Liberty Matters

Qualified Immunity: A Qualified Amen to Judge Glock


I am totally on board with what I take to be one of Judge Glock’s key points: the Administrative Law debate of recent decades has been overly concerned with administrative rulemaking (delegation, Chevron, and all that), and sadly inattentive to adjudicative and enforcement proceedings. I am likewise on board with the enterprise of re-thinking the law, so-called, of “qualified immunity.” But the problem of what to do about official misconduct vis-à-vis citizens is quite complicated. And I am not sure that A.V. Dicey was entirely right on that broader issue.
The basic framework of government liability operates against a very firm background of sovereign immunity. You cannot sue the sovereign—the government; “the state,” as continental jurists would say—without its consent. (You can in the U.S. sometimes sue a state without its consent. But that just shows that the states aren’t quite as sovereign as they often think they are.) Still, any decent country will want to provide legal redress for government misconduct. There are two basic ways of doing that.
One is to permit suit against individual officers. That was the English system as Dicey described it, and it was our system throughout the 19th century and beyond. Importantly (we’ll see), those “officer lawsuits” were not claims directly under the Constitution. Rather, they were common law actions (for example, trespass), usually brought in state court, that ran against officers just as they would against private actors. The officer would then assert a defense that is not available to private parties: I acted on the lawful authority of the United States. The plaintiff would respond that the officer exceeded his authority (acted ultra vires) or else, under our system though obviously not England’s, that Congress had no business conferring that authority. That’s how the Constitution came into play in those days.
The alternative is to bar suits against officers and to permit suits against the government—to waive sovereign immunity, as we say. That was the French system as Dicey described it. And that, over a wide range, is our system now. The Administrative Procedure Act (APA) is a waiver of sovereign immunity. So is the Federal Tort Claims Act (FTCA).  In most suits for money against a federal officer, the government will substitute the U.S. as a defendant (and, if need be, remove the case to federal court).
Why was this difference so important to Dicey? The English system, he argued, embodied the rock-bottom rule-of-law principle that no man is above the law—not even an officer. In contrast, the French system of officer immunity, even and perhaps especially when coupled with relief against the state, encourages officers to go rogue. The argument has great appeal. But is it entirely right?
As to whether the French system encourages rogue conduct: that depends, no? One can easily imagine a legal system that provides prompt and effective redress against the government in an independent court, coupled with an internal administrative system of strict supervision and regularity. Presumably the government will want to create such a system, if only to protect the public fisc. Under that scenario it is hard to see the rule-of-law difficulty, except perhaps at a very high level of abstraction.
The officer lawsuit model, meanwhile, poses problems that eventually prompted its erosion.
One problem is that the fear of liability might prompt officers to be excessively lax in the performance of their duties. In the U.S. we addressed that problem by having tax collectors and customs agents work on a commission basis. How well did those rival incentives work? It’s hard to say, for want of empirical evidence. But on Jerry Mashaw’s account —no fan of rule-of-law obsessions he—it seems to have worked quite well. But this is not an abstract rule-of-law question; it is more of a public choice question. One has to think about the officials’ incentives, and it is hard to get them right.
A second question is what to do about innocent mistakes in the process of following official orders. A famous illustration is Little v. Barreme (1804). A congressional statute enacted during the quasi-war with France authorized the President to stop and seize American ships sailing to a French port. To prevent evasion, the Secretary of the Navy instructed his commanders that ships were often flying under a false flag (especially Denmark’s). Also, one cannot easily tell where to or from folks are sailing out there on the open seas. So, commanders were instructed to exercise sound judgment and “do all that in you lies” to implement the true intent of the statute. Pursuant to that order Captain Little capturee the Flying-Fish, sailing under a Danish flag, and dragged her into Boston. Turns out she was actually Danish and was sailing from a French port, not to one. So the owners wanted their ship back, and they wanted damages—from Captain Little, who acted on direct orders and on all accounts on reasonable suspicion. Chief Justice Marshall’s opinion highlights the conundrum. Initially, he thought that “though the instructions of the executive could not give a right, they might yet excuse from damages,” at least for military officers. But then, his brethren convinced him that the instructions could not “legalize an act which without those instructions would have been plain trespass.” As Dicey recognized and emphasized, there can be no respondeat superior liability under the officer model, and no “I was following orders” excuse from liability. The point is to hold each and every individual officer responsible.
But that’s way harsh, no? What’s a conscientious officer like Captain Little to do when faced with financial ruin? Ask Congress for relief, that’s what. Conversely, what’s a plaintiff to do when the officer is judgment-proof? Ditto: ask Congress. Halfway through the 19th century Congress tired of the ceaseless petitions and created a (quasi-)judicial mechanism, which eventually became the Court of Claims. A rough century later Congress enacted the FTCA, for similar reasons: routinize the claims process; provide a federal forum. Both of these solutions have their problems, but the basic point is this. Congress migrated away from the officer suit model not because it threw in the towel on the rule of law but for perfectly intelligible reasons.
Far and away the biggest problem with officer suits is that they are feast or famine. If the officer acted ultra vires (or in derogation of a mandatory duty) you get full-scale judicial review and relief. In contrast, so long as the officer acted within his authority, you get no relief—no matter how arbitrary, oppressive, and unreasonable the exercise of discretion may have been. The reason is that “reasonableness” review means control of executive discretion, and judges who exert such control might as well be the executive. Need a case cite? Marbury v. Madison. That’s the logic. English law still follows it, although even there courts have begun—under the baleful influence of continental and especially EU jurisprudence—to probe the reasonableness and “proportionality” of administrative decisions. For our part we have enacted the APA, which instructs courts to set aside agency actions that are “arbitrary, capricious, or an abuse of discretion.” That, too, breaks with Dicey-an precepts. But again, you can see the basic impulse: as bureaucratic discretion expands, so does the perceived need for some sort of judicial check. And again, while the APA has its problems (lots of them), it’s hard to characterize it as a horrendous breach with the rule of law.
What, though, of the “qualified immunity” that precludes monetary relief unless the officer knew or should have known that he violated well-established rights? That is indeed completely made-up, and deeply problematic—but not for Dicey-an “no man above the law” reasons. Or, to speak more reverently, the reasons are more complicated than a projection of Dicey-an notions on contemporary law would lead you to suspect.
In a very real sense, qualified immunity is a product of an excess of law and of the Warren-Brennan Court’s irrational exuberance about rights. One manifestation was a fantastic expansion of forward-looking declaratory or injunctive relief against public officers in their “official capacity.” But that leaves cases where money is the only meaningful remedy and the claim has to run against the officers in their “individual capacity,” for sovereign immunity reasons. In the 1960s, the Court radically revamped the framework for “Section 1983” actions, which are the principal means of enforcing federal constitutional and statutory rights against deprivations “under color of [state] law.” Such actions, the Court held, could go forward even if the officers’ conduct was quite plainly prohibited by state law, and even if state law provided perfectly adequate legal remedies. In 1971, the Court created so-called Bivens actions for monetary relief against federal officers, directly under the Constitution—as if that instrument traveled with its own causes of action.
Having opened this barrel of worms, the Court soon backtracked. Wait: there can’t be Bivens actions when Congress has provided an alternative remedy, however meager. (At this point, Bivens actions are hopeless outside the Fourth Amendment context.) Wait: there can’t be “implied” rights of action against states or their officers. And, wait: to be held individually liable, for money, the officer should have known that he violated clearly established rights. Because otherwise it’s open season on officers. So as the rights expand, the remedies contract.
(This was Justice Scalia’s account of the matter. William Baude, in an important article cited by Judge Glock, has disputed it, but I think it is basically right as a matter of judicial perceptions and intuitions, if not necessarily doctrine.)
I am no fan of the Jackson Pollock canvas the Court has painted over Section 1983 and the law of constitutional remedies. However, the robed ones are attempting to strike the same balance that the old law sought to strike: find some arrangement to protect individual rights, the rule of law, and effective government. (Dicey himself put the calculus that way.) Thus, if we want to re-examine the “qualified immunity” band-aid the Court has plastered on this corpus juris, we ought to examine the wound itself.
That wound is the conviction that the common law is somehow not good enough for a decent society. The Section 1983 cases strongly suggest that. Bivens says it in haec verba. If the Court could put those dogs out of their misery (as two of my comrades in arms have urged), the official immunity problem would largely take care of itself. Hit the officers for trespass, battery, and the like: no immunity. And then junk the made-up claims.
This Court, for good but mostly for ill, is not going to do any such thing, for two reasons. One, it is addicted to a policy of narrowing precedents into oblivion, instead of overruling them outright. Two, we are all originalists now, and so we must be hostile to anything that smacks of common law adjudication. (Rote cite to Erie Railroad.) The point remains though: we are not going to fix official immunity law in isolation. At an intellectual, academic level, one would have to rehabilitate common-law modes of thought and argument. At a practical level, one would have to find ways of putting this over on soi disant originalist judges.
Maybe in this theater, we should all be Diceyans after all.