I have argued that Mr. Trump’s pardon of Sheriff Joe Arpaio is potentially an impeachable offense. Others have gone further, asserting that Mr. Trump lacked the constitutional power to issue this pardon.  I think these arguments are wrong as a matter of constitutional law, and ill-advised if conceived as part of a larger effort to secure impeachment of Mr. Trump.

I have already addressed the contention that the Arpaio pardon violated the due process clause.  Today, I take up the claim by Dean Erwin Chemerinsky and others that it violates the separation of powers.

Dean Chemerinsky, together with prominent lawyers Michael and Jane Tigar, advances the separation of powers argument in an amicus brief filed with the district court judge who found Sheriff Arpaio in criminal contempt.  The brief is a learned, creative, clever piece of legal work.  It’s also, sadly, a lovely house of cards that collapses at a touch.

To see why requires a quick recap of the proceedings that resulted in Arpaio’s contempt conviction and a review of basic contempt and pardon law.

The Arpaio contempt: In 2007, a group of individuals filed a federal civil rights class action against Arpaio and the Maricopa County Sheriff’s Office alleging discriminatory policing and a variety of other misbehavior. The case dragged on for years, with Arpaio resisting every step of the way.  In 2010 and again in 2012, the Department of Justice filed its own civil lawsuits against Arpaio and the county. Both DOJ and the class action litigants secured court orders requiring changes in the behavior of the Sheriff’s Department, but it became evident that the orders were not being complied with.  In 2015, DOJ intervened in (meaning that it became a party in) the civil rights action to coordinate its enforcement efforts with those of the plaintiffs and the court. Arpaio’s resistance to the court’s authority continued.

In May 2016, Judge Murray Snow found Arpaio and others in civil contempt of orders issued in the civil rights action by then being pursued by both private plaintiffs and the Department of Justice. In July 2016, Judge Snow ordered a series of remedies for civil contempt designed to protect the rights of the plaintiff class – such as a monetary compensation scheme for victims and changes in sheriff’s office policies.  On August 19, 2016, he also formally requested the Department of Justice to “prosecute” Sheriff Arpaio and others for “criminal contempt” before a different judge.

A separate criminal case was opened, with a different caption and different case number — United States v. Arpaio, Case No. 2:16-cr-01012.  The only parties were the United States and Sheriff Arpaio, not the individual plaintiffs in the separate civil rights action.  The “cr” in the case number denoted a criminal, rather than a civil, case.  On July 31, 2017, Judge Susan Bolton found Arpaio guilty of criminal contempt in violation of 18 U.S.C. Section 401. On August 25, before Judge Bolton could sentence Arpaio, President Trump issued him a pardon in the criminal contempt case.  He did not pardon Arpaio in the civil contempt case, which at all events he lacked the power to do.

As illustrated by the Arpaio case, contempt can be either civil or criminal.  In civil contempt, the penalties are coercive and often conditional (such as fines that continue to accrue so long as offending party persists in violating the court’s order), and are designed to compel compliance with a court’s order and vindicate the rights of the litigants. In criminal contempt, the penalties are fixed, not conditional on the defendant’s subsequent behavior, and intended to vindicate the authority and dignity of the court.  Criminal contempt penalties may include incarceration and all the collateral consequences of any other criminal conviction, such as disqualification from voting or running for office.

The law of pardons and contempts: Article II, Section 2, of the Constitution states that the president “shall have power to grant reprieves and pardons for offenses against the United States, except in cases of impeachment.” When lawyers and judges use the word “offense” as a general term describing a category of legal cases, they mean “crime,” as distinct from a civil wrong.  Thus, the plain language of Article II, Section 2 gives the president the power to pardon crimes, but not the power to void civil judgments against federal litigants.

Criminal contempt “is a crime in the ordinary sense,” a point the Supreme Court has affirmed over and over again.  And a violation of 18 U.S.C. 401 is a felony.

Moreover, the Supreme Court has expressly held that the pardon power extends to criminal contempts. In Ex parte Grossman, a 1925 opinion written by a man who knew about the pardon power from first-hand experience, Chief Justice and former President of the United States, William Howard Taft, the Court addressed exactly the same separation of powers argument Chemerinsky makes and unanimously rejected it.

Chemerinsky tries, as lawyers say, to “distinguish” the Grossman case — which means to argue that Grossman doesn’t count because it decided a different issue than the one presented by the Arpaio pardon.  In a nutshell, Chemerinsky claims that the Arpaio contempt was different that than the Grossman contempt in a way that, for historical reasons, places it outside the president’s pardon power.

The argument is intricate, elegant … and wrong.  It runs something like this:

The Supreme Court in Grossman said that, when creating the president’s constitutional pardon power, the framers thought of it as roughly coextensive with the King’s pardon power in England before the American Revolution.  Therefore, Chemerinsky sets out to find some limitation on the King’s pardon power that, by analogy, would limit a president’s power to pardon the Arpaio contempt.

In Merry Olde England (1300s though the 1600s), there were several avenues of redress for victims of crime.  The first was an ordinary prosecution of the wrongdoer in which the Crown was theoretically the party bringing the case. I say theoretically, because until the 1800s there was no body of public prosecutors, and virtually all criminal cases were brought by the victim, who was commonly called “the prosecutor.” Douglas Hay, et al., Albion’s Fatal Tree 40 (1975). The punishments in such cases were those we would ordinarily think of as criminal – fines, prison, or death – plus some specialties of the period like flogging, the stocks, or transportation to the colonies.  In addition to this customary practice of 18th Century English criminal courts, there was a very old procedure called “appeal of felony” that allowed victims to bring private prosecutions which, if successful, could result in both restitutionary payments to the victim and criminal punishment of the wrongdoer.  Chemerinsky cites several legal treatises, and might have cited others, to the effect that the King could pardon defendants convicted in actions brought by the Crown, but could not pardon defendants convicted in private “appeal of felony” actions.

Chemerinsky argues that the constitution’s founders would have been familiar with the exclusion of “appeal of felony” cases from the King’s pardon power, and thus that they intended to limit a president’s pardon power in any modern case analogous to the old “appeal of felony” mechanism.  As an historical matter, this is deeply unlikely.

In the first place, “appeal of felony” does not seem to have existed in American colonial jurisprudence. Boston Railway v. Dana, 1 Gray  83 (Mass. 1854), in Charles E. Chadman, Chadman’s Cyclopedia of Law, Col. X1 (1906), at 12-16. There is no mention of “appeal of felony” in the records of either the constitutional convention or any of the state ratifying conventions. Moreover, even in England, the mechanism of “appeal of felony” was already falling out of favor by the early 1600s, and there are only a dozen reported cases of its use in all of the 1700s. “By 1800 the appeal was as obsolete as any institution can be that has not been formally abolished.” J.H. Baker, Criminal Courts and Procedure at Common Law 1550-1800, in  J.S. Cockburn, Crime in England 1550-1800, at 18.

In short, it is improbable that, by 1787, even an unusually well-read American lawyer would have any personal acquaintance with a legal mechanism that had effectively been extinct in England for nearly a century.  The likelihood that any significant number of the constitution’s drafters or ratifiers knew about “appeal of felony,” and were aware that it was outside the royal pardon power, and thought that bit of historical arcana relevant to the scope of the pardon power of an American president approaches zero.  As one scholar wrote about a similar issue, “Even in the unreformed common law, there was a distinction between precedents and fossils.” Martha Ziskind, Judicial Tenure in the American Constitution: English and American Precedents,” Supreme Court Review 135, 138 (1969).

But let’s indulge the fiction that the founders knew about “appeal of felony” and the King’s pardon.  If so, the analogy of “appeal of felony” to criminal contempt proves exactly the reverse of what Chemerinsky claims. In England, once the Crown undertook prosecution of a criminal wrong, the King could pardon the resulting conviction. Likewise, the King could, and very often did, issue pardons in cases where private prosecutors filed informations in the name of the Crown. John Langbein, Shaping the Eighteenth-Century Criminal Trial: A View from the Ryder Sources, 50 U. Chi. L. Rev. 1, 21 (1983). In Arpaio’s case, the Department of Justice became a party to the civil rights action and then, in a separate case heard by a different judge, prosecuted and obtained a criminal contempt conviction against Arpaio. Hence, even if this were 1787 and Mr. Trump were His Royal Highness Donald I, once the government entered the case seeking criminal penalties for violation of a criminal contempt statute, the resultant conviction would be pardonable by the Crown.

Finally, and most importantly, Chemerinsky argues that the contempt conviction in Grossman is different  because there the underlying lawsuit was brought by the government to enforce a federal statute, whereas the underlying lawsuit in Arpaio was a civil rights case initially brought by individual plaintiffs. Therefore, he contends, the contempt in Grossman was designed to protect government interests, while the contempt in Arpaio was designed to protect the interests of the individual civil rights plaintiffs in the original lawsuit. But in attempting this distinction, Chemerinsky either ignores or garbles two centuries of American case law on contempt and glosses over the procedural posture of Arpaio’s criminal contempt.

To begin, there is no suggestion in prior cases that the distinction between civil and criminal contempt turns on the identity of the litigants in the lawsuit in which the defendant behaved contemptuously.  Rather, federal courts have held over and over again that the difference between civil and criminal contempt lies in the nature and purpose of the penalties imposed.  As noted above, civil contempt penalties are designed to compel compliance with a court’s order and “vindicate the rights of the litigants.” Putting it another way, the point of civil contempt is either to give the party injured by the contumacious conduct immediate relief in the form of something like monetary compensation or to coerce a recalcitrant person into changing his future behavior by following the court’s orders so that litigants get the relief to which the court has found them entitled.

By contrast, the purpose of the penalty imposed following a criminal contempt is punishment.  It is delivered in the form of a criminal sentence indistinguishable in form and effect from a sentence for any other crime. It is fixed — so many months in prison, such-and-such a fine paid to the government — and cannot be later reduced or altered conditional on the defendant’s subsequent behavior.  (Indeed, under Rule 35(b), Fed. R. Crim. P., fourteen days after any criminal sentence is entered, the district judge loses the power to change it.  So a judge could not issue a conditional sentence in a criminal contempt even if she wanted to.) In short, the point of criminal contempt is not to compensate the injured party or coerce the contemnor into changing his future behavior to comply with the court’s orders.  It is to punish the contemnor for his past behavior in defiance of the authority and dignity of the court.

Of course, the threat of facing a criminal contempt prosecution may have the incidental effect of deterring a recalcitrant party from further defiance of the court’s orders, and an actual criminal contempt conviction may convince the convict to comply with orders he previously resisted.  Thus, a criminal contempt may incidentally benefit litigants harmed by the contemnor’s defiance.  But the object of criminal contempt remains protecting the authority and dignity of the court, not benefiting any party to the case.

Chemerinsky’s brief notes that the Supreme Court’s Grossman opinion alludes to the principle that a pardon in a contempt case is ineffective “to halt or interfere with the remedial part of the court’s order necessary to secure the rights of the injured suitor.”  But that snippet is nothing more than a nod to the fact that, in some prior cases, the same judge may have issued both civil and criminal contempt judgments. The Court is merely clarifying that, in such cases, a presidential pardon could void only the judgment of criminal conviction and any characteristically criminal punishments, but not the characteristically civil remedial components of a contempt order.

There is no danger of any confusion on this point in the Arpaio case.  The civil and criminal contempt judgments against Arpaio were separated by over a year and came in separate actions. The court’s order in the civil contempt case contained nothing but classic remedial measures.  We can’t know what sentence the judge would have imposed in the criminal contempt, but the penalties could only have been those permissible under federal criminal law.

In short, Grossman cannot be distinguished. And in Grossman, the Supreme Court held that a presidential pardon of a criminal contempt does not violate the constitutional principle of separation of powers.

Should Grossman’s holding on separation of powers be reexamined and overturned? 

Of course, even though Grossman is good law and indistinguishable from the Arpaio case, the Supreme Court could change its mind. Dean Chemerinsky argues that it should.  He makes three basic arguments, none of which withstand scrutiny.

First, he contends that aggrieved litigants in federal court have a “right to redress” implied from Article III, of which they would be deprived if judges could not employ criminal contempt sanctions to coerce compliance with their orders.

Second, he argues that the power to compel obedience to the judgments of courts through contempt sanctions is an inherent component of judicial authority that exists independent of any statutory authorization. He reluctantly concedes that Congress can, and has, limited that authority in a variety of ways. But he strongly implies that either removing or significantly limiting the judiciary’s contempt power would violate Article III.

Both of these claims are debatable, at least in the sweeping form advanced by Dean Chemerinsky. A general right to redress does not make constitutionally mandatory every possible means of judicial coercion. Likewise, the existence of an inherent judicial contempt power does not necessarily imply that criminal contempt is a constitutionally mandated attribute of judicial authority.  If, for example, Congress were to repeal Sections 401 and 402 of the criminal code and decree that henceforward federal judges would enjoy only civil contempt authority, it seems doubtful that such an action would be unconstitutional.

But even if we concede that litigants as a class have a constitutional entitlement to redress and that judges have a constitutionally implied power to hold in criminal contempt those who defy judicial orders, neither proposition creates a constitutional argument for voiding the Trump pardon of Arpaio.

The general principle that litigants have a right of redress is, at most, a guide to the kinds of processes that ought to be built into the judicial system as a whole.  It does not imply a rule that every litigant must receive perfect justice or complete satisfaction of all his legal objectives. Nor does it imply that judges are to be the exclusive arbiters of how justice should be apportioned. Once the pains and stigma attendant on criminal conviction enter a case, Anglo-American law has long reserved a place for executive judgments about clemency.

As for the Arpaio pardon in particular, it changed no laws, procedures, or rules of court.  It has no effect on the right of redress of any litigant in any case other than that involving the Maricopa County Sheriff.  And even there, the civil rights plaintiffs sued, won, and secured injunctive relief and monetary compensation. Arpaio and the county resisted the court’s orders, a resistance that begat further court orders, a civil contempt verdict, and additional remedial measures. Arpaio was actually convicted of criminal contempt, even though he escaped punishment for it by virtue of the pardon.  That is a deeply regrettable circumstance and one that reflects on Mr. Trump’s fitness for office precisely because it suggests his personal disregard of both the individual plaintiffs and the sanctity of their constitutional liberties. But it is not an outcome that denied the civil plaintiffs all “redress.”

Likewise, the general power of judges to hold recalcitrant litigants in criminal contempt remains unchanged by the Arpaio pardon. By issuing the pardon, Mr. Trump repealed no statute, promulgated no new Justice Department policy, advocated no new interpretation of Article III, and raised no challenge to the criminal contempt power of judges.  Here, too, his action reflects adversely on his fitness for office because it manifests a personal disregard for the role of an independent judiciary and a disposition to employ the powers of the presidency to distort the rule of law in favor of friends or political allies.  But the baseness of his motives in exercising a power granted the president by the constitution does not deprive judges as a class of their contempt power.

Which brings us to what might be characterized as the third of Dean Chemerinsky’s arguments, which is that the authority to hold persons in criminal contempt for violating court orders is such an indispensable attribute of judicial power that voiding it, even by presidential pardon in a single case, violates the constitutional principle of separation of powers.  With the greatest respect to Dean Chemerinsky, I do not believe that invocation of the phrase “separation of powers” can nullify the plain pardon language of Article II, Section 2 — “The President … shall have power to grant reprieves and pardons for offenses against the United States, except in cases of impeachment.”

After all, the constitution nowhere says, “There shall be a separation of powers.” Instead, it creates a structure of three branches of government and specifies how powers are to be distributed  among them.  The separation principle is inferred from the textual power distribution. In cases where the language of the constitution is unclear or fairly open to interpretation, or in cases that plainly could not have been anticipated by the founders, it may be appropriate to employ the inferred separation principle to decide the constitutional propriety of a contested use of power by one of the branches against another.  But, at least absent some extraordinary justification, one cannot use a general principle that is, after all, merely an implication from the structure to void an explicit, unequivocally worded, part of the structure.

Moreover, the inferred separation principle does not mean that the three federal branches occupy three non-intersecting silos of authority.  Rather, as we all learned in high school civics, the Madisonian constitution is one of checks and balances — three co-equal branches, each endowed both with its own characteristic powers and powers to check abuse of power by the others.  The presidential pardon power is a checking power.  It was designed to provide a case-by-case executive branch check on judicial overreach and an avenue of redress for individuals oppressed or misjudged by the courts.

Note that, unlike Professor Martin Redish and others who argued that the Arpaio pardon in particular might be subject to judicial review on its special facts and voidable as a violation of due process, Dean Chemerinsky’s argument is not limited to the Arpaio case. Instead, he is necessarily arguing that no president can ever pardon any criminal contempt, regardless of its circumstances. Judicial power over contempts, he says, is absolute and untouchable by executive clemency.

For this categorical exclusion to succeed in the face of the unequivocal language of Article II, Section 2, would require some powerful reason to believe either that occasional use of the pardon power would subvert judicial authority generally, or that criminal contempt cases are peculiarly immune from the danger of judicial misjudgment, meanness, or malignancy.

But the Arpaio pardon, however repellant one may find it, is not an instance of one branch (the president) preventing another branch (the judiciary) from functioning at all. The judiciary is still in business. Nor does it present an example of one branch preventing another branch from exercising a particular type of power.  Mr. Trump didn’t pardon everyone who has been found guilty of criminal contempt of a federal judge.  Nor did he threaten to do so.  The contempt power remains available to federal judges to enforce their orders, and they do so daily.

Moreover, there is no reason to think that judges are any less prone to misjudgment or injustice in cases of criminal contempt than in any other class of crime.  Indeed, there is every reason to fear that, in cases that necessarily involve defiance of their official authority and may closely touch their own outraged professional dignity, at least some judges may lose perspective.  This risk does not outweigh the imperative requirement of effective means of enforcing judicial orders, but it makes the case for an executive check on particular instances of judicial overreach even stronger than in ordinary criminal cases. Presidents do, and should, have the power to pardon criminal contempts.

The bottom line is that Mr. Trump had the power to pardon Joseph Arpaio.  That does not mean he was right to do so.  Nor does it mean that the country is without remedy if, as I believe, he was egregiously wrong.  But, as I have written repeatedly on this site, the remedies are political — repudiation of Mr. Trump and his party at the polls or impeachment.

The liberal legal intelligentsia and impeachment

In writing about Dean Chemerinsky’s brief on the Arpaio pardon, the almost invariably perspicacious Dahlia Lithwick opined that if the Dean and others succeed in convincing judges that the pardon violated the separation of powers, that would strengthen the case for impeachment.  At least with respect to Dean Chemerinsky’s brief,  I am constrained to disagree.

Dean Chemerinsky does not argue that this particular pardon was a uniquely abusive exercise of presidential authority.  Rather, he asks the court to find that neither Mr. Trump nor any other president can pardon anyone for criminal contempt, regardless of the circumstances. To win, he would have to convince the Supreme Court either to overturn the Grossman case or distinguish it based on a novel, technical, antiquarian basis. But then what?  Arpaio’s contempt conviction would be reinstated, but not because this pardon was a reprehensible abuse of authority, but because the Court reversed itself and decided for the first time that all presidents lack the power to pardon anyone for criminal contempt.  One can hardly impeach a president for failing to anticipate a U-turn by the Supreme Court. The Arpaio pardon is only plausibly an impeachable offense if Mr. Trump either violated a well-established constitutional rule or grossly abused a concededly available presidential power in this particular case.

Which brings me to my final point in this long screed. To be candid, despite my immense respect for Dean Chemerinsky, who is one of the undeniable giants of the American legal academy, his argument for the unconstitutionality of the Arpaio pardon is strikingly weak. It might persuade a district judge understandably angry at Arpaio’s long and inexcusable defiance and at the crassly political interference of a crass and belligerent president.  But it will go nowhere thereafter. And it shouldn’t.  The general rule for which it argues is contrary to the plain text of the constitution, the plain holding of the Supreme Court, sound principles of constitutional interpretation, and simple good sense.

It feels, I am sorry to say, like another example of brilliant legal scholars sponsoring tenuous legal arguments in the service of resisting, and potentially removing, Donald Trump. Heaven knows, I sympathize with the objective. Mr. Trump’s frightening unsuitability for his office becomes more evident by the day.

But if I may be pardoned the solecism of quoting something I wrote several months back:

Those who would impeach a president for his disregard of democratic norms must come to the task with clean hands. It will not do to say that, because Trump flaunts some norms, other norms can justifiably be bent or broken to bring him low. If that becomes the stance of Trump’s pursuers, then they become little better than the object of their indignation, and what should be a righteous pursuit will seem no more than partisan vindictiveness. Which will degrade the very standards of conduct we should all be seeking to uphold, and lead, inevitably, to yet another round of bitter political warfare conducted with even fewer restraints and less decency.

If that happens, those who oppose Mr. Trump will have failed utterly, even if they succeed in expelling him from office.

When legal academics enter the fray over Mr. Trump, we must hew tightly to our own cherished norms of intellectual honesty and analytical rigor.  There is nothing wrong, of course, with creativity and even a bit of aggressiveness. Mr. Trump, after all, presents unique challenges to the legal order we seek to uphold and his novel transgressions are fair game for legal innovation.  Still, we ought not be sponsoring arguments against Mr. Trump that, on the same or analogous facts, we would not make against a President Obama or a President Bush. When we yield to that temptation, we weaken whatever authority we have as honest and more-or-less neutral interpreters of the law.  That authority will be needed in the months and years to come, I think.  We should avoid squandering it.