Balancing Equities and the Abortion Pill Cases, Part II
University of Chicago Law Professor Will “Shadow Docket” Baude asked a good question after my last Substack on April 24, “Equities Should Have Been Balanced in the 2023 Abortion-Drug Litigation: A Civil Procedure Post”:
One thing I had trouble parsing from the post -- and apologies if I just didn't read it carefully enough -- was what you think about the merits and the relevance of the merits. That is, if I think the standing argument is extremely weak for the plaintiffs and extremely strong for the government, does that justify being more lenient on the failure to argue the balanced equities in the way you prefer?
Professor Baude wrote the April 22 Volokh Conspiracy post, “The Supreme Court's Mifepristone Stay,” where cites to Professor Adam Unikowsky’s April 13 blogpost, “Mifepristone and the rule of law, part III: The Fifth Circuit's decision is wrong too,” as a good place to learn about the standing issue of whether, regardless of the merits of the case, the plaintiff doctors’ organization has enough at stake in the FDA’s abortion pill approval process to have the right to sue. That is a huge issue in this case, and one I have stayed away from. I am very interested, and fairly knowledgeable, about the issue of what an applicant should do when he applies to a court for a temporary remedy while the main case is proceeding, but there are three other areas of law that are relevant:
Standing. Does the doctors’ organization suffer enough direct harm from the FDA doing the safety approval wrong for the pill to have the right to sue to get the approval procedure redone?
Jurisdiction. Is the Texas federal court the right place to bring the suit, and does its ruling really apply all over the United States, even when the court in Washington State has made an opposite ruling later? (The “national injunctoin” problem).
The Merits in Administrative Law. Once we get past court procedure to The Merits, it turns out the merits are procedure too— FDA procedure. This is not, legally, an abortion case. The issue is not abortion, but whether for this particular pill the FDA followed its approval procedure correctly, in a way that ensures the pill is safe and effective for the woman taking it. An oddity is that if it is safe and effective for the woman, that makes it unsafe and deadly for the baby, but that’s not the issue in this case, because the Supreme Court has left that issue to each state to decide for itself.
So, I have not even tried to understand the merits, and I’ve pretty much ignored standing and jurisdiction too.1
This makes it both easy and hard to address Professor Baude’s question. The easy part is the answer to his question taken literally.
If I think the standing argument is extremely weak for the plaintiffs and extremely strong for the government, does that justify being more lenient on the failure to argue the balanced equities in the way you prefer?
The answer is Yes. Let’s make this concrete. Suppose the District Judge granted the doctors’ group a preliminary injunction to ban the pill and the drug company has appealed the decision to you, a judge on the 5th Circuit appeals court. The drug company is asking you to grant a temporary stay to prevent the preliminary injunction from going into effect. You thinks the drug company almost surely will be able to show later that the doctors’ group lacks standing, but the drug company makes no attempt whatsoever to balance the equities. What do you do?
What you should do is grant the stay. That is because if the probability of success on the merits is high enough, and the applicant will suffer any amount of irreparable harm, then it doesn’t matter how big the irreparable harm is to everyone else. The outcome of balancing the equities is so obvious the judge can do it in his head.
Suppose the judge thinks the drug company has a 99.9999% chance of winning on the merits, and will suffer $10,000 in irreparable harm. That’s enough not to bother with estimating the irreparable harm to everybody else concerned, the doctors’ group and everyone else in the world.
But let’s go deeper into this than just Professor Baude’s precise question. What if the judge thinks the drug company has just a 90% chance of winning?
Each court has its own precedents for the various factors, even though they all follow the same general common law rule of looking at probability of success on the merits, irreparable harm to everybody concerned, and the public interest. I will be my own court here too, and use the version of the rule I like, which is close to what is used by federal courts in the 7th Circuit (Illinois-Indiana-Wisconsin):
Does the applicant have a substantial likelihood of success— that is, over 10%, even if it isn’t 51%— before full consideration of the matter?
If not, stop. Applicant loses.
Similarly, does the applicant have an overwhelming likelihood of success— that is, over 99%?
If so, stop. Applicant wins.
Will anyone on the applicant’s side suffer irreparable harm during the time period before full consideration of the matter?
If not, stop. Applicant loses.
Will anyone on the side of the applicant’s adversary suffer irreparable harm during the time period before full consideration of the matter?
If not, stop. Applicant wins.
If we balance the equities, factoring in the probability of success, irreparable harm to the applicant, irreparable harm to the other parties, and irreparable harm to third parties, does the applicant come out ahead?
If not, stop. Applicant loses.
If we look to anything else involving the public interest, is there some special reason the applicant should lose?
If so, applicant loses. Otherwise, he wins.
This scheme is designed to conserve court time and energy as well as to achieve justice. If the applicant loses at test 1, there is no reason for the court to waste time on figuring out the answers for test 2, 3, and 4. In practice, judges do waste time by talking about all of these things in their opinions, even if the applicant loses at test 1. That is superfluous effort.
One reason you might object to me saying it is superfluous effort is that a full discussion of all four tests could be useful on appeal. Suppose the trial judge— the lowest court— denies the applicant on test 1 and doesn’t discuss the other three tests, but the applicant appeals to a three-judge appellate panel and the panel decides the trial judge is wrong on test 1. Then under my scheme the appellate panel has to either decide tests 2, 3, and 4 themselves or kick it back down to the trial judge for each test.
That objection is a feature, not a bug. We want to economize on mental effort. It’s great if the appellate panel doesn’t have to think about anything except test 1. They can quickly reverse the trial judge based on that test and kick the case back down to him. He can then start thinking about test 2. If he denies on that basis and the case is appealed again, the appellate court can start thinking about test 2.
(Does that make this an interlocutory appeal? This is like how Japanese courts operate, I think, and is one reason they operate better than U.S. courts. I’ll have to ask Mark Ramseyer and look at his Japanese Law: An Economic Approach, which I have on a shelf somewhere. )
Now let’s apply the scheme to the abortion drug case.
1a. Does the applicant have a substantial likelihood of success— that is, over 10%, even if it isn’t 51%— before full consideration of the matter?
Well, probably. It seems to me that the doctors’ organization is too little injured by unsafe abortion pills to have the right to sue, but standing rules are complicated, so maybe I’d change my mind, so let’s put my estimate of likelihood of success at 20%. Notice that test 1 is NOT “Who will win on the merits?” That would be to decide the entire case now. At this stage, we just want to filter out the frivolous suits, of which there are many. Nonetheless, judges often go wrong here: they decide who they think should win, and grant a preliminary injunction or stay based on that while just paying lip service to the other three tests by scribbling down a few more paragraphs about them. Anyway, the applicant passes this test, so let’s go on.
1b. Similarly, does the applicant have an overwhelming likelihood of success— that is, over 99%?
No, clearly not because 20% is too small. So let’s go on to test 2.
2a. Will anyone on the applicant’s side suffer irreparable harm during the time period before full consideration of the matter?
Yes, the babies who are aborted would be dead and unable to sue for money damages. So continue to test 2a.
2b. Will anyone on the side of the applicant’s adversary suffer irreparable harm during the time period before full consideration of the matter?
Yes, the women who want abortions will not get them. Note that I would *not* count the drug companies’ lost profits, because that is a money loss and the judge could require the plaintiff doctors’ organization to post a bond to repay them if the plaintiffs eventually lost on the merits.
3. If we balance the equities, factoring in the probability of success, irreparable harm to the applicant, irreparable harm to the other parties, and irreparable harm to third parties, does the applicant come out ahead?
Test 3 is the hard part. Now we have to start doing real work. We have to roughly quantify the probability of success and the losses to everybody concerned.
The sensible way to factor in everything is called the “sliding scale” method. In rough terms, that means that even if the applicant would suffer more harm, if he is very unlikely to win, he loses anyway. In the abortion pill case, I think if I balanced the equities this way, I’d end up deciding against the doctors’ organization. Their standing is very dubious, even though their merits on the question of whether the FDA followed proper procedures seems strong, so I’d estimate a 20% probability of eventual success. The irreparable losses are clearly there for each side, but they are very hard to quantify, and practically impossible for me as judge to quantify sua sponte— that is, without the lawyers having done a lot of research and putting numbers in their briefs. Most basically, how many abortions will be prevented if the preliminary injunction is granted? How can I estimate harms unless I have that number? As a judge, I’d know law well enough to have some chance of doing a good job of estimating the probability of success sua sponte, but when it comes to facts like this, judges have to depend on the parties to the lawsuit to do the work. So even though as someone who is strongly opposed to abortion I think the harm from an unlawfully approved abortion pill would be very great, I’d turn down the request for a preliminary injunction.
Let’s go on, though, to illustrate how a judge should balance the equities when he does have reasonable estimates of the harm. Suppose I decide to rate the loss to the doctors’ organization and people on their side from keeping the pill legal until the decision on the merits as 100, and the loss to the drug company and people on its side from having a preliminary injunction against pill sales until the decision as 10. The exact number is not what matters: it’s the ratio between them: I’ve decided that the plaintiff-side loss is ten times the defendant-side loss.
Now I can balance the equities. Earlier, I estimated the plaintiff’s chance of success at 20%. Multiplying this by their loss if the preliminary injunction is denied wrongly, we get 20% (100) = 20. Doing the same for the defendant’s probability of winning and loss if the preliminary injunction is granted wrongly, we get 80% (10) = 8. Since 20 is bigger than 8, I will grant the plaintiff their preliminary injunction.
I know that this use of numerical values will seem unduly precise to most readers. You may think that just making up numbers like 20%, 100, and 10 is wrong. Indeed, my numbers are almost surely wrong. In the end, one side or the other will win on the merits, so the true probability is either 0% or 100%. The losses to each side are extremely hard to estimate for the same reason that they are “irreparable” harm: they do not translate easily to money. You may think back to what Aristotle said in Chapter 3 of The Ethics:
Our discussion will be adequate if it has as much clearness as the subject-matter admits of, for precision is not to be sought for alike in all discussions, any more than in all the products of the crafts. Now fine and just actions, which political science investigates, admit of much variety and fluctuation of opinion, so that they may be thought to exist only by convention, and not by nature. And goods also give rise to a similar fluctuation because they bring harm to many people; for before now men have been undone by reason of their wealth, and others by reason of their courage. We must be content, then, in speaking of such subjects and with such premisses to indicate the truth roughly and in outline, and in speaking about things which are only for the most part true and with premisses of the same kind to reach conclusions that are no better. In the same spirit, therefore, should each type of statement be received; for it is the mark of an educated man to look for precision in each class of things just so far as the nature of the subject admits; it is evidently equally foolish to accept probable reasoning from a mathematician and to demand from a rhetorician scientific proofs.
There is some truth in what Aristotle says, but not in the context of stays and preliminary injunctions. Here, we know that probability of success and losses to each side all have to be balanced. We can balance them by thinking hard about what our estimate is of the numbers, or we can shut our eyes and mutter vaguely about one loss being big and the other being small. We can say, “pretty small, but not extremely small” instead of “20%”. But we would be doing the same kind of calculation, implicitly, with words that we do with numbers, just less transparently. It is much like I say in the Introduction to my book, Games and Information: An Introduction to Game Theory about how intelligent laymen object to the amount of math in economics:
Intelligent laymen have objected to the amount of mathematics in economics since at least the 1880s, when George Bernard Shaw said that as a boy he
(1) let someone assume that a = b,
(2) permitted several steps of algebra, and
(3) found he had accepted a proof that 1 = 2.
Forever after, Shaw distrusted assumptions and algebra.2
Despite the effort to achieve simplicity (or perhaps because of it), mathematics is essential to exemplifying theory. The conclusions can be retranslated into words, but rarely can they be found by verbal reasoning. The economist Wicksteed put this nicely in his reply to Shaw’s criticism:
`Mr Shaw arrived at the sapient conclusion that there “was a screw loose somewhere”— not in his own reasoning powers, but—“in the algebraic art”; and thenceforth renounced mathematical reasoning in favour of the literary method which enables a clever man to follow equally fallacious arguments to equally absurd conclusions without seeing that they are absurd. This is the exact difference between the mathematical and literary treatment of the pure theory of political economy.’ (Wicksteed [1885] p. 732)3
Here, too, applying arithmetic and algebra makes it clear what you are implicitly and unknowingly assuming when you try to the same thing just with words.4
Balancing the equities ought to be the grand finale of the scheme, but in the traditional order, we put the public interest test last:
4. If we look to anything else involving the public interest, is there some special reason the applicant should lose?
Test 4 is usually easy. There’s no special reason here, and there almost never will be. Sometimes this test has been used to incorporate irreparable harms to third parties on both the plaintiff and defendant sides, but we’ve already done that in test 3, which is where it ought to be considered.
In conclusion, in the abortion pill cases I’d probably have granted the preliminary injunction and denied the stays in the appeals courts if the parties had briefed me properly on balancing the equities. But I present this as an illustration of the method, not as a serious decision. I haven’t read the briefs carefully, and using this same method I might come out with the opposite answers if I did, based on their sketchy submissions to me.
Not that standing, jurisdiction, and administrative law aren’t interesting. Indeed, New York State ex rel. Eric Rasmusen v. Citigroup, Inc. (220 F.Supp.3d 523 [2016], 75 N.Y.S.3d 903 [2018]) was about these three things as well as tax law and corporate law. One reason I lost was because the Court ruled I knew of Citigroup’s underpayment of taxes from “public” information as defined in New York’s whistleblower statute, so I had no right to sue on behalf of New York State, the beneficial plaintiff. The case was initially filed by me in New York State court, but Citigroup removed it to Federal court, and then the Federal judge kicked it back to state court, ruling that he had no jurisdiction because it was about state taxes, not federal. My case was based on the U.S. Treasury not having the right to exempt a corporation from losing its net operating loss carryforwards as tax deductions after an ownership change (temporary acquisition by the U.S. government) just because there was a financial crisis and they wanted to throw money at banks. You can see why judges didn’t like to have to deal with this case. If it had been simpler, maybe I’d be $100 million richer.
A fallacious proof that 1 = 2. Suppose that a = b. Then ab = b2 and ab − b2 = a2 − b2 . Factoring the last equation gives us b(a − b) = (a + b)(a − b), which can be simplified to b = a + b. But then, using our initial assumption, b = 2b and 1 = 2. (The fallacy is division by zero.)
That’s a Shaw quote within a Wicksteed quote within a Rasmusen quote quoted by Rasmusen, so I don’t blame you if you feel confused. Substack doesn’t let me indent it and use font type and size the way I’d like to to make it clearer.
Alfred Marshall’s 1906 letter to Bowley adds some caveats to take us back a bit to Aristotle’s side:
I had a growing feeling in the later years of my work at the subject that a good mathematical theorem dealing with economic hypotheses was very unlikely to be good economics: and I went more and more on the rules---
(1) Use mathematics as a short-hand language, rather than as an engine of inquiry.
(2) Keep to them till you have done.
(3) Translate into English.
(4) Then illustrate by examples that are important in real life.
(5) Burn the mathematics.
(6) If you can't succeed in 4, burn 3.
This last I did often.
Perhaps I should Substack on these rules sometime. I will also add the caution that both user and audience should beware of the use of mathematics to obfuscate and win arguments; see my March 29 “(a+b^n)/n = x. Therefore, God Exists.”