Academic

Chief Justice Roberts has just issued his year-end report on the federal judiciary. As in past years, he urges Congress to raise the salaries of federal judges. The accompanying data can be found here. The latter web page might seem to make up in beautiful graphics for what it lacks in statistical rigor, but let?s hope that lower-court judges do not use it as a guide for evaluating expert testimony. Roberts? argument boils down to two points: that the real wages of federal judges have declined since 1969 because of inflation; and that today first-year associates in some law firms, law school deans, and many senior law professors make more than judges do.

(Last year Roberts also noted that federal judges make less than counterparts in the UK; however, it turns out that in most developed countries with sophisticated legal systems, judges make less than American judges do.  The judiciaries in those countries are also less prestigious, yet they seem to perform just as well as ours does.)

The fact is that in 1969 federal judges were extremely well paid, near the top of the income distribution. Today, they are in the top 5 percent or so. There is no particular reason to think that Congress got judicial pay right in 1969; it could have been too high. In the meantime, the real wages of lawyers have increased, so that now many lawyers make more than federal judges. It is also true that many CEOs make more than the president of the United States. There is no particular reason to think that government salaries should be as high as, or higher than, private sector salaries, even salaries of jobs that seem to draw on comparable skills. It all depends on the relative value of government employees, and no one knows how to measure the social value of the federal judge.

Roberts worries that as the relative salaries of judges decline, the quality of the federal judiciary will decline, because fewer people will give up high private-sector salaries in order to become judges, and that judges will retire in order to seize opportunities for higher pay.

However, the data do not support this argument. Albert Yoon has performed a statistical study of judicial retention, and finds no evidence that judges have retired at higher rates as their relative salary has declined. See also this CRS study, which comes to the same conclusion (and debunks many other claims Roberts has made about judicial pay). It is worth pointing out that the non-pecuniary benefits of being a judge are enormous?one has power, status, an interesting job, job security, and an extremely generous pension?and few people want to give this up. Indeed, many judges continue to work even after their pension fully vests, and since their pension is the same as their salary, these judges, in effect, work for free.

Scott Baker has also conducted a study; his study exploits the different cost of living levels across states, to see whether there is any correlation between judicial quality?measured in multiple ways?and the effective standard of living of particular judges. He finds none of any importance.

Finally, with Steve Choi and Mitu Gulati, I have looked at whether the salaries of state high court judges?which vary greatly?are correlated with judicial quality, and found little evidence of such a correlation.  Our study is here.

Why wouldn?t more highly paid judges be better? There are several possible explanations.

First, the non-pecuniary benefits of being a judge might well swamp the cash benefits. We already know that partners will give up a million dollars a year to become a judge; would it make a difference if they give up only, say, $950,000 per year?

Second, federal judges are appointed, and so the quality of the federal judiciary depends in the first instance on the quality of the appointments process. If presidents and senators use appointments as patronage opportunities (think of Harriet Miers), then a salary increase just improves the value of patronage, not the quality of judging.

Third, even high quality people do not necessarily want to work hard once they become judges. So a high salary might attract higher quality people, but if these people do not work hard because there is no sanction for not working hard, and there is none, then judicial output will not necessarily improve.

A final point, one that is rarely made in the debates, is that federal judges should not be paid more than the value of their activity. If you overpay federal judges, you attract from the private sector people who are more productive as practicing lawyers.

There is no easy way to figure out the social value of a federal judge, but until some progress is made in coming up with a figure, Justice Roberts? arguments will remain in the realm of speculation.

For other comments on the Chief Justice's arguments (past and present) in favor of a pay raise, see:

The WSJ law blog;

Ilya Somin;

Dahlia Lithwick