As a general rule, if the government appropriates your property, it is considered a taking of private property for public use under the Fifth Amendment and the state must pay you "just compensation." This is true even if the government has a very good reason for taking your property (e.g. - it needs it for vital infrastructure or for a military base). As the Supreme Court famously explained in Armstrong v. United States (1960), the Takings Clause is "designed to bar Government from forcing some people alone to bear public burdens which, in all fairness and justice, should be borne by the public as a whole."

Unfortunately, the U.S. Court of Appeals for the Federal Circuit has recently ruled that this principle does not apply in a case where the government appropriates the property of innocent people as part of of a criminal investigation. In AmeriSource Corp. v. United States (hat tip: my soon-to-be colleague T.J. Chiang), the Court held that there is no taking and no compensation is required in a case "when the government seizes an innocent third party's property for use in a criminal prosecution but never introduces the property in evidence, and it is rendered worthless over the course of the proceedings." In this case, the government seized some $150,000 of drugs belonging to the AmeriSource Corporation for use as evidence in a criminal case against another firm. By the time the government decided that it no longer needed the drugs (which were never actually introduced into evidence), they had been rendered worthless by the passage of their expiration dates. Although AmeriSource was never even accused of any wrongdoing, it will get no compensation whatsoever for the loss of this valuable property.

I. Public Use and the Police Power.

The Federal Circuit's ruling may ultimately be correct under current Supreme Court precedent. But some of its reasoning is extremely dubious. The Court's main argument is that there is no taking here under the Fifth Amendment because, when the government seizes property for use in a criminal investigation, it is exercising its "police power" (which, includes, among other things, law enforcement) not taking property for a "public use," as the text of the Amendment indicates. However, the Supreme Court has ruled in several cases (most notably Berman v. Parker and Hawaii Housing Authority v. Midkiff) that the scope of public use is "coterminous with the scope of a sovereign's police powers" (Midkiff, 467 U.S. 229, 240-41 (1984)). I think that this is an overly broad interpretation of "public use," but it does clearly indicate that the mere fact that a government action involves the police power doesn't mean that it can't also be a taking for a public use. Thus the Federal Circuit is wrong to draw a sharp dichotomy between "public use" on the one hand and "police power" on the other.

II. Supreme Court Precedent.

That said, the Federal Circuit may be on more solid ground in relying on the Supreme Court's extremely permissive jurisprudence on asset forfeitures. For example, in Bennis v. Michigan (1996), The Supreme Court held that there was no taking in a case where Mr. Bennis' car was confiscated because he had engaged in illegal sex with a prostitute in the vehicle. Although Mrs. Bennis was a co-owner of the car and she had not been convicted of any crime, the Supremes held that she wasn't entitled to any compensation for the loss of her interest in the car. I think that Bennis was wrongly decided. But obviously the Federal Circuit had to obey this Supreme Court precedent.

Whether Bennis does in fact determine the outcome of the present case is a close call. Unlike in Bennis, the property seized in AmeriSource was not connected to any wrongdoer whatsoever. Thus, the two cases might be considered different. The Bennis decision did not rule that the absence of any wrongdoing by any of the owners would lead to a different outcome. But it also didn't preclude that possibility. Indeed, Bennis explicitly relied on "a long and unbroken line of cases holds that an owner's interest in property may be forfeited by reason of the use to which the property is put even though the owner did not know that it was to be put to such use." Unlike in Bennis, there was no such illegal use of property by AmeriSource. On balance, therefore, I think that AmeriSource can be distinguished from Bennis, but I don't blame the Federal Circuit too much for failing to do so.

III. Why Compensation Should be Required.

Much more problematic is the failure of the Supreme Court to properly apply the Takings Clause to innocent property owners who have their property appropriated by the state in the course of criminal investigations. Even if the government has a legitimate need for the items in question (e.g. - because they are important evidence), denying compensation is a classic example of "forcing some people alone to bear public burdens which, in all fairness and justice, should be borne by the public as a whole." Apprehension and punishment of criminals is a general public interest and the costs should be paid by the public as a whole, not arbitrarily imposed on individual property owners who were unlucky enough to get caught up in an investigation. There is also no doubt that the government is "using" the property in question. Therefore, there is every reason for the courts to conclude that private property appropriated during a criminal investigation has indeed been taken for "a public use" and that the Fifth Amendment therefore requires compensation for the owners.

Some might fear that the government will be hamstrung by having to compensate all owners of evidence used in investigations. Perhaps the cost will be too great. In reality, however, there is always a cost when the government takes property away from owners. The only question is whether the owners will be forced to bear that cost or whether it will be borne by the public fisc. If the government is forced to pay compensation, it may have stronger incentives to correctly balance the benefits of appropriating the items in question against the costs imposed on innocent owners. It is actually a good thing if the government is deterred from seizing property that is of great value to owners, but perhaps of only marginal use to the prosecutors' case. Currently, the government can afford to ignore the costs imposed on innocent owners completely, unless the latter happen to have a lot of political clout. That is both unfair to property owners and likely to promote the use of investigative tactics whose costs to the innocent outweigh their benefits in promoting the conviction of the guilty.

UPDATE: It is worth noting that these cases seem to bring out some of the worst instincts of both conservative and liberal judges. The liberals tend to support the government because of their general tendency to devalue constitutional property rights. The conservatives do the same because of their general reluctance to support anything that might impede law enforcement. That said, three liberal justices signed on to Justice Stevens' excellent dissent in Bennis. One conservative (Justice Kennedy) also voted that way. Another (Justice Thomas) wrote a concurrence suggesting that the case should be construed relatively narrowly. Hopefully, the Court will limit or overrule Bennis in a future decision. I suppose I should emphasize that the above points about conservative and liberal jurists don't necessarily apply to conservatives and liberals more broadly. For example, many liberals outside the Court were outraged by the Kelo decision, which was supported by all four liberal justices on the Court itself.

UPDATE #2: Economist David Friedman (son of Milton) makes some good points in his comment on this post.

In his interesting post responding to my critique of the Federal Circuit's AmeriSource decision, Orin argues that the Takings Clause does not require compensation in these cases.

Orin claims that "the argument that the Takings Clause applies is hard to make from an originalist perspective [because] Innocent third parties have long had their property rights interfered with in criminal investigations." There is a big difference between having your property "taken for public use" (the term used in the Fifth Amendment) and merely having it "interfered with." A search of a house is not a taking. There is, however, a taking if the government kicks you out of your house and takes control of it - either permanently or for an extended period of time. To my knowledge there is no evidence that the framers and ratifiers of the Bill of Rights accepted the view that there isn't a taking in cases where the government actually seizes the property of innocent third parties during a criminal investigation, as opposed to merely "interferes" with it. As the Court explained in Bennis, there were some early cases when the courts tolerated a seizure of property that had been used in criminal activity by a person whom the owner had entrusted it to. But that is very different from the seizure of property that the owner had never entrusted to a wrongdoer who went on to use it to commit a crime.

I recognize, of course, that the line between taking and interference may sometimes be fuzzy. But that doesn't mean that there aren't cases that clearly fall on one side or the other. Seizing someone's property and holding it for months (as happened in AmeriSource) is clearly a taking. Since I a more a textualist than an originalist, I think it highly significant that the text of the Fifth Amendment does not distinguish between the taking of private property for the "public use" of investigating crime and other kinds of takings. All require the payment of "just compensation." Since there were no professional police forces or extensive government evidence-gathering in the 1780s, the Founders may never have been confronted with a situation like that in AmeriSource; as a result we may never be able to discover their specific opinions on the matter (if indeed they had any). However, the text of the Fifth Amendment is more than broad enough to cover these cases. And it is the text, not the subjective intentions of the Framers, which is the law.

Orin also contends that "allowing such claims under the Takings Clause would be quite difficult to administer" because property rights of innocent people are so often "interfered with" during investigations. Part of this concern may be mitigated by the distinction between takings and interference discussed above. However, I don't deny that there will be at least some administrative problems. But the government can mitigate these by establishing a system of administrative compensation that can, in most cases, obviate the need for litigation. As a practical matter, property owners are unlikely to sue for recovery of small losses because of the cost of litigation. Some administrative costs will arise anyway. But that is an inevitable consequence of the enforcement of any constitutional right. For example, as Orin knows better than I, Fourth Amendment enforcement raises all kinds of difficult administrative problems arising from the difficulty of determining whether the police had adequate reasons for searching a particular location at the time they did it.

That said, Orin and I might well agree on the issue of administration more than we disagree. Orin writes:

I tend to agree with Ilya that the government should provide compensation in cases such as that raised by the Federal Circuit case. Plus, a statutory regime could be written to make it much more administrable.

I think that Orin is right on both of these counts. A statutory regime of compensation would be highly desirable and might well be enough to provide adequate compensation to property owners. It could also obviate some of the administrative difficulties of case by case litigation. However, it is unlikely that the government will have any incentive to enact such a scheme unless the judiciary starts enforcing the "just compensation" requirement of the Takings Clause in these cases. If Orin and I are right to believe that the administrative difficulties can be overcome by a statutory or administrative compensation procedure, then judicial enforcement might help facilitate that happy outcome.

Finally, Orin argues that my position - if adopted by the Court - would lead to a narrowing of the exclusionary rule. This aspect of the matter is beyond my expertise, so I can't comment on it intelligently.

Unfortunately, I will be on the road most of tommorrow, so may not be able to continue this debate. However, I may return to it on Saturday if additional points occur to me at that time.

Orin makes some interesting new points in his latest post in our debate. But I remain unpersuaded. To review, I argued that when the government seizes the property of innocent people during a criminal investigation (e.g. - to use as evidence), the government is required to pay compensation under the the Fifth Amendment's Takings Clause, which requires the payment of "just compensation" whenever "private property" is "taken for public use."

In his most recent post, Orin claims that my interpretation of the Takings Clause fails to take account of the Fourth Amendment, which protects homeowners against "unreasonable" searches and seizures:

From a textualist standpoint, the Fourth Amendment has a very explicit regulation on searches and seizures; its language was explicitly addressed to the rules that govern the "seizing" of "effects."

Given this language, it seems odd to construe the one phrase at the very end of the Fifth Amendment — "nor shall private property be taken for public use, without just compensation" — as implicitly providing another limitation of "The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures" found in the Fourth Amendment. Further, the word "use" implies action, not just the passive act of storing something away over time (which, on the other hand, is clearly a seizure). Given that, I just don't think Ilya's reading is a natural one.

To my mind, there is no tension between the Fourth Amendment and my interpretation of the Takings Clause. Both protect property rights to some degree, but in different ways and against different threats. The Fourth Amendment forbids "unreasonable" searches and seizures even if compensation is paid. However, the text also assumes that there are at least some "reasonable" seizures of the property of innocent people that are not categorically forbidden; the Fourth Amendment says nothing about the rules governing those situations. In such cases, property rights are protected (to some degree) by the Fifth Amendment's requirement of "just compensation." Thus, the Fifth Amendment protects innocent property owners in precisely those cases where the Fourth does not. I don't know if this synergy was deliberately intended by the framers. But it is certainly a natural and logical interpretation of the text.

To recap, the Fourth Amendment protects us against unreasonable seizures by categorically banning them, but says nothing about what happens if the seizure isn't unreasonable. The Fifth Amendment, meanwhile, protects us against "reasonable" seizures that are severe enough to constitute a "taking" of property for "public use" by requiring the payment of compensation. That way, the government is allowed to take property for use in a criminal investigation when it is reasonable to do so. But it is not allowed to arbitrarily impose the cost on innocent property owners. As I noted in my very first post in this sequence, this is a classic example of one of the main purposes of the Takings Clause: "bar[ring] Government from forcing some people alone to bear public burdens which, in all fairness and justice, should be borne by the public as a whole." Armonstrong v. United States (1960). It seems to me perfectly logical to have one Amendment that forbids "unreasonable" seizures, and another that protects people against uncompensated takings of property (including some types of "reasonable" seizures).

Orin also argues that "public use" implies "action," not just "the passive act of storing something away over time." Perhaps. But seizing property and keeping it away from its owner is certainly an action. So is the act of moving the items to a government-owned facility for storage and locking it up (which certainly doesn't seem "passive" to me). And going through the stored property in order to evaluate its potential usefulness as evidence is even more clearly an action that requires "using" that property. Moreover, even "passive" storage of an item for potential future use is plausibly interpretable as "using" it as well. For example, when I store away my second pair of glasses, I am "using" it as a backup for my primary pair. If I had seized that backup pair from the rightful owner, it would surely be a "taking" for "Ilya's use." If the government does the same thing, it's a "taking" of the glasses for "public use."There may be some cases where the government seizes property for a criminal investigation in such a way that it doesn't "use" it at all; but such situations are likely to be rare if they exist at all.

I do not mean to suggest that any government use of property in an investigation requires compensation. As I noted in an earlier post, many such uses do not infringe on property rights severely enough to constitute a "taking" of that property (e.g. - a search that does not require the government to seize the owner's property or take exclusive possession of it for any significant length of time).

Finally, Orin disputes my claim that the framers may have been unfamiliar with seizures of property owned by innocent third parties in the course of criminal investigations. He notes that they obviously did know about cases involving the forcible return of stolen goods that had found their way into the hands of innocent third parties. Orin himself suggests the logical response: "the government is acting in a very different capacity when it is retrieving stolen goods than when it is collecting evidence that just happens to belong to a third party."

To elaborate slightly on Orin's point: if the goods are stolen, the innocent third party is not the legitimate owner of those goods, and she or she isn't entitled to compensation under the Fifth Amendment because she didn't have any right to the goods in the first place. Orin also notes that even the original owner was not entitled to compensation for the time when the government had his property in its custody prior to returning it. But there is a clear difference between a case where a thief steals someone's property and the government then takes it back from the thief in order to return it to the rightful owner, and one where the government itself takes the property from an innocent person in order to use it for its own purposes. The former scenario is not a "taking" from the rightful owner (at least not by the government), and also doesn't involve any kind of "public use" of the property by the state.

Orin says that this commonsense distinction is inconsistent with this 1967 Warren Court decision. Possibly. I'm not sure if Orin means to endorse the Warren Court's reasoning or not. But either way, a case decided by the decidedly nonoriginalist Warren Court in the 1960s says very little of relevance to the original meaning of the Fourth and Fifth Amendments in 1791. It is particularly irrelevant to the Takings Clause, since the case in question does not seem to consider that clause at all.

As I noted in my previous post, my argument is primarily textual rather than originalist. I am not claiming that the Framers specifically envisioned the approach I advocate. I merely suggest that they did not intend to preclude it, and may not even have considered the specific point at issue, given that the problem it addresses was unlikely to be a common one in their time.

I think my debate with Orin over the Takings Clause and seizures of property during criminal investigations is likely to enter the realm of diminishing returns soon, if it hasn't already. So in this post I will merely recap my key points and respond briefly to Orin's latest post.

I argued in my original post that the Takings Clause, which requires "just compensation" for "takings" of "private property" for "public use" requires compensation for the taking of innocent persons' property during a criminal investigation. The textual basis for this is very simple: that the Fifth Amendment does not distinguish between takings of property for use in a criminal investigation and takings for other public uses. The text simply states: "nor shall private property be taken for public use, without just compensation." There is no exception for takings related to criminal investigations or any other takings of any kind. Thus, it is reasonable and natural to assume that the text requires compensation for all takings alike, whether they occur during criminal investigations or not. Orin then argued that this interpretation of the Fifth Amendment is inconsistent with the text of the Fourth Amendment, and I tried to explain why there is no contradiction: the Fourth and Fifth Amendment both protect property rights, but in different ways and against different threats.

Now, Orin claims that my latest post merely shows that it is possible to interpret the Fifth Amendment as requiring the government to compensate innocent property owners for takings that occur during criminal investigations, not that such an interpretation is required. I'm not sure I understand his post correctly. But it seems to me that it conflates my response to his Fourth Amendment point with the whole of my argument. If I am correct in that latest post in arguing that there is no contradiction between my interpretation of the Fifth Amendment and the functions of the Fourth Amendment, then my textualist approach to the Takings Clause stands (at least against Orin's objections): it is logical and natural to interpret it as applying to all takings of private property for public use. There is no implicit, nontextual exception for takings that occur during criminal investigations. Thus, compensation for such takings is not only permitted but required by the text. My core argument is simply that the Takings Clause says what it means and means what it says. It applies to all takings of private property for public use, not just some subset of them.

Orin also interprets my statement that the framers did not "intend" my reading as suggesting that it is merely a possible reading rather than the correct one. However, I have from the outset stated that my argument is textualist, not originalist. The sole originalist point I sought to make was that there is no compelling originalist evidence against my interpretation. That point is merely a negative defense against Orin's claim that originalism counts against my argument. My affirmative argument is textual, not originalist, and I respectfully suggest that Orin hasn't - at least so far - succeeded in refuting it.

To put the point another way, I doubt that the framers ever considered the specific problem of takings of the property of innocent people during a criminal investigation. At the same time, however, the text they enacted is clearly broad enough to cover this situation, just as it is broad enough to cover many other types of takings that were rare or even completely unknown in their day. Consider, for example, the taking of property for the purpose of building an airport. Similarly, when they drafted the First Amendment, the Framers could not and did not consider the regulation of speech on television and radio. But that does not mean that the Free Speech Clause can't apply to regulation of speech on broadcast media.