Why does doctrine get so complicated?

(Or does it?)

In 1998, a Texas jury convicted Rodney Reed of capital murder and sentenced him to death. He maintains his innocence but remains on death row.

Reed wants to conduct DNA testing of the crime-scene evidence against him, claiming it will establish his innocence. The District Attorney has mostly refused this request. In the D.A.’s view, the evidence is tainted and evidence of other people’s DNA at the crime scene wouldn’t prove Reed’s innocence anyway.

A Texas law called “Article 64” permits inmates to obtain DNA testing of crime-scene evidence, but only if certain conditions are met. Reed brought a claim in Texas state court under Article 64. He lost. He then sued the D.A. in federal court, contending that Article 64’s procedures are unfair.

This dispute reached the Supreme Court. On April 19, 2023, the Supreme Court issued its opinion in Reed v. Goertz. It held that (1) Reed has Article III standing; (2) Reed’s claim is not barred by the Rooker-Feldman doctrine; (3) the D.A. has no valid sovereign immunity defense; (4) Reed’s claim is timely because it accrued at the time of the Texas Court of Criminal Appeals’ denial of rehearing. One dissent argued that the lawsuit is barred by Article III and the Rooker-Feldman doctrine; another dissent argued that the claim accrued at one of three earlier possible times.

Huh?

After reading Reed v. Goertz, one gets the feeling that the American legal system has failed. Maybe Reed should get DNA testing and maybe he shouldn’t. But whatever the answer to this question, it should not turn on Article III, the Rooker-Feldman doctrine, sovereign immunity, and the selection of one from among four different possible accrual dates. Some disputes have convoluted facts, so one would expect the legal analysis to be correspondingly complex. But this dispute is simple. Reed says DNA testing would prove his innocence. The D.A. says it wouldn’t. If deciding this dispute requires the U.S. Supreme Court to resolve four difficult antecedent procedural issues, something has gone awry.

The purpose of this post is to conduct a kind of forensic analysis, so to speak, of how such a simple case could have yielded such a tangle of complicated questions. Do lawyers just love to make things complicated? Or does Byzantine complexity naturally emerge from seemingly simple legal rules?

Maybe the legal system isn’t that Byzantine?

(In 1453, Constantinople fell to the Ottoman Empire. When that happened, Eastern Europe was no longer Byzantine. This image depicts Dall-E’s rendition of that event, retro style.)

Actually on second thought, is Reed v. Goertz really that complicated? Perhaps we should start by questioning the assumption that legal doctrine can get soooo complicated.

I’m not the first to point this out, but law requires the shallowest degree of subject-matter expertise of any intellectual profession. Let’s suppose you’re an ambitious freshman at MIT who wants to learn about supersymmetry. It would take you at least five years of sustained study to even get to an introduction to this topic. Working through the course catalog:

• First year, fall: Physics I; Calculus I

• First year, spring: Physics II (Prerequisite: Physics I), Calculus II (Prerequisite: Calculus I)

• Second year, fall: Physics III (Prerequisites: Physics II, Calculus II); Differential Equations (Corequisite: Calculus II)

• Second year, spring: Quantum Physics I (Prerequisites: Physics III, Differential Equations)

• Third year, fall: Quantum Physics II (Prerequisite: Quantum Physics I)

• Third year, spring: Quantum Physics III (Prerequisite: Quantum Physics II)

• Fourth year, fall: Quantum Theory I (Prerequisite: Quantum Physics II); Electromagnetism II (Prerequisite: Physics III, Differential Equations)

• Fourth year, spring: Quantum Theory II (Prerequisite: Quantum Theory I, Electromagnetism II); Relativistic Quantum Field Theory I (Prerequisite: Quantum Theory I)

• Fifth year, fall: Relativistic Quantum Field Theory II (Prerequisites: Relativistic Quantum Field Theory I, Quantum Theory II)

• Fifth year, spring: Relativistic Quantum Field Theory III (Prerequisite: Relativistic Quantum Field Theory II)

The prerequisites here are real; each class really does require mastery of the prior class in the sequence. Also, it’s assumed later in the sequence that students will have taken other physics and math classes not officially listed as prerequisites. Finally, after all this, the last topic in the syllabus of Relativistic Quantum Field Theory III is “introduction to supersymmetry.” So, after five years, you are still taking an intro class.

Other professions are like this too. Medicine, architecture, engineering require years of study before achieving competence.

How about law school? The prerequisite to being admitted to law school is a college degree in anything. I studied physics and electrical engineering/computer science in college, which were very interesting topics but had no relevance to law school. As another example, noted practitioner Elle Woods matriculated at the top of her Harvard Law School class after completing a degree in fashion merchandising.

In law school, the first-year curriculum, at least when I was there, included five intro classes that could be taken in any order: Civil Procedure, Torts, Contracts, Criminal Law, and Property. (I agree the first four should be in the 1L curriculum. As for Property, I have never encountered many of the concepts learned in that course, such as future interests and the rule against perpetuities, at any point in my legal career.) In 2L and 3L year, virtually all courses were optional. The sole other required law school course was Professional Responsibility, where you learned that, when possible, it is best not to sue your own clients. Other than that, completing law school required taking two years’ worth of whatever random intro courses you wanted. Art Law, Japanese Law, whatever. You didn’t even have to take Constitutional Law, although most people did. After two years of this, the Dean pronounced you able to “think like a lawyer” and you were declared a Juris Doctor.

After finishing law school, many law students immediately become judicial law clerks, in which they are expected to draft judicial opinions in any area of law, including areas to which they had zero exposure in law school. If a judge asks a law clerk to prepare a judicial opinion in (say) an employment discrimination case, and the student expresses concern that she did not take Employment Law in law school, the judge will assume that the law clerk is making a whimsical joke. This practice continues at the U.S. Supreme Court, where it’s quite possible that the first draft of a seminal opinion in (say) antitrust law will have been prepared by a law clerk who had literally zero exposure to antitrust law until picking up the briefs in that case.

(I personally would make law school one year, or at most two years, on the perhaps simplistic view that if courses aren’t required, you don’t need them to become a lawyer. Students could optionally attend law school for additional years, and perhaps get some additional certificate or something for those who find that alluring, but it wouldn’t be a requirement to get a bar license. Alas, this idea has been proposed many times and has gone nowhere.)

I don’t mean to criticize law or legal education. The fact that legal disputes are generally understandable to people lacking expertise in a particular legal field is a feature rather than a bug of the American legal system. Indeed, virtually any Supreme Court case is understandable to even a non-lawyer after perhaps an hour of explanation, which is a good thing. Law is, after all, for the people, not the lawyers. In most cases, people should be able to understand the rules that govern them without paying a lawyer. Even when a lawyer is needed, the lawyer should be able to explain the case to the client without too much difficulty.

There are other reasons legal disputes should, and do, remain comprehensible without sustained study of the relevant legal field:

• The decisionmakers are non-experts. Judges must instruct lay juries on the law, which is impossible if the law is too complicated. Judges themselves are usually generalists and vary significantly in professional background. Also, many state-court judges are elected officials or political appointees not selected for their legal acumen. The legal system could not operate if legal rules were too complex.

• In addition to being generalists, judges may also have strong political views and face the temptation of motivated reasoning. The more complex the legal rule, the easier it is for the judge, consciously or unconsciously, to apply the rule in a skewed manner.

• Finally, one reason that topics like quantum physics can get so complicated and remain coherent (so to speak) is their extreme precision. The underlying mathematical concepts, while complicated, aren’t ambiguous and applying them yields objectively right and wrong answers. As such, one can build on these concepts to achieve very sophisticated intellectual regimes. By contrast, the substrates of legal analysis are verbalisms, which tend to get ambiguous when they get interesting, plus the balancing of incommensurate factors. As a result, any attempt at hyper-complex legal analysis inevitably collapses under its own weight. Consider the famous Georgia-Pacific factors, which are 15 (!) different factors that may be considered in calculating a reasonable royalty. One can imagine some earnest legal scholar coming up with some very complicated way of balancing these factors together, assigning them weights, etc. But in the real world the juice wouldn’t be worth the squeeze, because each individual factor may yield ambiguous or disputed results and figuring out some new way to balance them together wouldn’t produce better outcomes.

Overall, it’s both inevitable, and a good thing, that the legal system is comprehensible to lay people.

Maybe not Byzantine, but still too complicated.

(This is Dall-E’s surrealist take on Byzantium. No idea what’s going on here.)

Let us refine our critique of Reed v. Goertz. The doctrines at issue aren’t that complicated. For example, Article III requires you to show that you’ll be better off if you win. The Rooker-Feldman doctrine holds that you can’t file one lawsuit challenging the result of a different lawsuit.

The problem with Reed v. Goertz is that it seems arbitrarily complex. Why do simple facts require application of all these doctrines?

Well, as a starting point, the case is complex because Reed chose to split his lawsuit in two. He first brought the Article 64 suit in state court and lost. But in his first lawsuit, he didn’t argue that the Article 64 procedures were unfair; he just accepted them as gospel and argued that he should prevail under them. Then, he brought a new lawsuit in federal court, arguing that Article 64, as applied in state court, is unfair. There are lots of doctrines designed to prevent plaintiffs from bringing new lawsuits challenging the outcomes of previous lawsuits, so much of the complexity concerns whether Reed could overcome those doctrines.

But that’s not a satisfactory answer to why Reed v. Goertz got so complicated. Why was the legal system structured to give Reed an incentive to proceed in this way?

The basic reason, I think, is that historic ambivalence about state courts’ ability to protect legal rights has yielded ambivalent legal doctrine, which produces strange incentives.

On the one hand, in 1790, Congress enacted the full faith and credit statute, now codified at 28 U.S.C. § 1738, which requires federal courts to give full faith and credit to state judgments. This means, to first approximation, that if you lose an issue in state court, you can’t relitigate it in federal court. The full faith and credit statute reflects trust of state courts: once a state court resolves an issue, it’s over.

On the second hand, in 1871, Congress enacted the Civil Rights Act, now codified in relevant part at 42 U.S.C. § 1983, which permits plaintiffs to sue state officials in federal court for violating their constitutional rights. There’s no exhaustion requirement; you can come right to federal court and sue. Section 1983 reflects distrust of state courts: Congress conferred a cause of action in federal court because it didn’t think state courts would protect federal rights.

On the third hand, in District Attorney’s Office for the Third Judicial District v. Osborne, 557 U.S. 52 (2009), the Supreme Court held that an inmate lacked a federal constitutional right to post-conviction DNA testing when the litigant didn’t even try to use the state’s DNA-testing procedures. Alaska, like Texas, allows litigants to get DNA testing in some instances but not others. Osborne didn’t try to use Alaska’s procedures and instead filed a 42 U.S.C. § 1983 suit directly in federal court saying he had a constitutional right to DNA testing. The Supreme Court ruled against him, finding that he had to give Alaska’s procedures a shot first: “These procedures are adequate on their face, and without trying them, Osborne can hardly complain that they do not work in practice.” Osborne reflects trust towards state courts.

So you’ve got a 1790 law, a 1871 law, and a 2009 decision, which are protective, unprotective, and protective of state courts respectively. All are still on the books—as a result of which Reed’s strategy make perfect sense. Why did he have a right to sue in federal court? Because the Congress of 1871 didn’t trust state courts. Why didn’t he start off in federal court? Because the Supreme Court of 2009 believed this would be an unfair intrusion on state courts. Why didn’t he at least try to raise his due process argument in state court before coming to federal court? Because the Congress of 1790 decided that state court decisions are preclusive in federal court. The only way to navigate this labyrinth was to bring the state court suit but withhold the due process claim, and then bring the due process claim in federal court.

If you got the 1790 Congress, the 1871 Congress, and the 2009 Supreme Court in a room, they would all agree that this process doesn’t make sense. But because Congress and the Supreme Court trusted state courts more at some times than other times, there’s a welter of conflicting doctrines from which complexity naturally emerges.

Let’s add a fourth doctrine to the mix: the Rooker-Feldman doctrine. This doctrine holds that, in general, you can’t appeal state court decisions to a federal district court. If you think a state court decision violated your federal rights, take it up to the U.S. Supreme Court. This is another doctrine that reflects respect of state courts.

To get around this, Reed says he’s actually challenging Article 64 itself, not the state court’s decision. However, under Osborne, he’s required to avail himself of the state court’s process before complaining about it, which indeed he did. So he threads the needle by saying that he’s challenging Article 64 as interpreted by the state court, not the state court’s application of Article 64. If this seems like a thin distinction to you, it should. If the legal system was being designed from the ground up, no one would think this is the type of distinction on which an inmate’s freedom should turn; these arguments emerge because of the difficulty of balancing multiple conflicting doctrines together.

One more doctrine. Reed’s position is that he’s not seeking to establish that his conviction should be overturned. Why would Reed take this position? Doesn’t he want his conviction to be overturned?

As mentioned above, 42 U.S.C. § 1983 usually permits plaintiffs to sue directly in federal court without exhausting state remedies. But there’s an exception. State prisoners can’t use § 1983 to get out of prison. Instead they must resort to the habeas corpus statute, which requires exhaustion of state remedies as well as lots of other requirements that are extremely difficult to meet.

At first glance, this exception seems bizarre. If we’re concerned that state courts aren’t adequately protecting federal rights, shouldn’t we be most concerned about prisoners alleging they are incarcerated in violation of their federal rights? Indeed we should. But Congress and the Supreme Court tightened the screws on habeas corpus in an era when both institutions were focused on protecting federal/state comity.

Here, Reed brought a § 1983 claim. The reason he’s allowed to do so, without facing the crucible of the habeas corpus statute, that he doesn’t contend that victory will actually get him out of prison; instead it’ll just get him a DNA test which may or may not vindicate him. See Skinner v. Switzer, 562 U.S. 521 (2011) (buying this specific argument).

So the degree of respect for state courts has ebbed and flowed over time, and that’s why we’re in this strange place where Reed has an incentive not only to withhold an argument from state court, but to insist he’s not trying to get out of prison.

Wait, what was Reed even about anyway?

I haven’t even gotten to the primary issue addressed in Reed—the date the statute of limitations clock started. The majority agreed with Reed’s proposed date (the denial of his petition for rehearing in the Texas Court of Criminal Appeals), which rendered his suit timely. Justice Alito’s dissent proposed three other, earlier dates. Why did that question get so complicated?

It’s because Section 1983 is an extremely general law. General laws sometimes make legal analysis easier, but in this case the generality of Section 1983 makes legal analysis harder.

Usually it’s easy to know when the limitations clock starts for a Section 1983 claim. For instance, if a plaintiff sues a police officer under Section 1983 alleging the officer used excessive force, the clock starts on the date of the excessive force.

But in Reed’s case, it is not so easy.

• Reed is challenging Article 64 itself, not the state court decision (remember the Rooker-Feldman doctrine). So maybe the state court’s decision shouldn’t matter? And maybe the clock started when the D.A. refused to let him test his DNA?

• But wait, he’s challenging Article 64 as construed by the state court (remember Osborne). So maybe the clock started when the state court construed Article 64. But when did that happen? When the trial court construed Article 64? When the appellate court agreed with the trial court’s construction?

• But wait, he didn’t even know if he had a claim until he exhausted state court procedures. After all, if he won in state court, he couldn’t reasonably claim the state court procedures were unfair. And he didn’t know if he’d win in state court until the state court procedures were over. So maybe that’s when the clock started?

This is a hard problem for two reasons. First, each possible accrual date has its advantages and disadvantages. The right answer does not jump out at you.

Second, Section 1983 is silent on this question because it is so broad and general. It just says, plaintiffs can sue in federal court for violations of their federal rights. This particular type of lawsuit was totally outside of the contemplation of the drafters of Section 1983. Broad, general laws have their virtues, but they also have the drawback of requiring judges to invent solutions to difficult problems out of nowhere.

A word about path dependence and accretion

If we were designing the legal system from the ground up, how would we structure this litigation? If we set aside the concern that state courts are less protective of constitutional rights than federal courts, I think there’s a straightforward answer. Reed would bring one lawsuit under Article 64, where he’d argue both that Article 64 entitles him to the DNA, and that if it doesn’t, Article 64 is unconstitutionally unfair. This would allow a single judge to interpret Article 64 against the backdrop of Reed’s constitutional objections, and eliminate the need to figure out which state-court decision started the statute of limitations clock. Reed wouldn’t be any worse off than in the status quo, because he’d still get a hearing on his claim that Article 64 is unfair—he’d simply get that hearing in his first lawsuit rather than bringing a second lawsuit.

That’s what Justice Thomas’s dissent says Reed should have done. And boy, does Justice Thomas make a good case.

Many times when one reads Justice Scalia’s dissents, especially from the first 10-15 years of his career, his analysis seems so persuasive that it is difficult to understand how anyone could possibly have reached a different conclusion. Justice Thomas’s dissent in Reed v. Goertz is like that. He offers an exceedingly logical and tightly-reasoned analysis, explaining that Reed is challenging the state-court judgment and that a challenge to a state-court judgment cannot be brought in federal court. He also explains, persuasively, that Reed doesn’t have standing to sue the D.A. because he’s not seeking an order forcing the D.A. to do anything. Why did no other Justice join it?

Undoubtedly some members of the Court were concerned about an inmate facing execution with lingering questions about innocence. They were also likely concerned that Reed did not get a fair shake in state court. Death row inmates do not have a good track record in the Texas Court of Criminal Appeals.

But I’m not sure this concern explains the votes of some of the conservative-leaning Justices. If I had to guess, those Justices declined to join Justice Thomas because they felt bound by Skinner, which allowed a similar lawsuit to proceed. As Justice Thomas takes pains to point out, Skinner didn’t actually decide the specific issues in the Reed case. But, Skinner came pretty close, and those Justices didn’t want to cast a vote that would be perceived as overruling Skinner.

I suspect that if Reed had come before Skinner, Justice Thomas’s position would have carried more than one vote, and perhaps a majority. But because Skinner came first, the Court decided to stick with it and tolerate some doctrinal impurity.

In other words, the doctrinal outcome was path dependent. The degree of doctrinal complexity turned on the order in which the cases reached the Court. It’s hard to know how often this happens—one cannot rewind time and bring petitions to the Court in a different order to test whether the outcome would be different—but I suspect it happens pretty often.

More generally, legal complexity is a natural outcome of the adjudicative process. Courts use particular cases to announce general legal rules. Those general rules may be influenced by the particular facts of the case or the particular arguments of counsel. The courts may lack a complete picture of the consequences of their rulings. Later on, new cases arise that may lead the judges to question the wisdom of their prior decisions. Yet, they feel compelled by stare decisis to stick to precedent, perhaps with some modifications to account for unexpected developments. In this way, doctrines accrete, and law grows more complex.

In any legal system where judges enact binding precedents based on incomplete information, the inevitable outcome is that doctrine will get complicated.

On that optimistic note, see you all next time!