Read America's Unwritten Constitution: The Precedents and Principles We Live By Online
Authors: Akhil Reed Amar
Under this system, Richard Nixon’s administration was obliged to hire a renowned Harvard law professor, Archibald Cox, to conduct a credible outside investigation of alleged administration wrongdoing. When Cox
pushed very hard very fast, Nixon had Cox fired, in the so-called Saturday Night Massacre. But Nixon paid an enormous political price and was politically obliged to replace Cox with another respected outsider, Leon Jaworski, who completed the job and indeed brought down the administration. Had Nixon tried to fire Jaworski or to pardon the targets of investigation in an effort to oblige Jaworski to close up shop, Nixon would have been quickly impeached and removed from office in exactly the way that the framers sketched the system on the drawing board, with Congress—not a judicial panel—making the key decisions.
Although this Watergate system of political independence offered less formal legal independence than the 1978 statute that arose to replace it, it actually worked better, and did so precisely because it did not stretch the roles and rules laid down in the terse text. Viewed from one angle, Archibald Cox and Ken Starr were almost identical twins. Cox was a well-respected Democrat who had served as a distinguished solicitor general to the man who had run against Nixon—John F. Kennedy. Starr, in turn, was a well-respected Republican who had served as a distinguished solicitor general to the man who had run against Clinton—George H.W. Bush. Starr’s appointment was thus perfect poetic justice. Why, then, was Clinton much more successful in discrediting Starr than Nixon had been in his attempt to demonize Cox?
*
One big reason was that Starr started off with less credibility than Cox precisely because Clinton himself had not picked Starr. In addition, Starr’s appointment had arguably involved judges doing nonjudicial things in nonjudicial fashion. (Recall, for example, the controversial lunch between Judge Sentelle and leading anti-Clinton senators.) Had Clinton, like Nixon, been politically obliged to name an outside lawyer such as Starr, the outside lawyer would have had special political authority from the start, having been chosen not by a clump of mostly Republican-appointed judges, but by the Democrat president himself. After completing his service as independent counsel, Starr echoed Scalia and Reno in criticizing the basic structure of the 1978 act.
The complete collapse of the statutory independent counsel system
should teach us that modern reformers ignore the written Constitution at their peril. Innovations that work within or work around the document’s formal rules survive. Innovations that run roughshod over these rules do not. A Watergate-style system of special prosecutors has worked before and can work again precisely because nothing in this political improvisation violates the Constitution’s text—or, more precisely, because this improvisation nicely meshes with the written document’s schema of institutions and incentives. By contrast, the 1978 system of legally independent counsels failed precisely because it dishonored the proper written roles of each of the three branches by placing too little reliance on congressional oversight and impeachment; putting too much confidence in judges, even as it obliged them to do nonjudicial things; and paying no heed to how presidents may usually control prosecutors via the pardon pen.
HAVING JUST SEEN A COUPLE OF FAILED
twentieth-century institutional improvisations, let’s conclude with a couple of modern success stories.
One clever, albeit highly technical, separation-of-powers gadget is known to beltway insiders as the
Saxbe fix
. Here’s how it works: Under Article I, section 6, “[n]o Senator or Representative shall, during the Time for which he was elected, be appointed to any civil Office under the Authority of the United States…the Emoluments whereof shall have been encreased during such time.” The obvious aim of this anticorruption provision is to prevent members of Congress from improperly inflating the salaries of executive or judicial offices and then benefiting personally by resigning from Congress and being appointed to those overpaid offices. The strict letter of the rule could be read to disqualify any member of Congress from any executive or judicial office for which the salary was increased during the member’s current term. But the spirit of the clause is satisfied by a more sensible, if less literalistic, approach—the Saxbe fix—that allows the appointment so long as the appointee receives only the old (pre-increase) salary and thus does not pocket any salary increase that may have been recently adopted.
For example, if the salary for a given cabinet office swells from $100,000 to $105,000 on a senator’s watch, and the president thereafter wants to appoint that senator to this office, a literalist might say the appointment would be irremediably illegal (since the salary was indeed upped on the
senator’s watch). But under the less literalistic Saxbe-fix approach, the appointment would be proper so long as the salary is reduced—“Saxbe fixed”—back to $100,000 before the senator takes office.
55
While the letter and the spirit of the emoluments clause arguably tug in opposite directions on this nice question, actual practice—approved by all three branches and both major political parties over a long stretch of time—sensibly breaks this interpretational tie. No justice ever refused to sit alongside ex-senator Hugo Black, who would have been ineligible for appointment to the Court under a hyper-literalist reading. Also, for more than a century, presidents and senators of both parties have continued to appoint and confirm resigned or resigning members of Congress to cabinet positions for which salaries had recently been increased, so long as the new appointee would not receive the increase. Indeed, outgoing presidents have repeatedly signed on to statutory “Saxbe fixes” aimed at accommodating the cabinet preferences of incoming presidents—even presidents of the other party. Cabinet members appointed under this approach include Treasury Secretary Lot Morrill in the Grant administration, Secretary of State Philander Knox in the Taft administration, Attorney General William Saxbe in the Nixon administration (whence the phrase, “Saxbe fix”), Secretary of State Edmund Muskie in the Carter administration, Treasury Secretary Lloyd Bentsen in the Clinton administration, and Secretary of State Hillary Clinton and Interior Secretary Ken Salazar in the Obama administration.
On reflection, the general acceptance of the Saxbe fix makes perfect sense precisely because this particular institutional improvisation rests on a plausible—albeit not incontestable—reading of the terse text. To see this clearly, let’s imagine a hypothetical Senator Smith elected to serve a six-year term beginning at Time T1. Early in his term (at Time T2), Congress increases the attorney general’s salary from $100,000 to $105,000. (It matters not whether Smith voted for or against this increase—the relevant constitutional rule in no way hinges on this fact.) Still later (at Time T3), the president makes clear that he would like to name Smith as the next attorney general. Congress then (at Time T4) passes a “Saxbe-fix” statute restoring the AG salary to $100,000. (It matters not how Smith votes on this, or, indeed, whether he is still in the Senate or has already resigned in anticipation of his executive service.) The Senate then confirms Smith at
Time T5, and Smith, after resigning from the Senate (if he has not already done so), assumes his $100,000 office at Time T6. A replacement senator, Jones, is named to fill out the remainder of Smith’s Senate term, which ends exactly six years after T1. Call this end date T7. Did Smith’s appointment violate the emoluments clause?
Recall the relevant words: “No Senator…shall, during the Time for which he was elected, be appointed to any civil Office under the Authority of the United States…the Emoluments whereof shall have been encreased during such time.” A hyper-literalist might say that Smith’s appointment was indeed unconstitutional. The AG’s “emoluments” (i.e., salary) were indeed “encreased” at T2—and the fact that these emoluments were later decreased at T4 cannot change what did in fact occur at T2. And the $5,000 increase did occur on Smith’s watch—“during the Time for which [Smith] was elected.” The obvious counterargument is that the AG’s salary really had not “encreased”
on balance
and
for Smith
between T1 and T6—a characterization that captures the two legally relevant moments in time and properly focuses on the “Senator” himself in keeping with the letter and logic of the clause. The salary was $100,000 both when Smith started his Senate term (T1) and when he entered office (T6). This is a perfectly sensible way to understand the word “encreased”—especially once we understand the need to read the Constitution not literally but faithfully.
Even if we did not have the benefit of
Chapter 1’s
extended case study of vice-presidential impeachment, the emoluments clause itself makes clear that constitutional words must be read sensitively and in context, with reference to their obvious spirit and purpose. First, what does the opening phrase “during the Time for which he was elected” mean? Suppose at Time T1 Smith is beginning his
second
term, and that he was in fact first elected to serve a term beginning six years before T1. If the AG’s salary had been increased in that
first
term—from, say, $96,000 to $100,000—why couldn’t it be said that Smith was ineligible to be appointed even during the interval between T1 and T2? True, before T2, Congress had not “encreased” the AG’s “emoluments” during Smith’s
second
term. But these “emoluments” had increased “during the Time for which [Smith] was elected,” if that phrase is read in a literalistic, flatfooted way. The clause is not sensibly read that way, however, because this reading does not make good common sense or structural sense and because the clause can be construed more sensibly.
(The reason why this flatfooted reading makes no sense is that it would disqualify Smith between T1 and T2, but in this very same time period it would not disqualify ex-senator Smythe, who served alongside Smith in Smith’s first term, and who then left the Senate at T1. What sense does it make to treat Smith worse than Smythe? Indeed, Smythe may have voted for the $4,000 increase, whereas Smith may have voted against it. And before T2, nothing in Smith’s second term has happened in Congress that seems relevant to the emoluments clause.)
Now consider the emolument clause’s final phrase: “during such time.” During what time? Under a literal reading, “such time” obviously refers to earlier language, namely, “during the Time for which [Smith] was elected.” In Smith’s case, this first “during” phrase clearly covers the precise six-year period between T1 and T7. But upon reflection, it cannot be right that the final “during” phrase means the same thing as the opening “during” phrase. Suppose Congress had never raised the AG’s salary at Time T2 in Smith’s second term. If so, there would have been no problem whatsoever with his appointment at T6—and no need for any sort of Saxbe fix at Time T4. But what if Congress
later
increases the AG’s salary on Jones’s watch—that is, sometime after T6 but before T7? Surely this increase does not somehow retroactively oust Smith from office. Even though the first “during” phrase covers the entire period from T1 to T7, the second phrase only covers the period until Smith’s appointment—T1 to T6. The closing phrase “during such time” cannot sensibly be read to mean the same thing as the opening phrase “during the Time for which [Smith] was elected,” even though this might at first seem to be the literal meaning.
Just as other phrases in the emoluments clause—the opening “during” phrase and the closing “during” phrase—must be read with reference to their purpose and spirit, so, too, must the word “encreased” be construed functionally. The Saxbe fix is thus a highly plausible gloss on a genuinely ambiguous text—a classic illustration of how America’s written and unwritten Constitutions generally cohere.
CONSIDER, FINALLY, THE ROLE OF
various
independent agencies
that have been created over the past century, such as the Federal Trade Commission, the National Labor Relations Board, the Federal Reserve Board, and the Consumer Products Safety Commission. All told, several dozen
such agencies currently exist, making up a substantial portion of the federal government’s regulatory apparatus. Many casual observers and even some scholars who should know better have suggested that the very existence of these agencies proves that real institutional practice in America broke free from the written Constitution long ago, and remains as free as ever today. A close look at both the text and the practice suggests otherwise.
The very label “independent agency” can be read in different ways, and some readings lead only to confusion. “Independent” agencies are of course not independent of the Constitution itself. Nor are they independent of the document’s tripartite scheme. Constitutionally speaking, they are executive-branch agencies of a certain sort.
True, some of these agencies perform multiple functions—promulgating rules of conduct (as does a legislature), enforcing civil laws and prosecuting violations of criminal statutes (in classic executive fashion), and also performing adjudicatory tasks between government and individuals and sometimes even between private parties (much like a court). But this fact does not suffice to relegate these agencies to some counter-constitutional “fourth branch” outside the written Constitution’s three-branch structure. Rather, this mixture of functions places “independent agencies” squarely within the second branch—the executive branch, a branch that has always performed a wide range of tasks. Interstitial rule-making within the bounds of a vague or ambiguous statute is a common executive function, as is applying law in the first instance to specific facts involving specific persons.
The label of independence may also mislead some into thinking that actual agencies either freely float between the Congress and the president or can be statutorily sited anywhere along the continuum between legislature and executive. In fact, these agencies conform to a strict pattern.