Don't be a visionary

Why Judge Cannon should have gone with the flow.

On July 15, 2024, Judge Cannon granted Donald Trump’s motion to dismiss the indictment against him, holding that the appointment of Trump’s prosecutor, Jack Smith, was not authorized by statute.

I admire Judge Cannon’s opinion, which is well-written, thoughtful, and thorough. Contrary to some of the commentary I’ve read, it’s not a crazy opinion at all. That said, I disagree with the ruling and hope it is reversed by the Eleventh Circuit.

(It is surprisingly difficult to create AI images depicting a federal judge’s determination that a special prosecutor’s appointment was not authorized by statute. With that in mind, this week’s post will feature images of Jack Smith packing up his office.)

I (can’t) do what I want.

Attorney General Merrick Garland was appointed by the President and confirmed by the Senate. So was Markenzy Lapointe, the U.S. Attorney for the Southern District of Florida. But Jack Smith was not. Instead, on November 18, 2022, Attorney General Garland appointed him to serve as Special Counsel in charge of prosecuting Trump.

Unlike Lil Uzi Vert, Attorney General Garland cannot simply do what he wants. Article II, Section 2, Clause 2 of the Constitution, a/k/a the Appointments Clause, says that the President “shall nominate, and by and with the Advice and Consent of the Senate, shall appoint … Officers of the United States, whose Appointments are not herein otherwise provided for, and which shall be established by Law: but the Congress may by Law vest the Appointment of such inferior Officers, as they think proper, in the President alone, in the Courts of Law, or in the Heads of Departments.”

Attorney General Garland is a “Head of Department.” Smith contends¹ he is an “inferior Officer.” If that’s so, then Attorney General Garland had the authority to appoint him—but only if “Congress,” “by law,” “vest[ed] the Appointment” in the Attorney General. Translated into English, there has to be a federal statute that authorized Attorney General Garland to appoint Smith.

Trump moved to dismiss the indictment, arguing that no such statute existed. This may seem like a technicality, but I can’t blame him for giving the argument a shot; making arguments based on technicalities is how I am putting my kids through college.

Judge Cannon agreed with Trump and dismissed the indictment, finding that no statute authorized the appointment of Smith. In the view of Judge Cannon, Smith’s appointment violated the Appointments Clause (because the Appointments Clause requires Congress to authorize the appointment of an Officer, and Congress never did), and also violated the Appropriations Clause (because Congress didn’t appropriate funds for a prosecutor who wasn’t lawfully appointed).

I don’t agree. Smith makes three arguments that I find compelling, any of which would be sufficient to prevail. These arguments are as follows:

• 28 U.S.C. § 515(b) authorized Smith’s appointment.

• 28 U.S.C. § 533(1) authorized Smith’s appointment.

• The Supreme Court decided 50 years ago that Smith’s appointment was authorized.

Can a lawyer be “specially retained” without having already been “specially retained”?

Let’s start with 28 U.S.C. § 515(b). Here’s what it says:

(b) Each attorney specially retained under authority of the Department of Justice shall be commissioned as special assistant to the Attorney General or special attorney, and shall take the oath required by law. Foreign counsel employed in special cases are not required to take the oath. The Attorney General shall fix the annual salary of a special assistant or special attorney.

(Yes, I skipped over 28 U.S.C. § 515(a). I’ll get to that.)

Did this statute give the Attorney General the authority to appoint Smith as Special Counsel? It’s ambiguous, in my opinion.

On the one hand, it refers to an attorney being “specially retained under the authority of the Department of Justice” and serving as “special assistant to the Attorney General.” This implies that the Department of Justice—which is run by the Attorney General—has the power to specially retain an attorney to serve as a special assistant to the Attorney General.

On the other hand, the statute lacks the pristine clarity of, say, Lake Tahoe. It doesn’t say anything about appointments and it doesn’t say anything about officers. Also, it seems to be talking about attorneys who have already been “specially retained under authority of the Department of Justice,” rather than affirmatively granting authority to specially retain an attorney.

Here is 28 U.S.C. § 543(a), by comparison: “The Attorney General may appoint attorneys to assist United States attorneys when the public interest so requires, including the appointment of qualified tribal prosecutors and other qualified attorneys to assist in prosecuting Federal offenses committed in Indian Country.” That’s Lake Tahoe. But it doesn’t apply here because Smith isn’t assisting a U.S. Attorney.

Juxtaposing Section 543(a) with Section 515(b) suggests that Section 515(b) doesn’t confer appointment authority. On the other hand, these sorts of comparisons are often uninformative. Laws are enacted at different times, with different words, for different reasons. It’s not a huge shock that some laws would be clearer than others.

I think we have no choice but to look at the history of Section 515(b). Ordinarily I hate looking at history when interpreting statutes … as someone who is both lazy and an ardent textualist, I think you should be able to figure out what a statute means by reading it. But Section 515(b) is obscure enough that unfortunately, you need to put in the work.

Smith makes three historical arguments that, taken together, are strong:

• When the original version of this statute was enacted in 1870, it wouldn’t have made sense unless Smith’s interpretation is correct.

• The circumstances surrounding the enactment of Section 515(a) strongly support Smith’s argument.

• Attorneys General have repeatedly appointed special prosecutors over the years whose appointments would have been illegal if Smith’s argument is wrong.

Let’s go through these one by one.

• When the original version of this statute was enacted in 1870, it wouldn’t have made sense unless Smith’s interpretation is correct.

First, the original version of Section 515(b) was enacted in 1870 as part of the Act to Establish the Department of Justice. Yup, before 1870 the Department of Justice didn’t exist. There were “district attorneys”—what we call U.S. Attorneys today—but no centralized DOJ.

(Before 1870, the office of Attorney General did exist, but it was a part-time job. The job was so unimportant that, until 1853, there was actually a statute saying that the Attorney General would get paid less than other members of the Cabinet. Those were the days.)

Anyway, when Congress created the Justice Department in 1870, it said this:

And every attorney and counselor who shall be specially retained, under the authority of the Department of Justice, to assist in the trial of any case in which the government is interested, shall receive a commission from the head of said Department, as a special assistant to the Attorney General, or to some one of the district attorneys, as the nature of the appointment may require, and shall take the oath required by law to be taken by the district attorneys, and shall be subject to all the liabilities imposed upon such officers by law.

This is a bit different from the current version of Section 515(b), but close enough: it has the same “specially retained under the authority of the Department of Justice” language and refers to “a special assistant to the Attorney General.”

In 1870, no other statute existed that would have authorized the appointment of a special assistant to the Attorney General. Thus, Smith’s argument goes, the “specially retained” language must have been intended to confer that authority, rather than refer to a lawyer who was already “specially retained” under a different statute.

It’s a good argument. Not dispositive. But a good start.

• The circumstances surrounding the enactment of Section 515(a) strongly support Smith’s argument.

Smith’s second argument has to do with Section 515(a). Around the turn of the 20th century, merchants in New York became concerned about “fraudulent importations of Japanese silks.” So they asked the Attorney General to investigate. The Attorney General said he didn’t have the money to conduct an investigation. So the merchants suggested that their own private lawyer be appointed as a special assistant to the Attorney General, and helpfully offered to pay for his expenses. (Conveniently, the private lawyer also had the last name Smith, although his first name—“Wickham”—was much cooler.) Amazingly, the Attorney General was fine with this arrangement, and so in 1902, he appointed Wickham Smith to be a Special Assistant to the Attorney General. Wickham Smith then went ahead and obtained grand jury indictments against the purported fraudsters, gentlemen by the name of Rosenthal and Cohn.

To me, this looks like an insane conflict of interest—merchants are paying a lawyer to obtain grand jury indictments against their direct competitors! It looked that way to the federal judge, too, and he dismissed the indictments. See United States v. Rosenthal, 121 F. 862 (C.C.S.D.N.Y. 1903).

Here was the judge’s theory. At the time, the 1870 version of the DOJ Act was still in place. Remember, here’s what it said: “every attorney and counselor who shall be specially retained, under the authority of the Department of Justice, to assist in the trial of any case in which the government is interested, shall receive a commission from the head of said Department, as a special assistant to the Attorney General.” The judge reasoned that this statute conferred authority on special assistants to “assist in the trial of any case,” but didn’t confer authority to obtain grand jury indictments.

In 1906, Congress responded to Rosenthal by enacting the statute currently codified at Section 515(a). In its modern form, here’s what it says:

The Attorney General or any other officer of the Department of Justice, or any attorney specially appointed by the Attorney General under law, may, when specifically directed by the Attorney General, conduct any kind of legal proceeding, civil or criminal, including grand jury proceedings and proceedings before committing magistrate judges, which United States attorneys are authorized by law to conduct, whether or not he is a resident of the district in which the proceeding is brought.

The new statute unambiguously confers the power that the Rosenthal court said was lacking. And, although I realize that in some circles quoting legislative history is equivalent to summoning Satan himself, the legislative history of the 1906 statute clearly states that Congress’s purpose was to overturn Rosenthal: “The law proposed by the bill under consideration seems to be very necessary, because of the decision in the Rosenthal case."

Why am I telling you all of this?

Well, it supports a very strong congressional ratification argument. 99% of the time, I dislike these sorts of arguments. Courts shouldn’t be interpreting statutes by drawing delicate inferences from later-in-time lower-court decisions and subsequently-enacted amendments. If you want to know what a statute means, just read it.

But this case just might be in the 1%. The 1906 amendment giving rise to what’s now Section 515(a):

• Presupposes that Section 515(b) authorizes the Attorney General to hire Special Assistants, and

• Expands the scope of the Special Assistants’ authority.

It would be weird for a court to hold, 118 years later, that actually, the authority that Congress assumed to exist doesn’t exist.

• Attorneys General have repeatedly appointed special prosecutors over the years whose appointments would have been illegal if Smith’s argument is wrong.

As the Rosenthal case makes clear, Attorneys General were appointing Special Assistants as far back as the Theodore Roosevelt administration. Theodore Roosevelt’s Attorney General appointed Special Assistants to serve as prosecutors in some other famous cases as well, like the Oregon land fraud case that led to most of Oregon’s congressional delegation being indicted. There are also lots of old decisions noting that a Special Assistant to the Attorney General brought the case, and in some cases rejecting challenges to the Special Assistant’s legal authority. See, e.g., Ewert v. Bluejacket, 259 U.S. 129, 133 (1922); May v. United States, 236 F. 495, 498 (8th Cir. 1916). Fast-forwarding a little bit, there’s also Leon Jaworski, who was the special prosecutor during Watergate, as well as Robert Mueller, who you all know.

As far as I can tell, if Smith’s appointment is illegal, then all of these appointments would have been illegal, too.

This historical practice might be relevant for a few reasons:

• If lots of attorneys general have assumed that they enjoy the power to appoint Special Counsel, this is evidence of what the ambiguous words of Section 515(b) mean.

• If lots of attorneys general have exercised this power, and Congress hasn’t tried to stop them, this is evidence that Congress has acquiesced in that power.

• Traditions are worth keeping for their own sake. If you don’t love tradition, you don’t love America.

People vary on whether they find these types of arguments persuasive. From my perspective, I wouldn’t say the historical practice decides the case, but it’s a point in Smith’s favor.

Enough about what I think

To me, the points above, taken together, are a sufficient basis to interpret the ambiguous words of Section 515(b) in Smith’s favor. So how does Judge Cannon reach a contrary conclusion?

The bulk of Judge Cannon’s analysis focuses on the literal meaning of Section 515(b). Judge Cannon holds that “Section 515(b), read plainly, is a logistics-oriented statute that gives technical and procedural content to the position of already-‘retained’ ‘special attorneys’ or ‘special assistants’ within DOJ.” After conducting a close textual analysis, Judge Cannon concludes that the statute “transmit[s] the fairly mundane, descriptive point that already-hired attorneys within the Department shall be classified as special assistants or special attorneys and shall take an oath and have a fixed salary.”

If one stares only at the text, I think that Judge Cannon’s interpretation is reasonable, but Smith’s interpretation is reasonable too. So you have to look at the history. If Judge Cannon’s interpretation reflected what the statute unambiguously said, I would ignore the history and agree with her. But I can’t get there.

Judge Cannon then rejects Smith’s historical arguments. She says that the 1870 version of the statute isn’t any more favorable to Smith than the modern version, and disagrees that Congress ever “ratified” Smith’s interpretation when the actual wording of the statute never changed. The court doesn’t really engage with what are, in my view, Smith’s two strongest arguments: the argument that there wasn’t any other source of hiring authority in 1870 for special assistants to the Attorney General, and the ratification argument based on the circumstances of Section 515(a)’s enactment.

Judge Cannon also disagrees that there’s a consistent historical practice of appointing special assistants under Section 515(b). She points out that in the past, special attorneys were in some cases (1) appointed by the President, or (2) appointed pursuant to special legislation. She is correct that those examples do not support Smith’s historical argument. But there are lots of other historic examples of appointments that were by the Attorney General and couldn’t have been authorized by any statute other than Section 515(b) (or Section 533, which I’ll get to below).

With regard to those other historic appointments, Judge Cannon points out that many of the appointment orders didn’t cite Section 515(b), and some of them didn’t cite any statute at all. Her implication is that there’s no historic understanding that Section 515(b) authorizes the appointment of a special counsel, when prior attorneys general apparently didn’t even realize they were relying on Section 515(b). She scores some points here, but ultimately I’m not persuaded that one can psychoanalyze what the Attorney General was thinking via the text of an appointment order. There’s no law that says an appointment order has to recite the statute that the Attorney General thinks he’s relying on.

Finally, Judge Cannon emphasizes that many, perhaps most, prior Special Counsels were appointed from within the government and had less power than Smith. Perhaps so. But regardless of where they were hired from and how much power they had, the Attorney General couldn’t have retained them as Special Counsel unless he had the statutory authority to do so. Judge Cannon’s view is that Section 515(b) gives the Attorney General no power to “specially retain” someone, which I think is incompatible with a fair number of historical examples.

I’m not saying it’s open and shut. The language of Section 515(b) could certainly have been stronger. But, viewed in conjunction with events of the last 154 years, I think it’s strong enough.

Prosecutors prosecute

Smith offers one other candidate for a statute that authorizes his appointment: 28 U.S.C. § 533. It says this:

The Attorney General may appoint officials—

(1) to detect and prosecute crimes against the United States;

(2) to assist in the protection of the person of the President; and

(3) to assist in the protection of the person of the Attorney General.

(4) to conduct such other investigations regarding official matters under the control of the Department of Justice and the Department of State as may be directed by the Attorney General.

This section does not limit the authority of departments and agencies to investigate crimes against the United States when investigative jurisdiction has been assigned by law to such departments and agencies.

The legal arguments on Section 533 are the diametric opposite of the legal arguments on Section 515(b). Section 515(b), read literally, doesn’t clearly authorize the appointment of a special assistant to the Attorney General—but the history and surrounding context of the statute suggests that it was intended to do so. By contrast, Section 533, read literally, does seem to clearly authorize the appointment of a special assistant to the Attorney General—but the history and surrounding context of the statute suggests that it wasn’t intended to do so.

As I see it, Section 533 is so clear in Smith’s favor that the history and surrounding context can be ignored. Here is how I would analyze this issue:

• Jack Smith is an “official.” He’s doing official things. He’s certainly not doing unofficial things. They’re official things.

• Jack Smith is a prosecutor. He’s “prosecut[ing] crimes against the United States.”

• Therefore, under 28 U.S.C. § 533(1), Attorney General Garland had the power to appoint him.

End of case.

You might think this is a simplistic analysis. My response is, guilty as charged. One of the beauties of textualism is that it avoids delicate inferences and multi-factor tests and philosophical considerations. You just read the words and see what they say.

Textualism isn’t always easy. Some texts are confusing and there is no way to avoid conducting a sophisticated analysis and drawing delicate inferences. Indeed, that’s the case, in my view, for Section 515(b). But when a text is straightforward, we should celebrate, end the case in five minutes, and go enjoy a barbecue, rather than conduct a sophisticated analysis for sophistication’s sake. Section 533’s clarity is worthy of celebration. The Attorney General can appoint officials who prosecute. Jack Smith is an official who prosecutes. We can all go home.

So why does Judge Cannon conclude that Section 533 doesn’t authorize Smith’s appointment?

Judge Cannon opens with the “preliminary point” that “the Appointment Order issued in November 2022 is the first appointment order or regulation that has cited Section 533 as a source of special-counsel appointing authority.” The fact that this is Judge Cannon’s first point should set off alarm bells that an atextual analysis is ahead. The question before the court is whether Congress, by law, has vested the appointment of Jack Smith in the Attorney General. Whether or not anonymous no-longer-serving DOJ employees cited the right statute in old, no-longer-valid appointment orders should be extremely low on the list of things we should care about.

Turning to the statutory text, Judge Cannon opens with the argument that “[t]he term ‘officials’ is not synonymous with ‘officers.’” According to Judge Cannon, “definitions of ‘officer’ emphasize the elevated degree authority, responsibility, and duty that inheres in the position.” Meanwhile, “definitions of ‘official,’ … tend to describe a more general class of bureaucratic personnel.” Therefore, “while all officers may be officials, not all officials are officers.”

Let us accept, for the sake of argument, that Judge Cannon is right about this: all officers are officials, but not all officials are officers. In other words, the word “officials” is broad enough to encompass both officers and non-officers. So what? Section 533 authorizes the Attorney General to appoint “officials.” So he can appoint both officers and non-officers. Hooray for mindless literalism.

Judge Cannon interprets the words “officials” to exclude officers. In other words, she interprets “the Attorney General may appoint officials” to mean “the Attorney General may appoint officials, as long as they’re not officers.” That’s not what “officials” means. Apply all the canons of construction and delicate contextual inferences you want, there’s just no way “officials” can mean “non-officer officials.”

Next, Judge Cannon argues that “[s]ubsection (2) through (4) describe security and investigative employees within the FBI,” who are “bureaucratic personnel.” “It is implausible,” Judge Cannon reasons, that “Congress intended to wedge appointment power for special counsels … into a statute concerning low- and mid-level law enforcement personnel in a statutory section governing the FBI.”

So what does “prosecute” mean? According to Judge Cannon, “the meaning of ‘prosecute’ naturally encompasses FBI employees who are engaged or involved in federal investigations and prosecutions.” This could include “FBI attorneys and other legal staff,” as well as “non-lawyer FBI personnel.”

I’m not persuaded. First, subsection (4) refers to “officials” that “conduct such other investigations regarding official matters under the control of the Department of Justice and the Department of State as may be directed by the Attorney General.” That’s extremely broad; I don’t see any limitation to “bureaucratic personnel.” I don’t think you can artificially narrow subsection (4), and then use that artificial narrowing to hold that subsection (1) is also artificially narrow.

Second, I disagree that subsections (2) and (3) can transform the meaning of subsection (1). I am aware of the noscitur a sociis canon, which, as the Supreme Court recently explained, “teaches that a word is given more precise content by the neighboring words with which it is associated.” But to give a word “more precise content” is different from changing a word to a different word. “Officials” who “prosecute crimes” just doesn’t mean “FBI attorneys and other legal staff” and “non-lawyer FBI personnel.”

I realize that subsections (2) and (3) are really narrow, while subsection (1) is really broad. But sometimes broad things are listed alongside narrow things, and that’s OK. Article I of the Constitution enumerates Congress’s powers. Some of those powers are broad, like the power to regulate interstate commerce. Others are narrow, like the power to establish post roads. If the Framers were allowed to list broad things alongside narrow things, Congress is allowed to do it too. If subsection (1) was ambiguous, maybe I’d construe it narrowly, but it says what it says.

Judge Cannon observes that Congress usually uses the term “officers” when it desires for officers to be appointed. Smith submitted a filing with a long list of statutes that do use the word “officials” to encompass “officers,” but Judge Cannon concludes that all those statutes are distinguishable from Section 533—for example, some laws say that “officials” will be “appointed by the President” with “the advice and consent of the Senate,” which supports an inference that those officials are “officers,” whereas Section 533 doesn’t refer to presidential appointment and Senate confirmation, which implies that “officials” in Section 533 excludes “officers.”

Sorry everyone, I’m just a country lawyer here and this is way too complicated for me. There are lots of laws out there. To repeat a point I’ve already made, they’re enacted at different times, with different words, for different reasons. “Officials” is clearly broad enough to include “officers,” as both the dictionary and all the laws cited by Smith will tell you. I don’t see any advantage in spelunking through caverns of the U.S. Code dealing with the Tennessee Valley Authority and the African Development Fund (these are actual examples of statutes that are debated).

Finally, Judge Cannon points to “Section 533’s placement in the legislative scheme”—viz., its placement in a chapter concerning the FBI. She also points to Section 533’s heading: “investigative and other officials; appointment.” These are indications, she argues, that Section 533(1) refers only to investigative personnel, not special assistants to the Attorney General.

But Section 533 was placed in its current location as part of a reorganization of Title 28 in 1966. In that reorganization, Congress also said: “An inference of a legislative construction is not to be drawn by reason of the location in the United States Code of a provision enacted by this Act or by reason of the caption or catchline thereof.” (See pages 616 and 631 of this PDF). So I don’t think that any inference can be drawn from the factors Judge Cannon identifies.

Out of the frying pan, into the fire. In that same reorganization, Congress also said (at page 631): “The legislative purpose … is to restate, without substantive change, the laws replaced by those sections on the effective date of this Act.” And if I’m tracking this correctly, the pre-existing version of Section 533(1), then located at 5 U.S.C. § 300, said this:

Officials for detection and prosecution of crimes. For the detection and prosecution of crimes against the United States and for the acquisition, collection, classification and preservation of identification and other records and their exchange with, and for the official use of, the duly authorized officials of the Federal Government, of States, cities, and other institutions, such exchange to be subject to cancellation if dissemination is made outside the receiving departments or related agencies, the Attorney General is authorized to appoint officials who shall be vested with the authority necessary for the execution of such duties.

This gives me pause … “prosecution of crimes” is put alongside “acquisition, collection, classification and preservation of identification and other records.” It’s likely that whoever wrote this language was thinking about law enforcement personnel. But in the end, I’m sticking with Team Text. “Prosecution of crimes” encompasses people who prosecute crimes.

I’m not being anti-intellectual here. I think there is a strong philosophical case for adopting simple, literal interpretations of statutes. Ordinary people are most likely to understand judicial decisions when they depend solely on the ordinary meaning of written words. And easy-to-understand decisions grounded in the English language are most likely to achieve legitimacy across citizens of divergent political views.

These considerations, to me, powerfully support interpreting the text literally in this case. Lots of people care about this case. It’s extremely easy to understand an opinion that says, “Jack Smith is an official person who prosecutes.” Perhaps there is no way, in a case like this one, that a judge could make a decision that will persuade citizens on both sides of the political spectrum. But if such a decision were possible, it would be a decision that said, “Jack Smith is an official person who prosecutes, and I’m not going to consider anything else because that’s what the words on the page say.”

Finally, Judge Cannon expresses concern that Smith’s interpretation “undermines the separation-of-powers principles that animate the Appointments Clause.” She “declines to dilute the appointment power by reading Sections 515(b) and 533(1) as ceding a core legislative function to another branch.” I’m puzzled by this. The Framers explicitly authorized Congress to vest the appointment of officers in Heads of Department. Smith’s argument is that Congress did a thing that the Constitution explicitly authorized Congress to do. His argument may be right and may be wrong, but it certainly doesn’t undermine the Constitution.

Why are we even talking about this?

Smith’s third argument is that we shouldn’t even be talking about any of this because the Supreme Court already decided this issue in United States v. Nixon, 418 U.S. 683 (1974).

In Nixon, the Acting Attorney General, Robert Bork, appointed a Special Prosecutor, Leon Jaworski, to investigate the Watergate scandal. The statutes governing the Attorney General’s appointment authority were the same then as they are now: in other words, if Leon Jaworski was legal, then Jack Smith is legal.

The district court issued a subpoena requiring Nixon, the sitting President, to turn over some tapes that would make him look really, really bad. The question before the Supreme Court was whether the subpoena should be quashed.

Nixon lost. Nixon argued, among other things, that the dispute between Jaworski and Nixon was an “intra-branch” dispute that couldn’t be resolved by a court. The Supreme Court rejected this argument. Here’s what it said, with the key sentence bolded (footnotes are omitted):

Under the authority of Art. II, § 2, Congress has vested in the Attorney General the power to conduct the criminal litigation of the United States Government. 28 U. S. C. § 516. It has also vested in him the power to appoint subordinate officers to assist him in the discharge of his duties. 28 U. S. C. §§ 509, 510, 515, 533. Acting pursuant to those statutes, the Attorney General has delegated the authority to represent the United States in these particular matters to a Special Prosecutor with unique authority and tenure. The regulation gives the Special Prosecutor explicit power to contest the invocation of executive privilege in the process of seeking evidence deemed relevant to the performance of these specially delegated duties. 38 Fed. Reg. 30739, as amended by 38 Fed. Reg. 32805.

The Supreme Court then went on to say: “So long as this regulation is extant it has the force of law.” And: “So long as this regulation remains in force the Executive Branch is bound by it, and indeed the United States as the sovereign composed of the three branches is bound to respect and to enforce it.” Thus, the Supreme Court held, “the fact that both parties are officers of the Executive Branch cannot be viewed as a barrier to justiciability. It would be inconsistent with the applicable law and regulation, and the unique facts of this case to conclude other than that the Special Prosecutor has standing to bring this action and that a justiciable controversy is presented for decision.”

Those statutes in the bolded sentence are still on the books today. Thus, if the bolded sentence is accurate, then Smith is legal.

Smith argues that the bolded sentence was part of the Supreme Court’s holding, and because the Supreme Court has already resolved this issue, there is nothing for Judge Cannon to do other than follow the Supreme Court. It’s a serious argument—the D.C. Circuit bought it in a challenge to the appointment of Robert Mueller.

But Judge Cannon doesn’t buy it. She disagrees with the D.C. Circuit. In her view, the bolded sentence is dicta.

I don’t think dicta is the right word. “Dicta” generally refers to a statement that isn’t essential to a holding. So when the court says “if the facts of this case were different, then the case would come out this other way,” that’s dicta.

That’s not what happened here. The Supreme Court held that the dispute was justiciable because the regulations—which authorized the appointment of a special prosecutor—were binding. If the regulations aren’t authorized by statute, it’s as if they don’t exist. And if the regulations don’t exist, the Supreme Court’s decision comes out the other way. So I wouldn’t call it dicta.

But, Judge Cannon points out, “the issue of the Attorney General’s appointment authority was not raised, briefed, argued, or disputed before the Nixon Court.” In a lengthy discussion, Judge Cannon explains that the Attorney General’s appointment authority wasn’t raised as a question presented in the petitions for certiorari. She also notes that Nixon’s brief conceded that the Attorney General had the authority to appoint Jaworski. She concludes that “the disputed passage is located within a prefatory, stage-setting paragraph which merely served to tee up the case-or-controversy analysis that followed.” She also points out that there’s no reasoning accompanying the Supreme Court’s statement, so it’s not entitled to persuasive weight sometimes afforded to dicta. So she declines to follow it.

What do I think about all this? Well, the Supreme Court does sometimes say that “drive-by rulings”—bare assertions, accompanied with no explanation or evidence the Supreme Court realized it was deciding something—aren’t entitled to precedential weight. Judge Cannon makes a plausible case that the Supreme Court’s statement was a drive-by ruling, so I think she was permitted to do what she did.

I would have been more diffident, though. The Supreme Court could easily have thrown in a footnote observing that Nixon had conceded the Attorney General’s appointment authority and so it wasn’t deciding the issue. But it didn’t. The Court just came out and said that Congress had vested the power in the Attorney General to appoint Jaworski.

Maybe the Supreme Court was just being careless. But maybe the Court thought it was obvious that the Attorney General had the authority. Nixon was a unanimous decision, and you’d think that at least one Supreme Court Justice would have suggested dropping a footnote if this really was an open question.

It’s a big thing to say that the Supreme Court was careless, especially in one of the most famous cases of all time. Sometimes the Supreme Court is, indeed, careless. But in my opinion, it’s the Supreme Court’s prerogative to say that. I would have followed Nixon, careless though it may be.

Don’t be a visionary

I am going to get philosophical for a moment.

If I could summarize my view of good judging in one sentence, it would be: “Don’t be a visionary.” A visionary is someone who reaches novel, counterintuitive conclusions and boldly strikes her own path. If you’re a business leader, a public intellectual, even a political candidate, it’s great to be a visionary. If you’re a judge, it’s not.

Why? Here are a few reasons:

1. When deciding legal questions, the judge’s basic task is to figure out what other people meant. When interpreting statutes, judges figure out what the legislature meant. When interpreting case law, judges figure out what prior judges meant. Visionaries hate figuring out what other people meant. They want to pursue their own ideas. But if you’re pursuing your own ideas, rather than faithfully implementing other people’s ideas, you’re not doing what a judge is supposed to do.

2. When an individual judge elects to become a visionary, judicial decisions become less predictable. Novel and counterintuitive insights are insights that ordinary people can’t predict when they’re carrying out their affairs. Also, when a judge is a visionary, the judge, definitionally, will reach decisions that differ from decisions of fellow judges. Because cases are assigned randomly, a judge appointing herself as a visionary will increase the randomness of judicial outcomes. This is bad. Predictability is a cornerstone of the rule of law. And predictable legal rules help actual visionaries, like people starting novel new businesses, achieve their goals.

3. It’s hard to be a visionary. Conventional wisdom is usually right. Most people who think they’re visionaries are actually cranks. Some people are widely recognized as visionaries, but only because they’ve managed to persuade a lot of people, over an extended period, to buy their products or adopt their ideas. In other words, market forces—either in the actual or intellectual marketplace—ensure that visionaries succeed and cranks fail. It’s different when you’re a judge. Becoming a judge involves persuading a judicial-selection committee to choose you. For federal judges, this happens a single time in your entire life. After you become a judge, people don’t listen to you because they want to; they listen to you because they have to. The market forces in the physical or intellectual marketplace don’t exist. If you consistently go against the conventional wisdom, it probably doesn’t mean you’re a visionary; it probably means you’re consistently wrong.

The dismissal order in United States v. Trump is a visionary judicial decision. At every turn, it goes against the conventional wisdom.

• As to Section 515(b), the decision holds:

  • Congress messed up in 1870: It enacted a statute saying that people who had been specially retained to be Special Assistants to the Attorney General would be commissioned and paid, without realizing that the Attorney General didn’t have authority to specially retain them.

  • Congress messed up again in 1906: It expanded the authority of those Special Assistants, again not realizing that the Attorney General didn’t have the authority to retain them.

  • Numerous Attorneys General, dating back well over a century, messed up—they appointed Special Assistants without realizing this was illegal.

  • All of the people who were prosecuted by these Special Assistants for all of those years, and all of their lawyers, missed this issue. Generations of lawyers committed malpractice.

• As to Section 533(1), the decision holds that Congress messed up again. By its terms, the statute permits the Attorney General, without reservation, to appoint officials to prosecute crimes. But this was also sloppy drafting; Congress actually meant that the Attorney General could hire FBI lawyers and prosecutorial assistants.

• As to United States v. Nixon, everyone messed up. President Nixon’s army of lawyers messed up by failing to raise, or even notice, the winning argument. The Supreme Court messed up too. On its face, the Court’s opinion states without reservation that the Attorney General had the statutory authority to hire Leon Jaworski. All of the Justices who heard the case, and all of their law clerks, carelessly failed to notice that this was an open issue.

All of these conclusions, on their own, are defensible. I reiterate what I said at the beginning: Judge Cannon’s opinion is very well done.

But at the end of the day, in a close case, I do not agree with a decision premised on the view that so many legislators, lawyers, and judges made so many mistakes. On appeal, I hope the Eleventh Circuit goes with the flow.

1. Judge Cannon expressed doubt as to whether Smith was an “inferior Officer”—in which case Attorney General Garland could constitutionally appoint him, so long as a statute authorized the appointment—as opposed to a “principal Officer”—in which case Article II would have required presidential appointment and Senate confirmation. But Judge Cannon didn’t reach a firm conclusion on this issue. Also, if Smith is indeed a “principal Officer,” the constitutional problem might be fixable by giving the Attorney General additional supervisory authority and hence transforming Smith into an “inferior Officer,” similar to the Supreme Court’s remedy in United States v. Arthrex, Inc., 594 U.S. 1 (2021). Sufficient unto the day being the evil thereof, I will stick to the issue that Judge Cannon definitively decided.