The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
On Monday, Magistrate Judge James Orenstein in Brooklyn ruled that Apple cannot be ordered under the All Writs Act to unlock an iPhone pursuant to a search warrant. Importantly, this is not the California case where the FBI wants Apple to create a software update to help the government guess passwords to unlock the iPhone used by the San Bernardino attackers. This case involves an earlier iPhone that Apple can unlock. The question is whether the All Writs Act allows a court to order a phone manufacturer to unlock a phone. Some courts have ruled that the answer is "yes," but Orenstein ruled that the answer is "no."
This post analyzes Orenstein's central argument in the case. It concludes that Orenstein's core argument is very weak. Orenstein treats the statutory language as if no court had interpreted it before, and he then develops a new theory of what the language should mean. Orenstein then applies his theory to reject the government's application. But it turns out that the U.S. Supreme Court has already interpreted that language, and the Supreme Court's interpretation is very different from the one Orenstein adopts.
Here are the details. Orenstein's opinion analyzes a 1789 statute, the All Writs Act, which states:
The Supreme Court and all courts established by Act of Congress may issue all writs necessary or appropriate in aid of their respective jurisdictions and agreeable to the usages and principles of law.
Orenstein ruled that Apple cannot be ordered to assist the government because that assistance would not be "agreeable to the usages and principles of law." Here's his argument as best I understand it, which I will break down into seven steps:
1) Each of the words in the phrase "agreeable to the usages and principles of law" must have independent meaning because courts should assume that no word in a statute is superfluous.
2) The "principles" of a law should be read to mean the manner in which the laws were developed, while the "usages" of a law refers to the manner in which the laws have been interpreted and implemented.
3) When Congress considers passing a law that would grant an authority, but Congress rejects that law for reasons that in context make clear that Congress was against that new law, the congressional rejection of that authority makes that authority contrary to the "usages and principles of law" because Congress has made clear it doesn't want that law.
4) The scope of powers granted to courts under the All Writs Act therefore is cut back when Congress considers new legislation and rejects it in ways that make clear Congress didn't want that authority. Courts can and must look at the legislative record about laws it didn't pass, which includes bills introduced and hearings held, to ascertain the genuine congressional will.
5) Any other reading of the "usages and principles" phrase might render the All Writs Act unconstitutional because it would "upend the separation of powers by delegating to the judiciary a legislative power bounded only by Congress's superior ability to prohibit or preempt." Such a reading would "thoroughly trample" on Article I, and it is hard to believe that the First Congress—which had 17 members who were delegates to the Constitutional Convention—would do that. To save the constitutionality of the All Writs Act, the phrase must be interpreted to limit the judicial power to issue writs in support of the executive branch when Congress considers but rejects new legislation expanding executive power.
6) As explained in Orenstein's October 2015 order, Congress has at various times over the past 25 years considered but not enacted a law that would require providers to give plain-text access to the government in circumstances such as this case. Congress has had hearings about the issue as recently as last year. Reviewing the legislative record, it is clear that Congress has chosen not to confer the powers the government is seeking.
7) Because Congress has considered but not passed a law requiring access in such cases, an order requiring assistance to help grant that access is not "agreeable to the usages and principles of law" and violates the All Writs Act.
I'm not persuaded. I have issues with several of the steps Orenstein takes, but let me focus on just one: It appears that the Supreme Court has already interpreted the phrase "agreeable to the usages and principles of law," and the Supreme Court's interpretation bears no resemblance to the interpretation developed by Orenstein.
Some background may be helpful. Before the 20th century, there were important limits on the particular forms of orders ("writs") that courts could issue. First, there were the common law writs from England. Next, there were writs that state courts had developed before 1789. And third, there were writs that state courts developed after 1789. Before the 20th century, the details of writs were a really big deal. Writs could be incredibly specific, and the details had to be right for the writ to be lawful. We mostly forget about the details of the old writs because the legal system has long since moved on. These days, the precise forms of the old writs are just a historical curiosity.
That wasn't the case in the 19th century, though, and that led to litigation on the meaning of the phrase "agreeable to the usages and principles of law" in the All Writs Act. The big question was which writs could be issued by federal courts. Were federal courts limited to the forms of the common law writs? Alternatively, could they add in state writs from pre-1789 or even post-1789? Or could they even add new writs not recognized by the states as they were deemed necessary by the federal judges themselves? In a series of cases, the Supreme Court interpreted the phrase "agreeable to the usages and principles of law" as answering that question.
Bank of the United States v. Halstead, 23 U.S. (10 Wheat.) 51 (1825) is a useful starting point. Halstead considered what steps a federal judge could take to order U.S. marshals to sell property owned by a debtor that was needed to satisfy a federal court judgment. The trial judge had used the All Writs Act to issue a writ known as "venditioni exponas," which was a writ to sell seized goods held as part of a legal judgment for the best available price after the property didn't sell earlier. A new state law prohibited the sale if the market price was too low, and the question in the case was whether the venditioni exponas writ had to be modified by the state law because otherwise the writ would not be "agreeable to the usages and principles of law."
In ruling that the trial judge's writ complied with the All Writs Act, the Supreme Court offered this interpretation of the phrase "agreeable to the principles and usages of law" (with emphasis in original):
The precise limitations and qualifications of this power, under the terms, agreeable to the principles and usages of law, is not, perhaps, so obvious. It doubtless embraces writs sanctioned by the principles and usages of the common law. But it would be too limited a construction, as it respects writs of execution [that is, orders to sell property to satisfy a judgment], to restrict it to such only as were authorized by the common law. It was well known to Congress, that there were in use in the State Courts, writs of execution, other than such as were conformable to the usages of the common law. And it is reasonable to conclude, that such were intended to be included under the general description of writs agreeable to the principles and usages of law. If it had been intended to restrict the power to common law writs, such limitation would probably have been imposed in terms.
According to the Court, the All Writs Act authority to issue writs "agreeable to the usages and principles of law" implied a broad authority to add new forms of writs as necessary:
To limit the operation of an execution now, to that which it would have had in the year 1789, would open a door to many and great inconveniencies, which Congress seems to have foreseen, and to have guarded against, by giving ample powers to the Courts, so to mould their process, as to meet whatever changes might take place.
If Congress didn't like how the courts "mould[ed] their process," Congress could overturn those decisions with specific legislation:
[Federal courts] have authority, therefore, from time to time to alter the process, in such manner as they shall deem expedient, and likewise to make additions thereto, which necessarily implies a power to enlarge the effect and operation of the process. The exercise of this power is, to be sure, left in the discretion of the Court; but the object and purpose for which it is given, is so plainly marked, that it is hardly to be presumed the Courts would omit carrying it into execution, without some substantial reason. And, the better to insure this, authority is given to this Court, to prescribe to the Circuit and District Courts, such regulations on the subject as it shall think proper. And should this trust not be duly and discreetly exercised by the Courts, it is at all times in the power of Congress to correct the evil by more specific legislation.
The Court also rejected the argument that giving courts this broad authority violated the separation of powers by improperly delegating legislative power to the courts:
It is said, however, that this is the exercise of legislative power, which could not be delegated by Congress to the Courts of justice. But this objection cannot be sustained. There is no doubt that Congress might have legislated more specifically on the subject, and declared what property should be subject to executions from the Courts of the United States. But it does not follow, that because Congress might have done this, they necessarily must do it, and cannot commit the power to the Courts of justice. Congress might regulate the whole practice of the Courts, if it was deemed expedient so to do: but this power is vested in the Courts; and it never has occurred to any one that it was a delegation of legislative power. The power given to the Courts over their process is no more than authorizing them to regulate and direct the conduct of the Marshal, in the execution of the process. It relates, therefore, to the ministerial duty of the officer; and partakes no more of legislative power than that discretionary authority in trusted to every department of the government in a variety of cases.
The Court offered a somewhat similar interpretation in Riggs v. Johnson, 73 U.S. (6 Wall.) 166 (1867).
As I read Halstead, Riggs and other similar cases, the Supreme Court interpreted the phrase "agreeable to the usages and principles of law" to be about what kinds of writs were proper. Halstead's answer seems to be that federal courts have broad authority to issue writs beyond the forms at common law. The court read that phrase in the All Writs Act as giving courts "ample powers … to mould their process, as to meet whatever changes might take place." If courts went astray with this discretionary power, "it is at all times in the power of Congress to correct the evil by more specific legislation."
As far as I know, this line of cases has not been overruled by subsequent Supreme Court caselaw. After the federal courts rejected the precise forms of the old writs, courts paid less attention to the "usages and principles" phrase. There isn't a lot of 20th-century caselaw on it. It's like a vestigial limb left over from the era of common law writs. Where the "usages and principles" language has come up in modern caselaw, it has mostly been interpreted to mean that an assistance order and the order that it is supporting cannot violate any statute or constitutional provision. Consider the Sixth Circuit's discussion in United States v. Perry, 360 F.3d 519, 533 (6th Cir. 2004):
At first glance, the All Writs Act seems to authorize a district court to issue any subsequent order in furtherance of an initial order as long as the initial order was lawful. As one would expect, however, even if a court properly issues its initial order, it may not use the All Writs Act to issue a subsequent order to effectuate the first order if the subsequent order is itself unconstitutional. See, e.g., United States v. City of New York, 972 F.2d 464, 470 (2d Cir.1992) ("Nor can the All Writs Act confer on the courts the power to ignore the case or controversy requirement, which is rooted in Article III of the constitution's definition of judicial power."); In re Baldwin-United Corp. (Single Premium Deferred Annuities Ins. Litig.), 770 F.2d 328, 340 (2d Cir.1985) ("The All-Writs Act … cannot be used to circumvent or supersede the constitutional limitations of the Eleventh Amendment."). This is consistent with the text of the All Writs Act, which only authorizes courts to issue writs "agreeable to the usages and principles of law." 28 U.S.C. § 1651. "[A]greeable to the usages and principles of law" also suggests that courts may not use the All Writs Act to issue a second order to effectuate an initial, lawful order if the second order works by violating some other statutory provision.
This seems consistent with Halstead and Riggs, in that a writ that violates another law will run into the more specific legislation that Congress has enacted. And as Halstead also suggests in another part of the opinion, Congress can't permit an unconstitutional order, so the All Writs Act can't be used to enforce an unconstitutional order or to circumvent constitutional limits. The broad interpretation from Halstead also appears to have lived on in modern language from the Supreme Court that the All Writs Act can be used flexibly. As the Court stated in New York Telephone in 1977, "unless appropriately confined by Congress, a federal court may avail itself of all auxiliary writs as aids in the performance of its duties, when the use of such historic aids is calculated in its sound judgment to achieve the ends of justice entrusted to it." (quoting Adams v. United States ex rel. McCann, 317 U. S. 269, 273 (1942)).
These cases suggest to me that Orenstein was not free to create a new theory of the phrase "agreeable to the usages and principles of law." Orenstein's theory of what the phrase means is very interesting. But however interesting it is, it appears contrary to the precedents interpreting that language.
Orenstein touches on this issue only in a footnote. Here's what he says:
The Supreme Court wrote in 1952 that, "[i]n determining what auxiliary writs are 'agreeable to the usages and principles of law,' we look first to the common law." United States v. Hayman, 342 U.S. 205, 221 n.35 (1952) (quoting Price, 334 U.S. at 281). But that observation—made in the context of determining whether a particular type of writ of habeas corpus was available under common law, and therefore under the AWA—has no bearing on the instant dispute. Similarly, the Tenth Circuit's citation to Hayman for the same point in the context of determining the availability of a collateral challenge to a criminal conviction also offers no interpretive assistance here. See Rawlins v. Kansas, 714 F.3d 1189, 1196 (10th Cir. 2013) (citing Hayman, 342 U.S. at 221 n.35). In both cases, the discussion of the "usages and principles" clause was largely directed to identifying the types of writs that could be so described. Apple does not object that the type of assistance the government seeks here cannot find a close enough antecedent in the common law; the issue here—which Hayman and Rawlins had no need to address—is whether the issuance of such a writ is an appropriate application of the AWA's gap-filling function.
I don't think this argument works. To Orenstein, the caselaw on the "usage and principles" phrase is irrelevant because it was only about what kind of writs could be issued. That's not the question Orenstein is concerned with, so he says the caselaw "has no bearing on the instant dispute." But when the Supreme Court interprets statutory language to have a particular meaning, I don't think a magistrate judge can say that the caselaw is irrelevant because he's not interested in that issue. It's especially hard to start from first principles with a new theory of the phrase when the Supreme Court has already interpreted it.
Even putting that aside, Halstead does appear to address the gap-filling function of the All Writs Act in interpreting the "usage and principles" phrase. It interpreted the phrase to give courts "ample powers … to mould their process, as to meet whatever changes might take place," and it placed the burden on Congress to limit those powers "by more specific legislation." Maybe later Supreme Court opinions have modified that guidance; Halstead is almost 200 years old. It is still cited on occasion—see this Fifth Circuit opinion from 2000, which also cites Riggs—but maybe the details of its guidance have been altered by intervening caselaw. But however that should be resolved, I think it's too late to approach the "usage and principles" phrase as a tabula rasa. If there's a way to square Orenstein's reading with cases like Halstead and Riggs, it's not clear to me. Perhaps readers have some ideas.
Finally, I'm happy to admit that this particular argument in Orenstein's opinion was only the lead argument among several that he offers for why he rejected the government's application. It was the holding, but he indicated in extensive dicta that he had many other grounds for reaching the same result if his lead argument failed. I'm trying to finish and send out an article this week, so unfortunately I don't have time to delve into the details of the other arguments made (which may or may not be persuasive). Perhaps I'll return to the other issues next week when my article is out.