Courtiers of the Marble Palace: The Rise and Influence of the Supreme Court Law Clerk
By Todd C. Peppers
(Stanford University Press, 310 pp., $55)

Sorcerers' Apprentices: 100 Years of Law Clerks at the United States Supreme Court
By Artemus Ward and David L. Weiden
(New York University Press, 337 pp., $39)

Within months of each other, two remarkably similar books about Supreme Court law clerks have appeared. Both are the work of young political scientists; both are sober in tone, academic in style, meticulous in scholarship. In organization (basically chronological, but with some descriptive material at the beginning) and length they are almost identical, though Courtiers of the Marble Palace contains more historical and human particulars, and Sorcerers' Apprentices is more statistical and makes a greater effort at explanation. Both are excellent books only slightly marred by misleading titles, for their actual subjects are not "courtiers" or "sorcerers' apprentices," but nerdy professionals. The books should displace their gossipy (though pretty accurate) and unanalytical predecessors, Bob Woodward and Scott Armstrong's The Brethren and Edward Lazarus's Closed Chambers.

They are depressing to read, though. They put one in mind of Max Weber's image of the "iron cage" of modernity, in which the progressive bringing of all activities under the rule of reason--their subjection, in other words, to the rational methods of bureaucracy--robs life of its romance and its savor. For the implicit theme of both books is the bureaucratization of the Supreme Court. And this bureaucratization seems not to have improved the Court's performance, something that would have left Weber scratching his head in puzzlement.

Supreme Court justices were first authorized "legal secretaries" late in the nineteenth century, and they really were secretaries. Their evolution into law clerks was gradual. By the 1930s, however, all the justices had a law clerk in the modern sense of a law school graduate furnishing professional rather than secretarial assistance. The law clerks in that era were mainly research assistants and sounding boards, though they might also summarize petitions for certiorari for their justices. Virtually all opinion-writing was done by the justices themselves. Most of the clerks were fresh out of law school, and generally there were only a few applicants for each clerkship. Justices did not encourage applications, and often delegated the picking of their clerks to friends or law-professor acquaintances.

Today each justice has four law clerks (though the chief justice is authorized five), each of whom has had a year of clerking on a lower court under his or her belt and sometimes more extensive professional experience. Sometimes one of the clerks is designated the justice's chief clerk to assist the justice in supervising the others. Judging by my own experience as a Supreme Court clerk, clerks work longer hours today than they did forty years ago, even though the caseload per clerk is much lower. The hypertrophy of staff is a general characteristic of modern American government. It has the aspect of an arms race. If one justice has a large and able staff, other justices may feel placed on the defensive and want to catch up, even though the principal effect of the larger staffs may merely be longer and superficially more erudite opinions.

The clerkship positions have become highly coveted. There are many more, and many abler, law school graduates than there were in the 1930s, and they are falling over each other to become Supreme Court clerks. No one is hired without an interview, in advance of which the applicant studies the justice's opinions and seeks tips from the justice's former clerks on how to make a good impression. The applicants, much more carefully vetted than in the old days, are genuinely an elite, who after serving their clerkships are offered huge signing bonuses by leading law firms.

Clerks draft memoranda circulated to the entire Court advising on whether to grant petitions for certiorari. Often they write bench memoranda, which can run to fifty pages or more, for their justice in advance of oral argument of the handful of cases in which the justices grant plenary review. Except for Justice John Paul Stevens, who writes his own first drafts of opinions, law clerks write the first drafts of their justices' opinions. (According to Courtiers, Stevens's clerks rewrite his drafts extensively, thus producing an inversion of the normal relation of clerk-author to justice-editor. In another inversion, Justice Harry Blackmun, a genuine eccentric, left the opinion-writing to his clerks after his first years on the Court and concentrated on cite-checking their drafts. He was by all accounts an awesome cite-checker.) Some justices rewrite the clerks' opinion drafts extensively, others little. Sorcerers' Apprentices estimates that 30 percent of the opinions published by the Supreme Court are almost entirely the work of the law clerks; and as they are the primary drafters of most of the other opinions as well, probably more than half the written output of the Court is clerk-authored.

The clerks have the use of excellent computerized legal research tools, and their efficiency is further boosted by the device of the "cert. pool." Until the early 1970s, the law clerks prepared certiorari memos for their individual justices, a time-consuming task. A law clerk whose justice belongs to the pool (Stevens is the lone holdout) writes a memo for all the justices in the pool. So instead of nine memos being written in each case, there are only two--the pool memo and a memo written by one of Stevens's clerks. The pool memos tend to be longer and more careful than a memo for a single justice, because they are written for a larger audience and one that the law clerk does not know as well, but Sorcerers' Apprentices presents convincing statistical evidence that the aggregate time that law clerks spend on certiorari memos has fallen considerably because of the reduction in the number of memos written by each clerk.

Now here is the remarkable point, which is not to be found in either book (perhaps the authors did not wish to be regarded as Court-bashers): despite the substantial quality-adjusted increase in the Court's principal staff, the Court's decisional output has not increased. This may seem to contradict the basic laws of economics. It is true that the number of petitions for certiorari (and other applications for review) has grown substantially since the 1930s (from about 900 a year to about 7,500), and each one has to be decided. But most of the growth has been in petitions filed by indigents, mainly prisoners, who do not have lawyers, and most of these petitions are frivolous. The "paid" docket has grown from about 800 petitions a year to about 1,700, a mere doubling that the cert. pool alone could have comfortably accommodated with no increase in the number of clerks.

Owing to the complex jurisprudence of capital punishment that the Supreme Court has developed, motions for stay of execution have doubtless increased in number and in length since the 1930s, even though far fewer people are being executed. But the increase in the Court's dispositions of petitions for certiorari and motions for stay has been offset by the steep reduction in the number of plenary decisions (that is, decisions in cases in which the Supreme Court has granted review): from an average of 183 in the 1930s to an average of 84 in the 2000s. The average opinion is longer today, and there are more dissenting and concurring opinions, and so the total word output of the Court is greater; but whether this spells improvement may be doubted.

But aren't the cases more difficult today? I don't think so. Although American law is more complex than it was in the 1930s, this is primarily because of developments in areas of law--securities regulation, antitrust, taxation, federal pension law, intellectual property in high-technology areas such as computer software and pharmaceutical drugs, and complex financial and other commercial transactions--that the Court tends to shy away from in favor of constitutional and criminal cases. (This aversion may reflect the tastes and the aptitudes of the law clerks, whose certiorari memos influence the justices' decisions to grant or deny review.) And constitutional cases are often "difficult" in the sense of eluding satisfactory resolution. They are not inherently more complex. Instead they are indeterminate, as a consequence of the vagueness and antiquity of the constitutional text, and of the emotionality and political sensitivity that so many constitutional cases engender.

Of course, to measure output just in quantitative terms is seriously incomplete. One must also adjust for quality. This is extremely difficult to do in the case of Supreme Court decisions, and neither of these books attempts to do it. To ask whether the fraction of "correct" decisions is greater today than in the 1930s would be fatuous, because there is no benchmark for determining the correctness of a Supreme Court decision. Comparison over time is particularly difficult, because issues are in constant flux. Still, one can apply quality-related criteria, such as clarity, brevity, guidance provided to the lower courts, and candor in explaining the true grounds of decision, to the opinions in the two eras.

When one does this, one is not likely to find a dramatic, or perhaps any, overall difference in quality. Today's opinions are longer--a dubious virtue. There are more separate opinions, most of which are ephemeral. Today's opinions are more polished, more "scholarly," and more carefully cite-checked, but these are modest virtues. Neither judges nor their clerks are scholars. The scholarly apparatus of judicial opinions belongs to the rhetoric rather than the substance of judicial decision-making.

Could it be that the justices were abler in the 1930s than they are today, and so they needed less staff? There were some very distinguished justices back then, notably Brandeis, Cardozo, Hughes, and to a slightly lesser degree Stone. (Holmes and Frankfurter served only briefly during that decade.) No current Supreme Court justice has achieved the stature of a Brandeis, but the average quality of the justices is no lower today than it was then. Several of today's justices have very strong academic and professional backgrounds.

Although today's Supreme Court opinions are no more poorly written on average than opinions from the era in which the justices wrote their own opinions, there is nonetheless a loss when opinions are ghostwritten. Most of the law clerks are very bright, but they are inexperienced; and judges fool themselves when they think that by careful editing they can make a judicial opinion their own. If an editor changes thirty words on a page that contains three hundred words, the page will look heavily edited because handwritten changes are larger than type, but only ten percent of the words will have been changed, and many of the changes will be inconsequential. It is also possible that the cert. pool, which reduces the diversity of views presented to the justices concerning whether to grant or to deny review in a particular case, is distorting the justices' choices of which cases to hear.

In sum, a large quality-adjusted increase in inputs, having mainly to do with the increased quantity, quality, and efficiency of the law clerks, seems not have yielded a quality-adjusted increase in output. I suggested earlier that such a result contradicts fundamental laws of economics, but of course I was thinking of business firms, which will cut back if they discover that they are adding costs without generating commensurate increases in profits. The Supreme Court is not a business, and government agencies are not subject to market discipline. Congress controls the Supreme Court's budget, but respect for the Court as a co-equal branch of government (with Congress and the president) has assured generous funding for it; and anyway the Court's budget is such a small fraction of the overall $2.7 trillion federal budget that if the justices want more law clerks or more up-to-date computer software, Congress will not begrudge them these things. It will not insist that the justices "manage by the numbers" and thus demonstrate that an additional expense will add greater value to the justices' work product.

The question remains why the Court's inputs have grown. Sorcerers' Apprentices argues plausibly that the increased number and quality of the law clerks are mainly the unintended consequences of administrative measures beginning with Chief Justice Hughes's innovation of the "dead list." Until then, all petitions for certiorari were discussed at the Court's weekly conference. The chief justice would first summarize the petition and give his view on whether it should be granted, and the other justices would respond and then vote to grant or deny. When the chief justice decided that it would promote efficiency for him to circulate a list of cases that he thought should be denied without discussion at the conference, the other justices, deprived of his summary, had to evaluate the dead-listed cases on their own, and they turned to their clerks. The clerks thus ceased to be merely research assistants.

Clerk "networking," which enhanced the clerks' individual and collective influence, began when the justices moved into their new building, completed in 1935, in which each justice and his law clerk had offices. Until then, the justices had worked out of their homes and provided space there for the clerk to work, and so the clerks were dispersed and had little contact with one another. In 1947, in another change that enlarged the clerks' role, Chief Justice Vinson decided that each justice should be assigned the same number of majority opinions to write. Until then, faster justices had been given more assignments. The slower justices, now more heavily burdened with writing assignments, began delegating opinion-drafting to their law clerks. They also wanted an additional clerk, and Congress obliged. Finally, with the advent of the cert. pool in 1972, the law clerks began spending less time on petitions for certiorari, and this freed up their time to do more opinion-writing.

As the number and the experience of law clerks rose (increased competition for clerkships was one of the things that made previous clerkship experience a prerequisite--the justices could be pickier when they were fishing from a larger pool, and could thus insist on more experienced applicants), the clerks' ability to write good opinions increased. The justices began to conceive their own role as that of editor, not author. Even experienced writers find writing a first draft burdensome, and so with a competent ghostwriter at hand the temptation to delegate that drafting can become irresistible. The more clerks a judge has, moreover, the more time he must devote to recruitment, orientation, supervision, reading memos and other materials produced by the law clerks, and talking to them, and therefore the less time he has for writing opinions.

Another consequence of the justices' burgeoning staffs has been an increased concern, amounting to an obsession, with secrecy. I do not remember from my time as a clerk for Justice William Brennan anyone saying anything about confidentiality, though it was understood that one was not to gossip about the justices or discuss pending cases with outsiders. Now there is an elaborate code of conduct for clerks, enjoining them to utmost confidentiality, as if they were handling national defense secrets.

But leaks are a function mainly of the number of the secret-bearers, and cannot be stifled by rules. The effect of the rules is merely to delay the leaks. The clerks delayed blabbing about the internal machinations in Bush v. Gore for several years, but blab they eventually did. And the justices' official papers are increasingly voluminous (elaborate documentation is one of the defining characteristics of bureaucracy) and increasingly available to scholars, which works against the institutional desire to zip the law clerks' lips. Sorcerers' Apprentices makes especially good use of Justice Lewis Powell's extensive papers. Powell modeled his judicial role on that of the senior partner of a small law firm. It seems that rather than write judicial opinions, he wrote memoranda instructing, complimenting, and "mentoring" his law clerks. Chief Justice William Rehnquist was candid in acknowledging publicly the very large role played by his law clerks in the opinion-writing process.

Sorcerers' Apprentices endorses a proposal that has no chance of being adopted, though it seems as sensible as it is provocative. It is that the Supreme Court publish the pool cert. memos. There is no published statement of the reasons for denials of certiorari, and this deprives the bar of valuable information concerning the Court's work. The reasons given by a law clerk for wanting to deny certiorari in a case will sometimes differ from the reasons that persuade six justices to deny (six because only four votes are required to grant it). But there would be fewer futile petitions--the Court grants only a little more than 1 percent of the petitions--if the bar had a better sense of the reasons that the Court's key staffers give for turning petitions down. There might also be fewer denials.

The Court's preoccupation with the confidentiality of its internal workings makes an illuminating contrast with the English judicial tradition (now in rapid decline because of caseload pressures) of "orality." Everything English judges did was to be done in public, so that their performance could be monitored. They did not deliberate, they had no staff, they did not have libraries, they did not read briefs: on the bench they read the cases, the statutes, and the other materials that the lawyers handed up to them. (So appeals might take days to argue, which is why the tradition has eroded.) Our Supreme Court (imitated in this by most other American courts) has gone to the opposite extreme, imposing--or attempting with mixed success to impose--a regime of secrecy on the judicial decision-making process.

The standard argument is that secrecy is necessary for candid communications among the justices and between the justices and their clerks. But judicial decisions, unlike business and political and military decisions, are supposed to be based, and rightly so, on reasons that can be stated publicly without embarrassment. Sorcerers' Apprentices quotes some disreputably partisan clerks' memos in "hot" cases, for example involving abortion. If publicity deterred clerks from writing such memos, the nation would not be the loser.

Of course clerks must not leak information about cases not yet decided; but this is the only secret-keeping that should be required of them. As for deliberations among the justices, they are by all reports stilted and brief. Were transcripts of their deliberations to be published, the only embarrassment would be that the American public would realize that the Supreme Court of the United States is at bottom merely a committee of able lawyers. A bit more of the mystique of judging would be chipped away. It would be a further step toward the disenchantment of the world that Weber foresaw with prescience and regret, but it might not be a bad thing.

Richard A. Posner is a judge on the U.S. Court of Appeals for the Seventh Circuit and a senior lecturer at the University of Chicago Law School.

Copyright 2006 The New Republic, LLC