« Graded assignments | Main | Books Never Feel Real Until . . . »

Thursday, December 02, 2021

“Liquidation” vs. Stare Decisis in Dobbs: The Incoherence of Madison’s Theory of Precedent

I might be one of the few people whose views on precedent are more confidently felt than their views on abortion. I find the latter topic agonizingly difficult. On the former topic, however, I have this strong view: Madison’s theory of precedent, articulated most famously in his June 1831 letter to Ingersoll, is incoherent.

That theory maintains that, when some interpretation of a constitutional ambiguity is endorsed by “the uniform sanction of successive Legislative bodies, through a period of years and under the varied ascendancy of parties,” then that interpretation forms a binding precedent that settles the constitutional dispute. As I argue after the jump, this Madisonian theory requires the assumption of stable “parties” whose joint decisions virtually represent later generations who were not members of such parties. But “parties” are always in flux, as their constituents and platforms shift. The fact that Democratic Republicans of the Henry Clay-Calhoun stripe ratified the Bank of the United States prior to 1816 says little about the views of Jacksonian Democrats post-1832, because the latter simply did not exist as an organized political force during the Era of Good Feelings. Likewise, the ratification of Roe-Casey by Business Republicans on the Burger and Rehnquist Courts says little about the views of New Right GOP voters who despised the GOP Business Establishment.

The failure of Madison’s theory of “liquidation” (to use his term for the clarification of constitutional ambiguities by precedent) points to a deeper problem with every theory of constitutional settlement that relies on popular consensus to resolve contentious disputes: “The People,” as a statistically significant, demographically stable, inter-generational entity, rarely exists and, therefore, seldom have settled views on the sorts of questions that “the People” are asked to resolve.

1. A Quick Summary of Madison’s Theory of Precedent

For those of you who do not routinely assign Madison’s famous letter as mandatory reading in Constitutional Law (Aside: you should), here is Madison’s view of precedent in summary form. (Will Baude has an excellent extended exegesis here). In explaining why he acceded to the constitutionality of the Bank of the United States after more than a decade of arguing that the BUS was unconstitutional, Madison stated: “an exposition of the law publicly made, and repeatedly confirmed by the constituted authority, carries with it, by fair inference, the sanction of those, who having made the law through their legislative organ, appear under such circumstances to have determined its meaning through their Judiciary organ.” The “constituted authority” is the Sovereign People who, through their ratification of the Constitution, created various “organs” (judicial, executive, legislative) to act as their agents. The “repeated confirm[ation]” that counts as such popular “sanction” is not merely any past decisions by such agents, even if such decisions are “repeated.” Instead, genuinely binding precedents are decisions that have “the uniform sanction of successive Legislative bodies, through a period of years and under the varied ascendancy of parties” (emphasis added). The Bank of the United States had allegedly received such sanction, because both Federalists and Democratic Republicans like Henry Clay (who initially opposed the BUS) eventually came to support it in the wake of the War of 1812 when the lack of such a federal fiscal agent was sorely felt.

Stated more generally, when an “organ” of the People (say, SCOTUS) repeatedly “confirms” a decision (say, Roe v. Wade’s constitutional protection of the right to terminate a pregnancy) when rival political parties (e.g.,justices appointed by both Democratic and Republican Presidents) enjoy “varied ascendancy” over that “organ,” then that decision is owed extraordinary deference by political decision-makers, because such a decision amounts to an interpretation of the constitutional ambiguity by The People — in Madison’s words, “the requisite evidence of the national judgement & intention.”

2. What’s Wrong with Madison’s Theory?

So stated, Madison’s theory of precedent surely looks like it ratifies Roe-Casey as a super-strong constitutional settlement entitled to extraordinary deference. So why do I think that the theory is incoherent? Simply put, the “varied ascendancy of parties” that is Madison’s condition for settlement is impossible to meet when those parties’ identities (not just their names, but their platforms and constituents) are in flux. And they are always in flux. The Jacksonian Democracy simply did not exist as an organized movement in 1816, when President Madison signed the Second BUS’s charter. (Fighting stance: Jacksonians really did not exist as a coherent faction at all until after Jackson’s 1832 veto message). Why, then, should they be bound by the assent of parties that did not include their voice? Likewise, the New Right, consisting in large part of a coalition between Catholic and Evangelical voters that helped elect Ronald Reagan, did not exist until the 1980s. As a decisive influence on in the GOP’s program for selecting federal judges, this constituency really mobilized sometime between 1981 and 1987, as ably described by Professor Mary Ziegler. That’s a bit late to be a decisive influence on the appointment of Anthony Kennedy, hardly an outspoken exponent of Meese’s 1980s theory of originalism.

There is a more general point underlying this difficulty in defining “varied ascendancy of the parties”: Such “varied ascendancy” seems normatively relevant by the terms of Madison’s theory only to the extent that it speaks for “the People,” but those who oppose some precedent can generally (and reasonably) complain that the parties who handed down the precedent did not reflect the interests and values of the precedent’s current opponents, because those opponents were not politically mobilized at the time that the precedent was decided.

One might raise both an empirical and a normative objection to this rebuttal of Madison’s theory. As an empirical matter, one might argue that the dominant parties at any moment in history sometimes really do reflect the views of The People at the moment that the precedent was decided. As a normative matter, one might argue that, even if those earlier ascendant parties left out some important interests and values, the parties’ decisions are appropriately binding on everyone because all interests and values could have been raised and discussed earlier even if they were not actually part of the earlier political debate.

I am unpersuaded by these two ripostes. As an empirical matter, I subscribe to the old “Michigan School” of voters’ views, best described by Campbell, Converse, Miller, and Stokes. In a nutshell, the Michigan School maintained that, until mobilized by elites, the vast majority of voters are largely devoid of any stable views about public affairs. There’s always a “silent majority” who could oppose a precedent, simply because the majority of people are always silent about ideological abstractions and even specific political issues. Until some organization actually rallies some segment of public opinion to take a stance on some issue, therefore, the claims that actual partisan platforms represent the latent views of The People are fictions, because the People actually are view-less. As a normative matter, I am not impressed by the idea that such a fiction is a useful legal fiction. Such rests on the idea that, if voters remain silent when an important issue is decided, then they have tacitly consented to the decision. Put another way, the passive citizen has no right to complain about decisions made in his or her name, because they had the chance to bestir themselves and “ought” to have done so.

IMHO that normative theory makes for a good tee shirt slogan, not a plausible set of reasonable expectations about actual popular behavior. Put aside “oughts” for a second and focus on political reality: There is zero evidence that We the People are cowed by precedent-based rhetoric into abandoning our passionate desire to overturn precedents we hate. Jackson’s Democrats were unimpressed by the Madison/McCulloch Theory that the BUS’s constitutionality was settled by precedent. Likewise, Lincoln and the Republicans he rallied were unimpressed that the Compromise of 1850 settled the controversies over the Fugitive Slave Act. And so on. It turns out that Madison’s theory of precedent is a “directly self-effacing” theory in Parfit’s sense: We the People reject the idea that We the People settle constitutional ambiguities by the sorts of decisions specified by Madison.

3. Was Madison sincere about his theory of precedent?

Madison, of course, was a smart guy, which leads to the question: Did he seriously endorse the theory of precedent that he laid out in his 1831 letter?

I have my doubts. In any case, I suspect that Madison was actually addressing an entirely different constitutional problem than the BUS’s constitutionality, the ostensible topic of that letter. My suspicion is that Madison was hinting to Ingersoll that the problem of slavery’s status under the United States Constitution could and should be settled by morally arbitrary precedents like the Missouri Compromise. In June of 1831, the time was ripe for a Virginian with ameliorative instincts to make such an argument. William Lloyd Garrison had just founded The Liberator, incensing Southerners with fears of “incendiary” northern publications’ inciting slave rebellions. (Nat Turner would launch his rebellion just a couple months after Madison’s letter, confirming in Southern minds these fears). It would naturally be appealing to a member of the founding generation that the growing moral impasse over slavery’s constitutional status could be laid to rest by a precedent like the Missouri Compromise that assiduously avoided taking sides about slavery’s moral status.

Alas, if I am correct, that fond notion is unsustainable. The agreement of multiple parties to some deal through those parties’ “varied ascendancy” in government will always meet with the rejoinder that those parties really did not speak for Us the People. And the proof that this rejoinder is correct will always be the sheer brute fact that We the People won’t shut up about disagreeing with the precedent that is supposed to settle the dispute.

Posted by Rick Hills on December 2, 2021 at 02:25 PM | Permalink


Could someone please re-open the comments to the Hiring Spreadsheet thread? The market is still going on, and folks in the spreadsheet that's linked are asking.

Posted by: Request | Dec 4, 2021 2:20:34 PM

Precedent is legitimate as long as it does not contain an error in Substantive or Procedural Due Process Law.

Any precedent that would deny the personhood of any beloved son or daughter regardless of that person’s location, including if that beloved son or daughter is residing in their mother’s womb, is not legitimate and must be overturned.

Being a human person does not depend on one’s location.

Posted by: N.D. | Dec 3, 2021 8:55:25 AM

Rick, a right to privacy, in no way, shape, or form, gives any person a right to deny any innocent beloved son or daughter their inherent, Unalienable Right to Life, the securing and protection upon which their inherent Right to Liberty and The Pursuit of Happiness depends. Abortion, the intentional destruction of the life of a beloved son or daughter residing in their mother’s womb, is a direct violation of an innocent son or daughter’s Right to Life, thus every State has the obligation and the duty to protect the life of an innocent human person from harm, and Peter Kreeft is correct that abortion is not and never was protected in our Constitution as a private act.
You have been you since the moment of your conception and you were never, during any stage of your life, a place or a thing. Personhood is a reality for every son or daughter conceived by human persons. It is up to the State to demonstrate how it is possible for a man and woman to conceive a son or daughter who is not a human person, if they want to deny personhood to some sons and daughters in order to not secure and protect their inherent Unalienable Right to Life.

Viability does not change the essence of personhood, because viability does not change the essence of being, in essence, a beloved son or daughter.

Posted by: N.D. | Dec 3, 2021 8:43:58 AM


I've read the letter since our last exchange, and one of the parts of the letter that struck me was Madison's point that the BUS was renewed every year. That means congresses stretching over decades debated and renewed the charter annually. Plus, as Madison pointed out, the fighting in the different branches when the Act was being passed means that the matter was revisited over and over and over and over and over again with the same result.

That hasn't really occurred with Roe (or many decisions). This got me wondering what precedents might have an equal amount of review and the only one I could come up with was Judicial Review. That may be the only true precedent we have in a Madisonian sense. One might be able to add Incorporation.

This, then, led to the thought that maybe precedent doesn't really exist. After all, the Jacksonians rekt Madison's precedent, and their precedent was rekt, in turn, by learning from recurring financial panics.

In the end, it seems we are approaching a stance near to a theory of government legitimacy which states a government is legitimate so long as there is no successful revolution against it. The same may apply to precedents.

In any event, a very fascinating topic. I will have to devote more thought to it.

Posted by: thegreatdisappointment | Dec 3, 2021 3:33:23 AM

ND, here’s a pro tip: Take Justice Blackmun’s statement about the relevance of fetal personhood to the right at issue in Roe with the same skepticism that you give to the rest of the opinion. (For one thing, the “state action” limit on the 14th Amendment undercuts Blackmun’s claim about what the 14th Amendment allegedly guarantees).

And read Boonin to understand what’s really at stake with this particular liberty claim.

Posted by: Rick Hills | Dec 3, 2021 12:35:27 AM

“...because it might be that the right to an abortion doesn’t depend on any denial of anyone’s personhood.”

There is no inherent right to abortion although some persons justify abortion by denying the personhood of a beloved son or daughter residing in their mother’s womb.

“If this suggestion of personhood is established, [Roe's] case, of course, collapses, for the fetus' right to life would then be guaranteed specifically by the [14th] Amendment.”- Roe v. Wade, 410 U.S. 113 (1973)

Posted by: N.D. | Dec 3, 2021 12:22:49 AM

It is definitely true, greatdisappointment, that Madison imagines a liquidating precedent as a series of decisions taking place over an extended period of time — for instance, the initial enactment of the first BUS charter in 1791, the decade debate about the BUS during the 1790s, the debate over the lapse of the charter in 1811, the bad experience of the War of 1812, and, finally, the renewal debate in 1816. That’s a quarter century of decisions, more or less!

So maybe this last round of fights over Roe-Casey will finally lock in or defeat that precedent after close to a half-century of fighting. But…how will we know? As I was saying to Will Baude, it seems that the “popular consensus” test practically invites further fighting just to test whether there is a popular consensus.

Perhaps precedent is just like Potter Stewart’s porn: You know it when you see it.

Posted by: Rick Hills | Dec 2, 2021 10:37:57 PM

Rick, thank you for your response, and I think your response brings up the answer I was trying to provide, but was unclear.

My caveat here is that I haven't read the letter and so what I write may well not be borne out by the context of the letter.

Instead of "stable two party system" I may have been better served by using the phrase "moment in time". As a result, we may interpret "under the varied ascendancy of parties" to mean time passes (probably significant time) between when a decision is reached and when it becomes a precedent.

As a result, I'm not sure Madison's analysis means a particular group made a decision and so all people are bound forever and always, but rather that this particular group made a decision, and then the Jacksonians became ascendant and either did not challenge or failed in their challenges, and then another party became ascendant and either did not challenge or failed in their challenges, and at that point the decision begins to take on a more settled precedent air as several factions of American life have been ascendant and had their chance to challenge the decision and either did not do so or failed.

One may like it to Joseph Story's finding on one territory being subjugated rather than occupied: if a force occupies a territory long enough and there are no challenges to that occupation and it goes on so long that the countries of the world recognize the occupied territory as belonging to the occupier--again, with not contested claim--then the occupied territory becomes subjugated and sovereign territory of the occupier.

This idea seems to address some concerns vis-a-vis Roe. It was decided, and then the New Right gained ascendancy as is, really now, getting its chance to voice its objections and challenges to Roe, meaning we may understand that Roe does not have a Madisonian precedence yet because we're only just now dealing with the first round of new ascendant parties and their challenges.

Posted by: thegreatdisappointment | Dec 2, 2021 7:45:48 PM

Thegreatdisappointment suggests that “Madison isn't talking about a stable two party system, but rather may be discussing something more like ‘factions’ or what we may call ‘interests.’”

That seems right to me: The First Party System barely qualified as a party system, and the second one was not born until a year after Madison’s letter.

But does this concession, which I’m happy to make, change the analysis? Interests, factions, groups, etc, might all agree with each other in reaching a consensus position — but does that make such a position the view of We the People, or, more prosaically a critical mass of Americans? If Campbell, Converse, Miller & Stokes (and their latter-day intellectual heirs like Jon Zaller and maybe Larry Bartels) are correct, these mobilized groups still represent a tiny fraction of the body politic. Maybe 5% or so? What gave them the right to decide for the rest of us what the Constitution means? If someone else comes along — Andrew Jackson, say — and mobilizes a different set of folks with a different view, is it an adequate answer to say, “well, a cabal reached a different conclusion earlier about what some ambiguous phrase means, and you’re bound by that decision”?

Posted by: Rick Hills | Dec 2, 2021 7:26:37 PM

Rick, a very good post.

I wonder, though, when Madison wrote about "under varied ascendency of parties" if he meant established political parties, or meant it in a sense closer to his Federalist Papers essays use of "factions".

To be sure, there are decades between the two writings and Madison, by this point, may have given up the idea of a more parliamentary mixture of several factions forming ad hoc voting blocs on different issues (out of ten parties, parties A, E, and J may form a voting bloc on issue 1, but parties E and J may oppose party A on issue 2).

Still, the use of "varied" makes me think Madison isn't talking about a stable two party system, but rather may be discussing something more like "factions" or what we may call "interests".


Posted by: thegreatdisappointment | Dec 2, 2021 7:12:07 PM

ND asks, “I am wondering what precedent gives us permission to deny the personhood of a son or daughter residing in their mother’s womb?”

It is not obvious that ANY precedent does, ND, because it might be that the right to an abortion doesn’t depend on any denial of anyone’s personhood. For a defense of the idea that abortion should be legal even if the fetus is a person, see David Boonin’s “Beyond Roe,” at https://www.colorado.edu/asmagazine/2020/04/02/beyond-roe-why-abortion-should-be-legal-even-if-fetus-person

Of course, I am assuming that Judith Jarvis Thompson’s famous consent-based theory that Dave Boonin is expounding is consistent with Roe. Is it? Or is it “Beyond Roe,” as David assumes? If the viability principle really is not part of Roe’s core holding (despite what Casey said), then why not? CJ Roberts’ questions during Dobbs explored reformulating the Roe right as a consent-based right rather than a right based in interest balancing. If that’s a legit move consistent within a legit theory of stare decisis, then the personhood of the fetus is simply beside the point, because one should not be forced to rescue persons who happen to depend on one’s making immense bodily sacrifices on their behalf. Forced labor on behalf of anyone’s rescue might just be too great a burden for the state to impose.

I’m not defending the Thompson-Boonin consent-based defense of the right to terminate a pregnancy. I’m just suggesting that much of the current argument about this right, including ND’s invocation of a theory of personhood, might be beside the point if we stop thinking about precedents and start thinking instead for ourselves. Roe-Casey might be a ball and chain not only for these precedents’ opponents but also their supporters.

Posted by: Rick Hills | Dec 2, 2021 6:56:14 PM

I am wondering what precedent gives us permission to deny the personhood of a son or daughter residing in their mother’s womb?


Posted by: N.D. | Dec 2, 2021 6:19:36 PM


I think he was sincere. Deadly wrong, I do agree. Yet, he was sincere.

Not in vain, you couldn't read one crucial word in his letter:


The constitution, dictates principles, not laws for the lawmaker or judges as phrased by him. Sometimes, one constitution, dictates the law. But, it is so typically, when dealing with institutional provisions. Fixed law. Like, one president, would or can serve two terms, no more than that.

But, typically principles!

What does it mean:

It means, that, it is, like precedent, guiding, not binding. For, you can hardly find, two cases, where, both bear, precisely the same legal and factual configuration. As such, you must match, one unique case typically, with the law, precedents, and constitutional principles. So, the outcome, is the art of ad hoc adjudication, narrowly tailored for one off case one judge handles (and of course, others would blame then judges, for changing their mind, every Monday and maniac day).

Sometimes, principles are by nature, in conflict. One judge must prevail, what constitutional principle, would have the upper hand.

So, there in no way, one can claim, that this business is fixed, stable, and ever lasting one. For it would contradict then, the essence of the law and legal matters:

You deal with constitutional principles, and unique case or unique configuration typically.

Not mentioning, social and technological shifts, mentioned somehow in the post (through parties let's say).


Posted by: El Roam | Dec 2, 2021 4:39:35 PM

Will writes: “Well, maybe the permission can only successfully work as a shield if the speaker pretends (or even convinces themself) that it is more than that!”.

A theory that dares not speak its name! (I always get philosophical jargon mixed up, but I think that makes the “shield” theory an “indirectly self-effacing theory” in Parfit’s terms).

Posted by: Rick Hills | Dec 2, 2021 4:11:31 PM

Well, maybe the permission can only successfully work as a shield if the speaker pretends (or even convinces themself) that it is more than that!

Posted by: Will Baude | Dec 2, 2021 4:00:30 PM

I love Richard’s “Permissions” theory of precedents as “shields and shortcuts.” (Link here btw: https://texaslawreview.org/precedent-as-permission/ ). But is this theory a good fit with Madison’s letter? And can liquidating precedents of the Madisonian type provide the political cover that Richard’s theory requires, when the stakes are high and emotions, fierce?

On the first question, at least, as the 1831 letter is written, Madison seems to laying out a legal obligation, not just deflecting accusations of inconsistency with a “shield.” (Consider this sentence: “A veto from the Executive under these circumstances, with an admission of the expediency & almost necessity of the measure, would have been a defiance of all the obligations derived from a course of precedents amounting to the requisite evidence of the national judgement & intention.” It sure sounds like he is attacking the Jacksonians as defying obligations, not merely defending himself as avoiding disruption).

On the second question…we will see! Will Roberts deflect blame fromGOP voters if he sticks with Roe-Casey’s viability line? Or will he placate Dems if he ditches viability but generally sticks with the Roe-Casey’s protection for a right to terminate a pregnancy on “consent-based” grounds (upholding MS’s law on the ground that 15 weeks is long enough to know that one is pregnant and so provides for tacit consent to carrying the pregnancy to term)?

Let’s wait and see, but I have my doubts that precedent will provide much of a Re-style “shield” here, anymore than the Compromise of 1850 saved the Whigs in 1854.

Posted by: Rick Hills | Dec 2, 2021 3:32:55 PM

I wonder if it would be profitable to think about liquidation in terms of the "permission" model of precedent that Richard Re has laid out. Liquidation seemed to do the most work for Madison when he could use it to explain why he was not bound to do something radically unsettling that it might seem like his constitutional priors would otherwise obligate him to do. The same for, say, Ely on Roe, or many others on legal tender, etc.

Now I have my misgivings about the permission model as an approach to precedent, so this is a suggestion against interest, but I do wonder if it is useful to explain when the liquidation model is really doing work.

Posted by: Will Baude | Dec 2, 2021 3:07:54 PM

Can you please reopen the comments on the hiring threads?

Posted by: Anon | Dec 2, 2021 3:03:13 PM

Will, I am just not sure.

On one hand, I wonder whether such a theory of precedent would be too unsettling to do the work required of a settlement theory. After all, the whole point of any settlement theory (a worthy one IMO!) is to let us all get on with our lives by avoiding the tediously endless re-litigation of contentious issues. But a theory of settlement that allows liquidations to be unseated by a sufficiently determined response practically invites endless organizing by people intent on unseating the reigning precedent. After all, how can one tell whether some allegedly liquidating precedent is actually liquidating until one puts it to the test of aggressive mobilization? (Think of the Garrisonians, a tiny faction in, say, 1831, whose relentless attacks on slavery irritated everyone with their futility — until they didn’t).

On the other hand, maybe we know settlement when we see it. Like everyone, I have an intuitive (albeit ineffable) sense of when further debate on an issue is the work of an annoying nudzh rather than a policy entrepreneur with a prayer of nudging the sash of the Overton Window. (I usually feel this intuition during faculty meetings rather than SCOTUS oral arguments).

So, like I say, I am not sure!

Posted by: Rick Hills | Dec 2, 2021 3:02:13 PM

Rick, I love this post, as I love all of your criticism of liquidation, but (assuming it is correct descriptively) I wonder if what you see as a bug might actually be a feature. Maybe the force of liquidations simply decline over time if the issue in question is one that is sufficiently important that parties organize around unsettling it. Would that be so incoherent?

Posted by: Will Baude | Dec 2, 2021 2:35:07 PM

Post a comment