Yet Clinton’s impeachment by the House followed by the Senate’s unwillingness to remove him is one of four cases, each involving impeachment and acquittal, that can fairly be called epic confrontations, both politically and constitutionally. In the details of these four cases — Supreme Court Justice Samuel Chase in 1804, Judge James Hawkins Peck in 1830, President Andrew Johnson in 1868, and President Clinton in 1999 — lies a tale of lasting significance broader even than the tumultuous issues that came out as these impeachments unfolded.

In these four spectacular clashes, a fascinating pattern presents itself. It is the story of how resort to the Constitution’s ultimate sanction became inextricably entangled with one or another law that was itself fundamentally suspect constitutionally. These laws amounted to grave extra-constitutional disturbances to a carefully wrought constitutional system based on the separation of legislative, executive, and judicial powers. It was these disturbances around which sentiment for removal gathered in the first place — only to dissipate in the end.

It is, of course, possible to answer this question at various levels of abstraction, and accordingly to take his case as a window of one sort or another on the condition of American politics, culture, society, and the rest. There should be little doubt that Clinton’s impeachment and acquittal will long stand as a major feature of the American political landscape. Aspects of the events of 1998-99 will long be adduced as evidence in theories about our times and mores. But short of this abstract inquiry, highly subject to disagreement as it is, there are a few specific aspects of the Clinton affair that most people, perhaps, would readily agree were necessary conditions for impeachment.

First, the president was unwilling to restrain his own conduct at a time when the ongoing Paula Jones case put him at risk of exposure of the relationship, and he was willing to be deceitful in court about it. Obviously, if Clinton had walked away from Monica Lewinsky’s overtures, none of what happened next would have happened. Not only the president’s opponents, but also many of his allies, were quite clear in their view that the president recklessly and shortsightedly failed to conduct himself in a manner consistent with his office. He himself was responsible for that improper conduct. There was no deflecting responsibility to others.

Had the Jones case been settled earlier, before Clinton was called to answer about other women with whom he might have been involved, there would of course have been a political cost to the president — both for settling and in the event that information about recent sexual encounters became public — but no impeachment. Similarly, had the Supreme Court ruled that the president is constitutionally immune from civil suit while he is in office, as Clinton’s lawyers urged and the court unanimously rejected, or had Congress passed a statute creating such limited immunity, there might once again have been embarrassment in the event of disclosure, but there would likewise have been no impeachment.

And it seems unlikely that, if Clinton had told the whole truth in the Jones case, the ensuing sex scandal would have led to his impeachment. Although the president’s defenders argued otherwise, most of those who favored impeachment insisted that the sexual relationship itself was not the issue, but rather lying under oath and obstruction of justice. Given the narrowness of the eventual vote in favor of the two articles, it seems unlikely that a mere "morals" charge, in the absence of conduct arguably criminal, could have gained a majority vote in the House.

The second necessary condition was the control of the House of Representatives by the president’s political opposition. This point hardly needs belaboring. When the House voted December 19, 1998, approving articles charging the president with lying to a grand jury and obstruction by votes of 228-206 and 221-212, no more than five Democrats voted in favor. There is little reason to think that a Democratic House of Representatives would have voted even to begin an impeachment inquiry. It is possible, given the statements made by Democrats, that the House and then perhaps the Senate would have sought to pass some resolution of censure against Mr. Clinton for his conduct, whether out of the heartfelt conviction that what he did was wrong, as many Democrats professed, or out of a perceived political need to offer a response. But it’s hard to imagine more than that.

If these were necessary conditions for impeachment — the president’s own conduct in the context of the Jones case and the control of the body with impeachment power by his opposition — were they also sufficient conditions?

In the light of the huge role in the impeachment process played by independent counsel Kenneth Starr, it is difficult to say that the two preceding conditions would have been enough. It was Starr’s office that, upon obtaining evidence of perjury and obstruction in the Jones case, began an investigation. Nor was this solely a criminal investigation, designed to lead exclusively to a decision on whether or not to prosecute Clinton at some point, perhaps after he left office. The law requires the independent counsel to forward to Congress evidence of impeachable offenses by the president or other officials. If not as a constitutional matter, then certainly by statutory authority, Starr’s office carried the impeachment process forward — by gathering, over the course of nine months, evidence about the president’s actions; by discussing the investigation with the media, in a fashion Starr’s office believed consistent (but the president’s lawyers believed inconsistent) with the requirements of grand jury secrecy; then by forwarding to the House a report detailing that evidence and listing the independent counsel’s view of the impeachable conduct; then by the testimony of Starr himself as a witness before the House Judiciary Committee.

Moreover, at numerous points in the course of the Starr investigation of the Lewinsky matter, members of the Republican congressional majority, including the House leadership and many members of the House Judiciary Committee, preeminently Chairman Henry Hyde, made a point of deferring to Starr’s investigation. Hyde took it as his duty to let the Starr inquiry run its course before acting on the allegations of perjury and obstruction being bandied about. A July 30, 1998, statement was typical: "I don’t want to hurry or press or push the independent counsel. I don’t want it to appear that we’re driving his inquiry." Other members of both parties, and for different reasons, frequently said in response to questions about what the president’s fate should be that they were refraining from making any judgments until the independent counsel had finished his work. Whether this was sincere open-mindedness or simply a way of evading the question, these members of the House majority themselves placed Starr at the center of the process. And certainly the president’s staunchest defenders placed Starr at the center — as an out-of-control prosecutor obsessively pursuing the president for partisan and perhaps puritan reasons.

The Constitution vests the House of Representatives with "the sole power of impeachment," and one might say that until the House voted to undertake an impeachment inquiry, no such proceedings were under way. But this is surely an exercise in hair-splitting. The independent counsel statute that Congress itself passed and that the president himself signed explicitly assigned the independent counsel a fact-finding role that could lead to impeachment proceedings. The law requires that the independent counsel advise the House "of any substantial and credible information which the independent counsel receives, in carrying out the independent counsel’s responsibilities under this chapter, that may constitute grounds for an impeachment." In the relevant period, Congress, the White House, and anyone paying the slightest attention understood that Starr was the central player and awaited his report. It is absurd to contend that the arrival of the report in Congress marked the beginning of a process, rather than some point well advanced. And finally, the House Judiciary Committee itself legitimized the centrality of Starr’s role by conducting very little fact-finding of its own, instead basing its proceedings on the evidence he gathered. So firm was the committee’s embrace of Starr’s material that the majority did not feel a need to try to resolve the discrepancies Starr presented in various witnesses’ accounts of certain events; the committee majority felt, in Hyde’s view, that Starr’s fact-finding provided more than enough evidence to proceed with impeachment articles. "We had 60,000 pages of testimony from the grand jury, from depositions, from statements under oath," Hyde said. "That’s testimony you can believe and accept." The search for certitude in murky areas was a task that could be left to a Senate trial.

We might try to speculate about what would have happened had there been no independent counsel investigation under way at the time Linda Tripp met with Paula Jones’s lawyers and led them to questions that might ensnare the president if he answered untruthfully. Tripp professed, at one point, that her reason for taping her conversations with Monica Lewinsky was self-protection: She feared being called upon to commit perjury. Yet it seems clear she also intended to inflict as much political damage on the president as possible. In the absence of an independent counsel to whom to take her information, she might have delivered it to the Justice Department, or to the Jones case’s Judge Wright, or to Jones’s lawyers, or to the Judiciary Committee, or to the media. Regardless of which course she chose, it seems likely that the information would have become public, probably by her own agency — at least to judge by the behind-the-scenes actions of Tripp and her confidante, Lucianne Goldberg, described by Newsweek’s Michael Isikoff in his account of how he came to break the story, Uncovering Clinton: A Reporter’s Story (1999).

Then what? No one can say with certainty, of course. But surely there would have been a substantial outcry. The conservative editorial pages would have demanded an investigation. The president’s detractors would have had a field day. The salacious details of the story would have outed. Judge Wright, one assumes, would have viewed Clinton’s deception just as seriously and unfavorably. But what about the road to impeachment?

In order to conclude that Starr’s role wasn’t also a necessary condition for impeachment, we must imagine the House Judiciary Committee, on the say-so of its chairman, deciding solely on the basis of the Tripp accusations to look into the matter, notwithstanding the president’s denial. It seems reasonable to assume that Lewinsky would not volunteer further information. At some point, to proceed with the fact-gathering it would be necessary to compel her testimony.

What would Democrats on the Judiciary Committee have made of this? What about the president’s other defenders? They would surely not have idly acquiesced. Recall, too, that when the story broke, its main focus was the independent counsel investigation into perjury by the president in the Jones case. In the absence of such an investigation, and given only an investigation (or merely the possibility of one) by the Republican-controlled Judiciary Committee — without, in short, Starr’s official quasi-criminal, quasi-impeachment inquiry into the president’s actions — Clinton’s defenders would surely have had even better success portraying the matter as base snooping into private conduct motivated by mere partisanship.

The House leadership would presumably have had some sway over the course of events as well. Subsequent revelations about House Speaker Newt Gingrich’s own extramarital affair at the time give one cause to wonder whether, without an independent counsel to defer to, he would have supported an inquiry by the House whose first question would be whether or not Clinton had a sexual relationship with Lewinsky. In the event, Gingrich told the Washington Post in August 1998 that "a single human mistake" would not constitute grounds for an impeachment inquiry. "I don’t think the Congress could move forward only on Lewinsky, unless [Starr] had such a clear case, such an overpowering case." He also said the matter was "not about scandals in the gossipy sense or sexual behavior in the gossipy sense. It’s about whether or not the law has been violated, and if so, is it a pattern of violation [or] is it a one-time event." The speaker’s words here can perhaps be seen now to have had a greater sense of personal urgency than was apparent at the time.

In addition, let us not forget the legal mandate of the independent counsel: to advise the House of "substantial and credible information . . . that may constitute grounds for an impeachment." The Constitution calls for removal "on impeachment for, and conviction of, treason, bribery, or other high crimes and misdemeanors." Which conduct, precisely, falls within the compass of "high crimes and misdemeanors" has been disputed by scholars, politicians, judges, and the public since the debate over the ratification of the Constitution itself. The statute here gives the independent counsel the first judgment on the subject of what might constitute an impeachable offense. In the absence of an independent counsel to make a determination on what might or might not be considered impeachable, the Judiciary Committee majority would have to make that decision itself. To be sure, the Judiciary Committee did not accept all of Starr’s charges, and the House accepted fewer still. But it was Starr who, in effect, had the first opportunity to define the scope of impeachability. He presented a lengthy document and numerous boxes of evidence to the House, to which the Judiciary Committee would have to react. Would the Judiciary Committee have gone as far as Starr on its own?

Finally, as it happened, the full House rejected an impeachment article approved by the committee charging the president with lying under oath in the Jones case. The perjury-related article approved by the House concerned lying before Starr’s grand jury. In the absence of the president’s appearance there, House members would have been forced to confront head-on whatever underlay their greater reluctance to recommend Clinton’s removal for the conduct that started all of the trouble in the first place. Perhaps Judge Wright’s ire, coming not after impeachment and acquittal, as it did, but at her own pace, would have galvanized the Judiciary Committee or the House as a whole into action it would not otherwise have taken; on the other hand, one might as plausibly speculate that many members would have been perfectly content to let her contempt ruling and the fine she imposed on Clinton stand as sufficient rebuke to his conduct.

Perhaps a House vote to impeach the president could have resulted even in the absence of the independent counsel investigation. Obviously, the answer is unknowable. But the obstacles in the way of such an outcome look formidable indeed. If the president’s own misconduct in the context of the Jones case and the control of Congress by his political opponents were necessary conditions for impeachment, the independent counsel investigation, which was undeniably central to the process as it actually unfolded, looks to have been no less necessary.

In addition to the constitutional affront to the separation of powers represented by the office of independent counsel, critics pointed to numerous other problems with the law: Unlike an ordinary federal prosecutor, an independent counsel has a limitless amount of time and unconstrained resources to pursue wrongdoing on the part of a particular individual or group of people. Ordinary prosecutors begin with evidence of a crime and then try to identify its perpetrators. An independent counsel investigation can be triggered merely on the basis of "specific" information, requiring "further investigation," "from a credible source" about criminal conduct by a person the law covers — conditions easily met. Once appointed, the prosecutor, with the supervising court’s approval, can broaden the inquiry more or less at will into any and all allegations against the subject. Ordinary prosecutors face competing demands for their attention, and therefore must apply real-world standards of discrimination on the question of how worthy of pursuit a particular case is. Independent counsels are faced with almost an opposite pressure to be thorough — to give up on no avenue of investigation until it has been explored, no matter how unlikely the prosecutor thinks it is to yield anything of consequence.

Theodore B. Olson, a Reagan Justice Department official and the target (eventually completely exonerated) of an independent counsel inquiry, mounted a challenge to the constitutionality of the law under which he was being investigated. In 1988, the D.C. Circuit Court of Appeals ruled in his favor. The opinion, written by Judge Laurence H. Silberman, was witheringly thorough in detailing the defects of the law and its problematic constitutional character with regard to the separation of powers. The Supreme Court, however, disagreed. In a 7-1 ruling in Morrison v. Olson (1988), with Justice Antonin Scalia the sole dissenter, the court held, in effect, that there were sufficient safeguards for the executive branch in the law to avoid constitutional difficulties. Chief Justice Rehnquist, who would later preside over Clinton’s Senate trial, delivered the opinion of the court. Scalia reasserted many of the points Silberman had made, and added some more besides, but the law stood vindicated by the high court.

Conservatives were unpersuaded, and they found in Iran-contra independent counsel Lawrence Walsh the living embodiment of the law’s defects. Liberals, needless to say, saw matters differently. They defended Walsh, focusing instead on the gravity of the alleged crimes and the constitutional effrontery of Reagan administration officials. Neither side gave any ground through the effective end of the Iran-contra investigation at the conclusion of the Bush administration, with the outgoing president’s pardon of six Reagan-era officials.

The law expired in 1992. Its reenactment in 1994 had the support of the Clinton administration and bipartisan support on Capitol Hill. Attorney General Janet Reno testified to its importance before the Senate Governmental Affairs Committee in May 1993: "While there are legitimate concerns about the costs and burdens associated with the act, I have concluded that these are far, far outweighed by the need for the act and the public confidence it fosters." In the course of five years, however, the administration changed its mind. In March 1999, shortly before the statute was due to expire, Deputy Attorney General Eric H. Holder Jr. appeared before Congress attesting to the administration’s new view: "The experience of the Department over these last five years has been enlightening. It takes a close-up view of the operation of the Independent Counsel Act to understand that it has serious flaws. The Department of Justice has reluctantly come to the conclusion that the structural flaws we have identified here cannot be fixed."

In a stunning appearance on Capitol Hill in April 1999, Kenneth Starr himself announced his opposition to the law under which he had been investigating the Clintons, culminating in his report to the House. "By its very existence, the act promises us that corruption in high places will be reliably monitored, investigated, exposed, and prosecuted, through a process fully insulated from political winds," Starr said. "But that is more than the act delivers, and more than it can deliver under our constitutional system." Bipartisan support for the law had turned into bipartisan disillusionment, and this time the law lapsed without a serious prospect of eventual revival.

The explanation for this is hardly mysterious. Democrats had come to see the independent counsels of the Clinton period in the same light as Republicans had seen them in the Reagan-Bush period: as dangerously powerful if not out of control. Walsh now had an opposite number: Kenneth Starr was to Democrats what Walsh had been to Republicans. There were other independent counsel prosecutions for Democrats to resent as well, much as Republicans still rankled at the memory of the investigations into Olson, Attorney General Edwin Meese III, and others: Independent counsels had investigated gifts accepted by Clinton Agriculture Secretary Mike Espy (he was acquitted); lying to the fbi about payments to a former lover by hud Secretary Henry Cisneros (he pleaded guilty to reduced charges); and financial impropriety by Commerce Secretary Ron Brown (the investigator closed shop following Brown’s death in a plane crash on an aid mission to the former Yugoslavia).

It is beyond our purpose here to assess the relative merits of the claims of abuse at the hands of unaccountable prosecutors. This much, though, is plain: The subjects and their friends on both sides of the partisan divide, rightly or wrongly, feel abused by independent counsel investigations. The sentiment is not necessarily consistent or general: There are Democrats who would defend Walsh to the end but say Starr was completely out of control, Republicans who would defend Starr even as they continue to despise Walsh. But a new element of bipartisan consensus did form in Washington in the 1990s. It was that our political system is better off without an independent counsel law. We do not want these prosecutors anymore.

Those still concerned with the potential for executive dereliction of duty in pursuit of politically embarrassing wrongdoing will have to find other ways of holding the executive accountable. While the law remained in force, many Republicans demanded an independent counsel to look into the 1996 Democratic fund-raising scandal; Attorney General Reno declined to find that the statute’s terms for seeking the appointment of one had been met. But despite control of both the House and the Senate, Republicans made no attempt to reenact the law. On the contrary, the hearings Congress held on the subject seemed weighted toward reasons it should be allowed to lapse.

Except that manifestly, it is not good enough. The Clinton impeachment was, in its entirety, novel, and the source of the novelty was the independent counsel act. If not an unconstitutional instrument in furthering the impeachment of President Clinton, Kenneth Starr’s office was surely an extra-constitutional instrument in that endeavor. It is hard to imagine Congress, upon catching wind of possibly illegal, possibly impeachable conduct by the president, then immediately empowering an investigator to do what the independent counsel was required by law to do — take as much time, in secrecy, with the full powers of compulsion of a federal prosecutor and the ability to grant unlimited immunity, with no requirement of interim consultation with any member of Congress let alone supervision by any congressional committee or anyone else, but with leave to keep the press informed of his activities, to investigate and reach a conclusion for presentation to Congress about the president’s possible impeachable offenses. In fact, the notion is ridiculous.

And if, as seems clear, the role of the independent counsel was necessary for events to culminate in the House vote to impeach, then we have a process whose legitimacy is open to question on account of this extra-constitutional mechanism. Starr himself, in his testimony opposing reenactment of the independent counsel statute, said that the public could lose confidence in the integrity of the investigation because of political attacks on the investigators — attacks to which the independent counsel himself could not effectively respond. This is surely true. But the charge assumes its conclusion, namely, the integrity of the investigators and investigation. The problem is that one may conclude that Starr and his team acted with perfect propriety, proportion, judiciousness, and probity at all times in the course of their investigation — but that even so, there was something fundamentally wrong with the investigation, something beyond the control of the investigators.

These circumstances suggest that while the siege on the independent counsel laid by the president’s defenders was far from laudable, it was certainly understandable. The role of the independent counsel was the most dubious aspect of the impeachment process, its weakest link — not the facts, not the Constitution, not the statutes or case law on perjury or obstruction; not even, given its total deference to the independent counsel investigation while it was under way and to the facts amassed by the independent counsel once he presented them, the congressional majority itself. The role of the independent counsel would be an obvious place to focus a defense, and so it was.

The attack on Starr was waged at times in highly personal terms. Some of the president’s defenders implausibly characterized him as a sex-obsessed weirdo, and many even believed their superheated rhetoric. But the attack was not in fact personal, for anyone in Starr’s position would surely have been subjected to precisely the same thing. No one in Starr’s role — that is, an independent counsel obviously engaged in preparing an impeachment case against the president — could possibly be deemed fair-minded and upright by the president’s loyalists.

The Starr report, and the boxes of evidence that accompanied it, contained testimony and evidence about the facts of the president’s conduct as well as quasi-legal, quasi-constitutional conclusions that the president may have committed impeachable offenses. One might focus on the facts, as the House mainly did, and ponder the independent counsel’s conclusion that they constituted evidence of high crimes or misdemeanors; House majorities on two of the articles derived from Starr’s work agreed with his conclusion. Or one might focus, as the Senate mainly did, on the conclusion itself, and ponder whether the House, following Starr, should have reached the conclusion it did. In the last analysis, the reason the Senate conducted only a truncated trial, with no witnesses brought to the chamber, was simply that the constitutionally required two-thirds of the Senate could not be mustered for conviction even if everything alleged by the House was true. Sen. Jim Jeffords of Vermont, a Republican who voted to oppose the articles, said, "The facts and circumstances of this case are low and tawdry, but these same circumstances do not, in my opinion, cause his offenses to rise to the level of impeachable acts." Democrat Tom Harkin of Iowa said, "This case should never have been brought to the United States Senate." Democrat Richard H. Bryan of Nevada said, "I conclude that the evidence presented in this case does not reach the standard commanded by the Constitution to convict and remove a president."

Obviously, this was likewise the view of the vast majority of House Democrats who voted against the articles. Some, such as ranking Judiciary Committee Democrat John Conyers Jr. of Michigan, also argued that the facts did not support the allegations in the articles. But others simply rejected the independent counsel’s conclusions about impeachable offenses and were accordingly less interested (in some cases, uninterested altogether) in the facts. Rep. Barbara Lee of California said, "This process and this action are the real crimes against the American people and our democracy." Given the president’s consistently high job-approval rating; public opinion polls showing substantial majorities opposed to impeachment; and surprising Republican losses in the November congressional elections, expectations were running high that a sufficient number of House Republicans would preemptively reach the conclusion that the president should stay in office and the proceedings should somehow be brought to an end.

Some have argued that the more conservative powers in the House, specifically Majority Whip Tom DeLay, forced the hand of moderate Republicans by foreclosing the option of a censure resolution and insisting that the House vote up or down on the four articles approved by the Judiciary Committee. This is not terribly persuasive, especially in the light of the subsequent Senate action. Senators, too, faced an up or down vote on conviction, with no third option, and here enough Republicans voted to acquit to deprive both counts of even simple majority support. If a small number of Republicans in the House had adopted the majority sentiment in the Senate, the matter would have ended with the defeat of the articles on the floor. That didn’t happen. The real choice was between contending points of view: If the offenses the independent counsel identified as impeachable were indeed the high crimes and misdemeanors to which the Constitution refers, the evidence was sufficient to find against the president; if the offenses did not rise to the level of high crimes and misdemeanors, then the particulars the independent counsel provided didn’t matter and might well constitute the worst sort of snooping. The House inclined toward the former view and the Senate the latter, but both chambers were narrowly divided on which to embrace. Whatever the outcome, the votes both in the House and the Senate were decisions on whether to endorse the work and the conclusions of the extra-constitutional office of independent counsel.

As well, Johnson could hardly claim to have been blameless. He avowedly adopted a strategy according to which he would use all of the powers of his office to oppose and block the stringent Radical measures. As Michael Les Benedict describes it in The Impeachment and Trial of Andrew Johnson (1973), "Where it suited him, he had ignored the Senate’s right to confirm government appointments, disregarded the Test Oath law, and emasculated the Freedmen’s Bureau and Confiscation Acts." By all accounts a stubborn man who became inflexible once he had made up his mind, he was also willing to take to the hustings to explain in no uncertain terms why his opponents were absolutely wrong and scoundrels besides. In the context of the times, this was rather scandalous behavior, especially as perceived by Republicans in Congress. The presidency was a far less powerful office then than now. Lincoln had laid claim to far-reaching wartime powers, but the widely held expectation was that after the passing of the crisis of the house divided, the office would return once again to its smaller scope. Johnson’s actions ran consistently and deliberately against this expectation. Unlike Clinton, nothing Johnson did or was even accused of doing could be construed as felonious. But he was somewhat relentless in flouting the standards of the day — and in this respect an agent of his own subsequent troubles.

The specific articles with which Johnson was charged mainly related to his effort to dismiss his secretary of war, William Stanton. Under the Tenure of Office Act, which had been passed over Johnson’s veto in 1867, an official appointed during a president’s term who had been confirmed by the Senate held office until the Senate confirmed his successor — he did not hold it solely at the pleasure of the president. Stanton, a holdover from Lincoln’s Cabinet, found himself increasingly at odds with Johnson’s views on Reconstruction. Johnson, well aware that he was courting congressional ire, suspended him and named Ulysses S. Grant interim secretary. The Senate, however, refused to confirm Grant, whereupon Grant stepped aside, giving the office back to Stanton — much to Johnson’s displeasure, since he was apparently expecting Grant to stay on, forcing Stanton to mount a court challenge. In February 1868, Johnson fired Stanton outright, naming another interim secretary. But Stanton, by now a hero to the Radicals and a symbol of Johnson’s depredations, refused to leave. And the House quickly voted, on party lines, to impeach Johnson, shortly thereafter settling on charges related to the Tenure of Office Act as well as other articles.

Accounts of the Johnson impeachment usually cast it as a product of a vicious struggle between two radically opposed points of view, in which the specific charges in the articles of impeachment merely vented the underlying partisanship. In Presidential Impeachment (1978) John R. Labovitz writes, "The Johnson case is the prime example in our history of an impeachment based on a pretextual issue." Other historians have hailed the "recusant Republicans" in the Senate who voted against Johnson’s removal for preventing a constitutional miscarriage — ouster of the chief executive not for high crimes and misdemeanors but over a policy disagreement, a step down the road toward a quasi-parliamentary system of government, in which the executive serves at the pleasure of the legislative majority. One frequently hears the Clinton impeachment described in similar terms — as a proxy for partisan bile. Hillary Clinton stated the case at its baldest when she described the president as a victim of a "vast right-wing conspiracy." Although many of those who supported Clinton throughout the ordeal would not go as far as that, insisting that the president had behaved reprehensibly, many held that Starr’s actions went too far and rightly or wrongly saw the impeachment’s origins in a partisanship the Republican congressional majority shared.

Clinton did not need his own "recusant Republicans" to ensure acquittal; there were more than enough Democrats in the Senate to keep him in office, at least so long as the party held the line in his defense. But he attracted enough anyway to prevent even a simple majority in favor of conviction. As in the Johnson case, perhaps even more so, this judgment invites the conclusion that the charges in the articles were trumped up for political reasons, or at least secondary to the political considerations.

Except that this analysis ignores a third commonality of the Clinton and Johnson impeachments. In addition to a partisan opposition in Congress and highly dubious conduct on the part of the president, the Johnson case had at its very core, similarly, a piece of flawed legislation that did serious damage to constitutional arrangements regarding the separation of powers. In this case, it was not legislation that shaped the process of impeachment, as the independent counsel statute did; it was the legislation the violation of which constituted the principal charge against the president: the Tenure of Office Act restricting Johnson’s ability to remove senior officials.

The historical judgment on the tenure act is not in much dispute. In Myers v. United States (1926), the Supreme Court took up the question of whether the president "has exclusive power of removing executive officers of the United States whom he has appointed by and with the advice and consent of the Senate." Chief Justice William H. Taft, in a landmark opinion that discusses in detail the constitutional history of the president’s powers, the understanding of them reflected in early legislative history, and subsequent attempts to circumscribe them, concluded for the court that separation of powers concerns made the power of removal of officers the president’s alone. Taft described this view as firmly established and widely accepted from the republic’s earliest days — until the Reconstruction Congress passed the tenure act as well as other laws governing executive branch personnel that "attempted to reverse this constitutional construction," arrogating to Congress the power to decide where to place the removal power, and granting that power to the Senate.

Neither Johnson nor subsequent presidents accepted this usurpation of executive authority. Johnson’s successor, President Grant, argued urgently for the repeal of the act in his first message to Congress, in fear of "the embarrassment possible to arise" from leaving on the books a statute "inconsistent with a faithful and efficient administration of the government." In response to Grant’s appeal, the House voted to repeal the law, but the Senate, entering a period at the zenith of its power in our political system, didn’t act. The particular law the court took up and overturned in Myers, concerning the appointment of postmasters and purporting to restrict their removal, passed in this period, in 1876. The Tenure of Office Act wasn’t repealed until 1887.

But in the scheme of things, do the particular provisions of the law the House charged Johnson with violating matter all that much? If the charges were merely a pretext for trying to oust him for political reasons, wouldn’t any pretext do?

Gerald Ford, reflecting on impeachment in 1970, before the House majority leader would be elevated to the vice presidency and the presidency, once famously said that "an impeachable offense is whatever a majority of the House of Representatives considers it to be at a given moment in history; conviction results from whatever offense or offenses two-thirds of the other body considers to be sufficiently serious to require removal of the accused from office." This is, in some sense, unarguably true, since the House has the sole power of impeachment and the Senate’s judgment at trial is final. But as a historical description of impeachments past, it is rather off the mark.

The fact is that when considering impeachment, the House customarily looks for "high crimes and misdemeanors" first in violations of the law — particular violations of particular laws. The literature on impeachment is rich with consideration of the question of whether impeachment is justifiable for conduct that is not criminal. As Raoul Berger demonstrated in Impeachment: The Constitutional Problems (1973), such impeachments have ample precedent in the English tradition, from which ours derives. We also have the sad case of the impeachment and removal of Judge Thomas Pickering in 1803-04; this once estimable patriot of the Revolution fell prey to mental illness, which he combined with drunkenness. The House and Senate agreed that his conduct on the bench was intolerable, and were willing in effect to stretch the notion of "high crimes and misdemeanors" to include it and remove him.

But the Pickering case is exceptional for that very reason. One thing most American impeachment cases have in common in the House is concern about the law. Strong indication of this comes from the Johnson case itself: These were arguably the worst relations between a president and Congress ever. But it is almost certainly wrong to say that just anything could have served as the instrument with which the Republicans would seek to remove Johnson. The fact is, the House voted as a whole twice on whether to impeach Johnson. The first time was December 7, 1867. The vote failed, 108-57. Sixty-six Republicans voted "no." As Eleanore Bushnell notes in Crimes, Follies, and Misfortunes: The Federal Impeachment Trials (1992): "The failure to impeach is surprising, considering the antagonism Johnson had stirred. He had not, however, violated any law, and for that reason even some of his committed foes backed away from impeaching him." The House voted to impeach only after Johnson fired Stanton. Here was a clear violation of the Tenure of Office Act, in the view of House Republicans. Far from being incidental, the act was essential to impeachment, and the participants at the time understood its supposed violation to make the difference between a majority in opposition to impeachment and a majority in support.

The Senate disagreed — or perhaps one should say, in the Johnson case, that a sufficient number of senators disagreed to allow Johnson to stay in office. The tenure act had passed over the president’s veto; therefore, two-thirds of the Senate had voted in favor of it. But a similar supermajority in the Senate was not willing to enforce its encroachment on the executive’s removal powers by removing a president who had flouted it in this case.

Now, as it happens, the House managers’ case against Johnson was murky on more than one point. It is doubtful, for example — or at least it takes a somewhat expansive reading of the tenure act to conclude — that it actually applied to Stanton, who was appointed during Lincoln’s term. (One issue came down to what the meaning of "term" is. Was Johnson finishing Lincoln’s term or did he begin his own term upon taking the oath of office?) A major part of the president’s defense in the Senate trial consisted of his contention that because both he and his Cabinet believed the Tenure of Office Act was unconstitutional and wanted to subject it to a court challenge, it was proper for Johnson not to enforce it in relation to Stanton’s appointment. (Here, the oddity is that Johnson initially complied with the act, by suspending Stanton and submitting Grant’s name to the Senate — perhaps not the actions of a man testing a principle.) One defender argued that perhaps the president had run afoul of the act — an unconstitutional act, to be sure — but that if so, his actions hardly rose to the level of "high crimes and misdemeanors." Bushnell also notes: "The president’s counsel had experienced no discomfort in acknowledging, at an earlier time, the disreputable tone of Andrew Johnson’s speeches nor in acknowledging the possibility that his political positions could be considered abominable. But they held steadily to their course that distasteful conduct or unpopular policies did not fit the constitutional model of impeachable acts."

These various contentions were sufficient to persuade the seven recusant Republicans to join all 12 Democrats in finding the president not guilty. Bushnell notes: "The six recusants who filed for posterity the reasons for deserting their party did not make a strong case for Andrew Johnson. Instead, all reported they believed as a matter of judgment and conscience that offenses they considered impeachable had either not been charged or had not been proved." William H. Rehnquist, in Grand Inquests: The Historic Impeachments of Justice Samuel Chase and President Andrew Johnson (1992), a study the chief justice of the United States published seven years before he was called upon to preside over the Senate trial of President Clinton, noted: "When all the evidence was in, and the arguments of lawyers on both sides concluded, the essence of the case turned on the Tenure of Office Act." Of the six statements issued by recusants, Les Benedict observes, "five of them turned on the disputed point — whether Secretary of War Stanton had been covered by the Tenure of Office Act. They concluded that the act did not protect him from removal without the Senate’s consent."

In both of our presidential impeachments and acquittals, partisan division between the White House and Congress combined with questionable presidential conduct have been necessary conditions for impeachment — but not sufficient conditions. There have, after all, been other instances of deep partisan division; 13 presidents have faced a Congress controlled in both chambers by the opposition party. (This list does not include Johnson.) Nor is our history short of questionable presidential conduct. But another necessary condition for impeachment has been a law at odds with the previous (and subsequent) understanding of the proper separation of powers. And acquittal has depended largely on a repudiation in the Senate of the intrusion of these laws.

Justice Samuel Chase was an ardent Federalist appointed to the Supreme Court by President Washington. The character of his actions and statements in public, especially outside the courtroom, was more or less the opposite of the reticent, apolitical affect we expect from judges today. Chase campaigned for Federalist candidates, advocated Federalist causes, and denounced the Republican opponents of the Federalists as a danger to the young republic.

Needless to say, he was hardly a revered figure when Republicans took control of the White House and Congress in 1801. The Jeffersonian party was further enraged by the slew of Federalist judicial appointments rushed through by the departing President Adams. Many Republicans believed the Federalists sought to retain control of the government through their hold on the judiciary.

Some Republicans saw the impeachment of Chase as the first in a string of potential impeachments, by which they might gain a toehold in the judiciary. Their partisan motives were not in doubt. Likewise, of Chase, as Bushnell writes, "That he was impetuous, arrogant, and overbearing is well-documented." His contemporary on the bench, Judge Richard Peters, noted Chase’s "singular instinct for tumult." Stephen B. Presser in The Original Misunderstanding: The English, the Americans and the Dialectic of Federalist Jurisprudence (1991), a revisionist work largely in sympathy with Chase, notes his facility for "building a circle of enemies."

When the collision with Congress came, its essence was a fight over the Sedition Act, which provided for the prosecution of seditious libel — false statements injurious to the government. Chase, riding circuit and presiding at trials, in the practice of the day, was a zealous champion of the law, and he seemed to relish enforcing it against outspoken Republicans. While in its earliest years, the durability of the American republic and the government created by the Constitution were open questions, and the threat to both in the form of such events as the Whiskey Rebellion was taken very seriously, the Federalist insistence on criminalizing certain political speech leaves a constitutional bad taste now, and for Republicans did at the time. Quite reasonably, if Federalists distinguished between statements injuring the government and statements damaging to the Federalist cause, it was lost on the Republicans. One of the charges against Chase was that he would not allow challenges to the constitutionality of the Sedition Act in his court.

The House impeached Chase in 1804 and the Senate acquitted him in 1805. The essence of the case against him was his abuse of power as a judge. In the words of manager George W. Campbell, Chase acted to "oppress, under the sanction of legal authority, those who became the objects of his resentment in consequence of differing from him in political sentiments."

The essence of the charge is that, under cover of the Sedition Act, Chase became more than a judge: rather, an oppressor, something more in the nature of lawmaker, prosecutor, judge, jury, executioner. In Constitutional Construction: Divided Powers and Constitutional Meaning (1999), Keith E. Whittington describes Chase’s conduct as an effort to incorporate the common law into the activities of the federal judiciary: "The early sedition prosecutions typified the perceived abuses inherent in the criminal common law. Not only did such prosecutions tend to entangle the judiciary with the executive as prosecutor and judge were drawn together to initiate, define, and execute the criminal prosecutions, but they also placed judges in the role of legislators in an area directly affecting personal liberty." The specific charges in essence lay out the terms by which the House supposed Chase had asserted constitutional powers he did not possess — an incursion into either the powers of others or the just power of no one.

But Chase and his team of defenders admitted to no such abuse on any of the charges. Chase may have been zealous, and he likewise may have been sharp with those whom he regarded as foolish, they said, but at no time did he exceed his authority as a judge. A hatred of prosecutions under the Sedition Act should not lead to condemnation of the judge sitting at trial. To conclude otherwise would risk doing damage to the independence of the judiciary — a different sort of injury to the separation of powers. The nine Federalist senators, none of whom deserted Chase, argued against what they saw in no uncertain terms as a partisan attack on judicial independence. In the end, Chase was acquitted by a majority vote of the Senate on eight of the articles, and on the other three, the number in favor of removal didn’t come close to crossing the two-thirds threshold. But his trial was a showcase for the abusive character of the now-expired Sedition Act.

Peck was judge in Missouri in a series of land claim cases in the territory of the Louisiana purchase. The law was complicated, the interests involved huge. In the first such case, in 1825 (the account here draws mainly on Bushnell’s in Crimes, Follies, and Misfortunes), Peck ruled against the client of a lawyer named Luke Edward Lawless. Because of the high degree of interest in the case, Peck published his ruling in a St. Louis newspaper in 1826. Shortly thereafter, a detailed rebuttal of Peck’s ruling appeared in another newspaper under the byline, "A Citizen." Peck was furious at the attack. He believed the "Citizen" rebuttal, in addition to its flawed legal reasoning, was replete with errors and misrepresentations of his ruling. Lawless’s authorship soon became known.

Bushnell writes:

Peck held the letter to be a contempt of court, sentenced Lawless to twenty-four hours in jail, and suspended him from practicing in federal court for eighteen months [a serious blow to Lawless’s livelihood as a lawyer specializing in land claims before the federal courts]. As the basis of the contempt ruling, Peck found that Lawless acted "with intent to impair the public confidence in the upright intentions of said court, and to bring odium upon the court, and especially with intent to impress the public mind, and particularly many litigants in this court, that they are not to expect justice in the cases now pending therein."

 

Lawless felt he was entirely within his rights to criticize a published decision and saw the contempt ruling as a tyrannical affront to the Constitution. He began a long crusade against Peck that ultimately led to impeachment nearly five years later on one article dealing solely with the judge’s treatment of Lawless. The article accused Peck of acting "to the great disparagement of public justice, the abuse of judicial authority, and to the subversion of the liberties of the people of the United States." James Buchanan, who went on to be elected president in 1856, was chairman of the House managers.

Peck maintained that his contempt ruling was within his powers as a judge, and his defenders argued that even if it went too far, Peck did not, as the article alleged, act with bad intent, believing that he possessed sufficient authority for his actions. At a minimum, however, it seems fair to say that Peck’s actions from the bench were harsh enough to meet the test of genuinely dubious conduct.

Peck was acquitted with 21 votes in favor of removal and 22 against. Where was the abuse of the separation of powers here? In this case, not in the statute books but in the common law — the precedents Peck relied on to hold Lawless in contempt and to sentence him harshly. As Bushnell observes, Peck’s defenders "sought to refute the charge of abuse of the contempt power by citing English and American precedents supporting the authority of courts to punish for contempts like Lawless’s." The House tried to hold his conduct to the standard of its more circumscribed view of judicial contempt powers. The Senate was not willing to rely on the House’s assertions to the extent necessary to remove Peck.

But the Senate, like the House, can hardly be said to have found Peck’s conduct salutary. Both chambers amply demonstrated this by approving, within a month of Peck’s acquittal, legislation introduced by Buchanan restricting contempt findings in federal courts roughly along the lines of the terms the House managers had unsuccessfully tried to apply in Peck’s impeachment. Contempt could be found in misbehavior in a courtroom or close enough to it to disturb its proceedings; or in misbehavior in such business of the courts’ as filing motions and briefs; or in the failure to obey a lawful court order. It could not be found in a newspaper rebuttal to a court’s decision. Buchanan’s legislation governs contempts in federal courts to this day.

There is certainly nothing wrong with such exercises. But the approach does invite a sort of all-or-nothing view of the issues involved. It does not do justice to the historical richness and the lasting meaning of these impeachments and acquittals. Better to take these "grand inquests" for what they really are — times at which great issues about how government works, and therefore how we should govern ourselves, come to the fore.

The Clinton impeachment and acquittal were rich in both regards. The facts about Clinton’s misconduct in trying to conceal his relationship with Monica Lewinsky were stubborn and damning. No president should do what Clinton did, and his name will forever bear the mark of impeachment for it. Yet it would have been impossible to remove Clinton without simultaneously vindicating the process by which he came to stand trial in the Senate — which is to say, without vindicating the House’s deference to and reliance on independent counsel Kenneth Starr’s investigation, the conduct of the Starr investigation in all its particulars, and the authority for the investigation, the independent counsel statute itself. And it is noteworthy that within the year, the statute was gone, bringing to an end its 20-year disfiguration of our political system.

President Johnson’s determined efforts to stretch his powers to an extreme in order to thwart the Reconstruction Congress were offensive — as, indeed, was his obsequiousness toward the South. Yet it would have been impossible to remove Johnson without vindicating the Tenure of Office Act and its assertion of congressional authority over the president’s ability to remove officials. Two decades later, Congress would at last abandon that claim to authority, and 40 years after that, the Supreme Court would lay it to rest once and for all.

Justice Chase was political in a manner wholly unbecoming a judge and had a view of jurisprudence that expanded the role of judges beyond the bounds established in the American Constitution. Yet it would have been impossible to remove Chase without denying his claim to the authority he asserted from the bench, even if that claim did derive from the dubious and discredited Sedition Act, with its invitation to judges to act as lawmakers and prosecutors. The Sedition Act had already gone by the boards at the time of the Chase impeachment, and the judiciary was thereafter reluctant to assert executive powers so sweeping.

Judge Peck, personally slighted, gave vent from the bench to an intolerable view of press freedom. Yet, once again, it would have been impossible to remove Peck without denying him his claim to his contempt authority, even if that claim derived most of its legitimacy from precedent that had gone largely unexamined in the light of passage of the federal Constitution and its protections of individual liberty. It took no more than a month for the government to correct the problem posed by Peck’s broad assertions with the passage of a law governing contempt.

In these four historical cases, impeachment followed by acquittal has been a messy and partisan process, offering numerous invitations to rhetorical excess, and reducing the decision to a binary choice when "guilty" and "not guilty" have not begun to do justice to the issue. And while, in the end, these impeachments and acquittals have maintained the status quo, they have done so only after bloodying everyone who was party to them.