast majority of other comparable professions. And similar to the point I raised in my prior post about the potentially misleading data advanced by the Administrative Office, which appears to use extraordinarily high comparable salary data for senior law school professors and omits any discussion of the judiciary's very generous pension benefits, Posner points out that it is no accident that the Chief Justice uses 1969 as the benchmark for judicial pay given that federal judges received a 30% pay raise that year. In Posner's opinion, using 1969 as the benchmark is "misleading."

My point in this series of posts is not to minimize the important contribution of federal judges or to take a firm position on the merits of the judicial pay debate. (Indeed, I have written in the past that we should consider giving even more generous pension benefits to federal judges.) Instead, my hope is that the Administrative Office, the Chief Justice, and academics can produce a greater volume of data that more accurately represents the case for (or against) a judicial pay raise. It is only then that we can have a serious conversation about whether, and how much of, a pay raise is warranted.

Note: Cross-posted at Empirical Legal Studies blog


With respect, I am absolutely amazed by this. I suppose that the rationale is that such Judges, unlike, say, professors were insulated from a free market selecton process when initially designated for appointment and are similarly protected from the invisible hand of the market because, let us say, their "mobility" in the judging sector is restricted.

Where does this kind of rationale come from?

Judges are not civil servants and they are not factory (or law school) workers. We have a system that values the independence of judges, that strives to keep them wound in cotton batten for fear of contaminating the decision-making processes and we suggest that there should be a merit based assessment of their --apparently-- collective work?

The bottom line is this. There can be no true economic value in assessing the merit of the work of judges collectively where there are bound to be so many inconsistencies and variables.
There can be no question that an independent assessment -- short of the blunt and vague instruments of appeal or impeachment -- otherwise prejudices the very independence we strive to see met.

If there are problems with the appointment system, fix the appointment system. If the appointment system cannot be fixed without damage to other fundamental legal and constitutional values then leave it where it is and live with inevitable downside consequences.

And for heaven's sake, grading on prolixity or paucity of length of decisions?

Gheesh. Surely there are other navels to be gazed into.

Professor Stras,

Thanks for the link to the crosspost, that looks to be an interesting resource.

My gut reaction is that you, like Roberts, have a partisan agenda, and if there were empirical studies supporting your position your skepticism would evaporate---much as Scalia's fears of the counter-majortiarian nature of the bench seem a thing of the past. I would be happy to be convinced otherwise.

What do I suppose that agenda to be? Simply to attract more money motivated minds to the bench as ideals driven jurists are less likely to march in lockstep with GOP policy.


Always nice to see word like "paucity" so early in the morning.

I reckon one criterion Stras and Roberts would love to track and pay for is the ability to get opinions out without dissents---well, GOP-friendly opinions. Didn't Roberts take the seat claiming an intent to create a united bench? Of course that can be two edged. If such uniformity comes from reasoned argumentation and a deference to logical conclusion in the face of partisan preference, fine. If it comes from stonewalling and personal persuasion, perhaps not so fine.

Are some, perhaps many, judges seeking celebrity status outside of judging? It can be expensive maintaining a celebrity lifestyle. More and more SCOTUS justices are are coming out of the judicial closet to express their views on many subjects, as permitted by the First Amendment (subject to certain limitations, of course).

I wonder how much heavy lifting is done by the judiciary? Consider the decreased SCOTUS load of cases over recent years. And look at the support the Justices receive. And look at how long it takes for decisions to be made. Yes, Justices (and others in the judiciary) are entitled to their social lives and desire for adulation and celebrity. But they know this before joining the ranks of lifetime appointments. They no longer have to worry about rainmaking and billable hours and bad decisions handed them from time to time by judges.

Of interest is the mortality rate of members of the judiciary and the impact of stress, assuming their jobs are indeed stressful, at either the trial or the appellate level. I remember a Suffolk Probate Court (MA) judge back in the early 1960s responding to an attorney's reference to cases of the Supreme Judicial Court (MA) contrary to the judge's decision: "If you don't like my decision, take it up on appeal; those guys [judges] have plenty of time on their hands."

The "have a partisan agenda" argument is harder to make to the degree Stevens agrees.

C-SPAN recently aired a Q&A with him and a lower court friend of his & he agreed when the guy argued low pay was a big problem. It probably can be found by doing a search on the American & The Courts page.

As the "celebrity lifestyles" ... I think a stronger claim is made as to lower court judges on this issue. The celebrity angle is much less strong on that front.

I'm not crying for Supreme Court justices. I don't doubt they work hard enough, but the reduced docket, clerks, and prestige of the job helps a lot. As does life tenure.

And, on this limited stage at least, I think all these factors will provide enough qualified candidates.

Joe, Touche on the Stevens angle. One shouldn't reason from authority, and arguably that's what I've done in my sense that this is a Federalist Society talking point, based on the pens I've seen spill the most ink on it. Peace.

Imagine a world where our highest officers did it for free (i.e., for the sheer, naked power of it all). "...but I'm not the only one..."

I agree that the studies referred to in Liptak's article have their weaknesses (and, in fact, a couple coauthors and I published a response to Baker's paper in the BULR).

At the same time, the statement that "empirical tools are relatively ineffective in studying highly subjective questions like the quality of judicial work" strikes me as simply incorrect. As a social scientist, I study a range of phenomena -- political ideology, racism, social capital, religiosity, etc. -- that are both difficult to measure directly and critically important to understand. Any measure of such phenomena is necessarily imperfect, but those imperfections shouldn't lead us to throw up our hands and give up any hope of knowing about such things.

When one of my own Ph.D. students tells me that something -- judicial activism/restraint, or political efficacy, or whatever -- "can't be measured," my response is always "You're not thinking hard enough." Any science should be cumulative, which means later work will improve upon earlier work. For all their drawbacks, the Baker and Gulati et al. studies at least provide a starting point for addressing the question of judicial pay and performance empirically; the goal now should not be to dismiss their work because it isn't perfect, but to build upon and improve it.

I've got an idea, Prof. Stras:

Why don't we auction off judgeships to the lowest bidder? I'm sure the market will quite adequately take care of "quality" issues. People living in districts served by substandard judging can always "vote with their feet" and more to those with more skilled judiciary.


The best argument judges have is that we used to pay them more in real dollars. In fact, federal salaries in general peaked around 1900 -- TR got about $1.8 million in today's dollars, while we paid W $400,000 -- and have declined ever since.

At the Convention, Benjamin Franklin argued that federal officials like the president shouldn't be paid at all. The counter argument, which prevailed, was that if that were the rule only the rich would be able to take the jobs.

While I find that a persuasive argument, the fact is that today only the relatively rich and well-connected can even run for office at the federal level. I doubt the salaries add much to the incentive.

"While I find that a persuasive argument, the fact is that today only the relatively rich and well-connected can even run for office at the federal level. I doubt the salaries add much to the incentive."

Agreed. But is it worthwhile considering what the incentives are other than salary running for federal office? In the early days of my practice that included federal taxation, there had been developed a "net worth" theory in ascertaining whether a taxpayer was not disclosing all of his/her income. This theory seemed to have faded. But consider such a theory with elected officials (both federal and state). To what extent are campaign contributions somehow adding to their net worths? Or do these officials rely upon the revolving door for their just rewards?

While this post concerns the judiciary, perhaps federal, it might be considered for elected state judges and their incentives, other than salaries.


I appreciate your comment and realize that my statement may have come off as too strong. I think it tends to be the case that extremely subjective questions like judicial quality--where we cannot even agree on what factors are relevant to the inquiry--are not particularly susceptible to empirical measurement or analysis. That is, any measure of judicial quality will be subject to vehement debate before we can even get to the merits of the empirical results. Unlike ideology, racism, and judicial minimalism, I am not even sure that we can agree on what constitutes high-quality judicial work.

But I do think that there is every reason to continue to try to measure judicial quality. Before Martin-Quinn scores, which are by no means perfect, our measures of ideology were not particularly good. I have no doubt that the Martin-Quinn scores have advanced the ball as I have used them in my own research. I just have my own doubts that we will be able to make strides on measures of judicial quality, but I could be wrong.

Why not start by measuring the potential factors/outcomes and attempt to circumscribe (rather than measure directly) judicial quality? Surely outliers exist that would elicit consensus!

That may not count as "empirical measurement" in your book, but social scientists have been comfortable with the necessity of inference for some time now. Why is the concept of "judicial quality" any harder to examine than the concept of musical taste or what constitutes good leadership?


I know you're not being facetious, so can you go deeper? What, with the caveat that law isn't your first hat, suggests itself as a criteria set for judicial quality?

You set the task of putting salt on that bird's tail next to the task of similarly salting "musical taste" and "leadership". I suggest that leadership in particular is much more amenable to consensus, and even musical taste can be correlated to the market and adjusted for erudition. But judicial quality? Part of the nature of judging is that judges make pronouncements on the contentious with both possible outcomes most often at least plausible. Seems that makes the domain much more tenuous, a wispy little tail with hardly room for even a grain of NaCl.

Well, just to throw something out there, but doing content analysis of those venues in which professionals discuss decisions might be one quasi-objective way to measure the limits of judicial quality. You could start with a general hypothesis like "poor decisions lead to disparaging remarks in the literature" and refine from there.

The variables to which David objects (length of opinion, number of published opinions, speed of production) try too desperately to measure quality directly in some way. One could try to measure a scholar's worth by the length of his bibliography perhaps, but obviously that doesn't cut it, as the sources cited may be erroneous or irrelevant. Far better to pick a number of factors and outcomes that are less directly connected to the thing that you are trying to study. Think of "judicial quality" as the "petit objet a" in a Lacanian Venn diagram, and you'll get my point.

The important thing, however, is that we don't deify empiricism as the sole prerequisite for knowledge; it's certainly possible to use empirical methods to investigate processes that aren't empirically measurable. Yes, there may be subjectivity involved, and the author of such a study will have to defend their work in a manner you should find familiar. But all too often people cry "subjectivity" to avoid doing work--I agree with Christopher Zorn above on this point--and we fall into a self-abusive recursive "post-modern dilemma."

If the question to "Are some judgments better than others?" is "yes," there must be a way to measure it, otherwise it is a nonsensical statement. Whether it qualifies as an empirical technique is a different matter altogether.

I wish Anderson would weigh in on this one; his legal experience and his time in the Kantian trenches would probably prepare him better for the question at hand.

Post a Comment

Older Posts
Newer Posts