But, well, my post is pretty topical, so up it goes! Warning: it is longer than my normal posts. If you’re not in a longform mood (or if you are really hankering for Arkansas-specific content), feel free to nope out. Also, I’m adding some illustrations to keep it, well, tonally consistent with my ordinary posts. Enjoy!

Subscribe now

Sidney Hook’s The Hero in History is a brilliant, wrong book. Published in 1943, amid war and against the rising prestige of historical determinism, it asked a simple question: given the impossible complexity of broader human trends, how and when do individuals actually change history? Hook’s answer is still the best place to start thinking about the problem. It is also, I think, not quite right. The easiest way to see why is Donald Trump.

In this review, I am going to revise Hook’s theory, drawing on concepts from evolutionary biology.

Hook’s Theory

The Great Man Theory posits that history is the product of unique, heroic people who drive history forward. The opposite theory, historical determinism, argues that history is the product of forces larger than any one person. When a person seems to have singular influence over events, it is really a manifestation of bigger trends, like industrialization, class struggle, etc. The debate raged through the 20th century, eventually giving way to a spectrum of arguments about the extent to which individuals truly can or do alter the trajectory of history.

Sidney Hook was an anti-Soviet socialist, a leading figure in the philosophical pragmatist movement. In the early 20th century, Marxist scholars pioneered determinist theories of history in which economic forces and class struggle drove history forward while kings and presidents just happened to be present when those forces acted.

The determinist theories were sufficiently persuasive to render the pure version of Great Man Theory largely untenable. For example, 19th century histories often suggested the Industrial Revolution was the product of specific men’s genius. In the 20th century, a variety of determinist theories dislodged it, to the point that serious historians of the era today argue over which determinist factor mattered most rather than whether any particular individual mattered tremendously to the overall story.

Hook and other pragmatic philosophers seemed to recoil from the impersonal arguments of the Marxists. His theory offered a middle road between the Great Man and determinist camps.

Hook argued that broader trends create opportunities for a person to make a difference. He categorized the people who do so as either “eventful” or “event-making.”

An “eventful” person is like the Dutch boy who plugged a hole in a dike. That boy doubtlessly affected history (or would have if the story were true), but only by doing what basically anyone would have done. Broader forces dictated that some particular person would be in a position to affect history, even if it did not really matter who. This is not so different from determinist arguments.

By contrast, an “event-making” person is possessed of “extraordinary talent of some kind and not merely the compounded luck of being born and of being present at the right place at a happy moment.” Through a combination of willful purpose and extraordinary talents, an event-making person can permanently alter the course of history in accordance with his vision.

His most notable example is Vladimir Lenin. Broad forces led to the Russian Revolution of February 1917, but Hook argued that the Bolshevik revolution of October 1917 was the unique contribution of the event-making Lenin.

What made Lenin so special? Hook notes an “organizational genius” and talent for inspiration that the other Bolsheviks lacked. Without him, the October revolution fails. Without the October revolution, there is no Soviet Union. Russia does not end up a pariah after World War I. The politics of Germany are untroubled by fear of communism. The German extreme right does not develop in the same way or to the same extent—no Hitler. You can fill in the rest of the blanks yourself, but the idea is obvious: Lenin’s special talents transformed the world.

Hook also categorized various historical figures as eventful or event-making, and he offers some bold and thought-provoking takes.

• Constantine, the Roman emperor who converted to Christianity, is merely eventful—anyone would have converted to Christianity faced with the political incentives of Constantine.

• Justinian, the emperor who codified Roman law, is eventful—he was a “mediocrity” whose acts had great effects, but who did not contribute to them in any unique way.

• All of the women he discusses he labels as eventful—Cleopatra, Catherine the Great, Madame de Pompadour, and the Byzantine Empress Theodora. (He actually contradicts himself on Catherine the Great, describing her as both eventful and event-making over the span of three pages.) Of Theodora, he argues that “had she applied the habits of a thrifty housewife to the royal economy her influence would have been more lasting.” Sexism aside, this is probably the single worst take of the entire book, and I’ll explain why in a footnote for people not looking for a lecture on this specific subject.1

He mentions Thomas Jefferson and Abraham Lincoln, but says they resist easy categorization. Both had “enormous” influence, but Hook contends neither made a unique contribution. Jefferson’s phrasing was memorable, but his ideas were common in his circles. He didn’t really explain Lincoln, but we can fill in the blank. Perhaps whoever led the Union in the Civil War would have accomplished the same outcome and overseen a similar program of emancipation once the South’s political power was neutralized.

Hook’s Flaw

Setting aside specific problems discussed above, Hook is persuasive regarding the idea that complex historical circumstances create the opportunity for individual actors to leave a distinct imprint. We’ll get to Trump in a minute, but his Lenin example largely stands up to scrutiny.

The events of the October 1917 revolution were so fraught with contingency that Lenin being absent seems like it obviously would have screwed the whole thing up. For starters, in April 1917, the Bolsheviks in Russia were accommodating the provisional government set up by the February revolution. From exile, Lenin demanded a change in the Bolshevik line to oppose the provisional government, setting in motion the chain of events that would lead to the October revolution. In September and October, Lenin again spurred the Bolsheviks forward to take bigger chances in challenging the provisional government. So, yes, Lenin seems to prove that sometimes a specific person—a Great Man—can change history dramatically.

But here is where Hook goes wrong. Hook says event-makers are necessarily forceful, willful, talented people, possessed of singular focus and determination. But the Lenin example does not really prove that.

What Lenin obviously did that was irreplaceable was pushing relentlessly to move faster, sooner, and take advantage of what ended up being a small window of opportunity. Something in Lenin’s psyche drove him to take action at that moment when there was no obvious reason to do so—if there had been, the Bolsheviks would have already been moving in that direction without Lenin.

Hook was a clever philosopher and student of history, but his opinions on specific people come off as glib. He categorizes eventful and event-making people based on how impressive their specific talents seem to be, not on whether they contributed something idiosyncratic to the outcome of events. These errors in application betray a more fundamental problem: he assumes that an event-making person must have some noteworthy positive talent or trait. If they do not, they are written off as eventful.

This is where Donald J. Trump enters the debate. He embodies the flaw in Hook’s theory.

Evolutionary Landscapes

A fitness landscape, frequently discussed on Slate Star Codex/ACX, describes a concept from evolutionary biology. Imagine a two-dimensional graph, with the x-axis being some trait (e.g., intelligence) and the y-axis being evolutionary fitness. Obviously, this thinking can extend beyond one-dimensional measurements. In reality, evolution is working across countless dimensions.

Intelligence comes with some tradeoff. A big brain lets you do more complicated things, but it is expensive in terms of energy. So you can imagine that for a lot of animals, the sweet spot (a local maximum on the fitness landscape) comes from being just smart enough to do some important thing. Maybe being smart enough to open nuts is a big improvement, but the next useful trait from high intelligence doesn’t happen until you can use simple tools. This is illustrated in the example below, lightly adapted from Wikipedia—you can imagine “A” is opening nuts, “B” is simple tool use.

In the absence of some big environmental change, evolution can only really push an organism toward higher fitness (i.e., uphill in the graph above). The problem is that if you’re at local maximum A, the only way to get to absolute maximum B is by first descending, by becoming less fit temporarily. Evolution does not work like that. It works slowly, and the trait will not be selected for if it is not helpful.

In the real biological world, you can get from A to B in one of two ways.

First, the environment can change, shifting the landscape so that there is now a continuous path upwards from A to B. Say there’s an ape species that usually lives in the trees of dense tropical forests, where strength is very important. But then there’s a dry period where the forests retreat. Suddenly, point A might shift in such a way that there is a continuous upward path to B, the high-intelligence/low-strength option. That is one important part of the story of human evolution.

The other way is that some of the species can be plopped down in a starting position that is closer to B. Think of a fish being dropped by a bird into a new lake. The fish already in the lake might not be able to get from A to B, but the new fish might happen to already be closer to B, enabling it to reach the absolute maximum of the fitness landscape.

Now, how does this help explain Donald Trump?

The Evolutionary Landscape of Republican Presidential Primaries

Consider a 3D fitness landscape where the Z-axis is popularity of Republican candidates in a primary field (i.e., among Republican voters), the X-axis is candidate virtue (loosely defined, but basically “how good of a person do you seem?”), and the Y-axis is reaction (how much you promise a return to the past).

Obviously, in the real world, you could have any number of axes to define fitness with greater precision, but once we go beyond three dimensions, we lose the physical legibility of the evolutionary landscape. A word on definitions here:

• Virtue means a lot of different things, but for electoral purposes, it represents a non-policy vibe of “goodness.” “Respectability” or “attractiveness to elites” are in the same ballpark, and you can feel free to think in those terms if your ideology prefers them. Being “good” would superficially seem preferable in a candidate, but consider all the times people of every political party wish for a “fighter” or someone willing to “fight dirty.” High virtue can also seem dull, and low virtue can seem more authentic and charismatic—consider how much cooler Han Solo is than Luke Skywalker.

• “Reaction” is not Republican orthodoxy circa 2014 (e.g., free trade, anti-regulation), but a representation of yearning for the past manifested as reactionary personal and policy preferences. This would include, among other things, social conservatism (distaste for homosexuality and transgenderism), religiosity, opposition to immigration, opposition to free trade, etc.

Here’s what that 3D fitness landscape might look like:

You’ll note that I have listed John McCain as the exemplar of a maximally virtuous Republican candidate and Mitt Romney as a little more reactionary and less virtuous. I’ll stipulate that I mean perceived virtuousness—there was a perception that Romney was a disingenuous flip-flopper while McCain was a virtuous war hero. (Also, full disclosure, I was a McCain staffer, and I briefly volunteered on the Romney campaign in 2012.)

You’ll note that Rick Santorum (notable Republican candidate of the 2008 and 2012 cycles) is listed as a somewhat non-virtuous reactionary. By this I mean not that he is personally non-virtuous, but that his political appeal did not rest so much on perceived virtue. He was a plain-talking social conservative willing to say transgressive things while castigating homosexuality, free trade, and all sorts of perceived “libertarian” things. If you prefer, he was not attractive to elites.

Stuck at Local Maxima (McCain and Romney)

An ordinary candidate could not really make the jump from Romney to the Trump absolute maximum in the same way that evolutionary processes can’t decrease fitness. In real life, animals can’t see their own fitness landscape, so they can’t plan out a way to get to the absolute maximum. Similarly, Republican candidates could see examples of success representing local maxima, but not the Trump absolute maximum. (And, if you’re being pedantic, technically we only know that Trump reached a higher local maximum. It is possible there could be another absolute maximum elsewhere.)

From the perspective of candidates unaware of the Trump absolute maximum, moving further up the reaction axis tends to kill Republican presidential candidates. There were three immediate reasons why fitness decreased as reaction went up. First, extremely reactionary candidates couldn’t attract Republican big money donors, who were generally wealthy people who profited off of libertarian-ish policies and didn’t have time for culture war nonsense. Second, if the candidates were too reactionary, the press turned against them and poisoned them for the electorate. Third, if they were too reactionary, their chances of winning the general election went down, leading to an exodus of pragmatic voters.

This was why Santorum’s campaign basically “died” (i.e., his candidacy failed) in the valley between Romney and Trump. (Also, I can’t put “died” in this context without hearing Trump’s voice add “like a dog!”) That space was not good for maximizing Republican attractiveness.

How Trump Discovered the Absolute Maximum

Trump was a candidate who could explore the low-virtue, high-reaction space in a way other Republican candidates could not. He was the quintessential fish dropped from a different lake.

Unlike other candidates, he had money and name recognition such that he did not urgently need donors. But that alone would not guarantee he found the absolute maximum. Rather, it meant that if there was an absolute maximum, he would have the resources to find it.

Consider another situation involving a vastly wealthy billionaire with name recognition. In 2020, former New York Mayor Michael Bloomberg entered the Democratic presidential primary intending to run as a kind of moderate, pro-business liberal. Several other candidates, most notably Elizabeth Warren and Bernie Sanders, torched him for various personal problems and for being a billionaire. This basically worked, and his campaign failed.

In light of Trump’s experience, I think we can say that what Bloomberg did “wrong” was that he stayed in the conventional presidential primary fitness landscape when he did not have to. Obviously, he deserves some credit for expressing what seemed to be his true, moderate beliefs. But that area has already been thoroughly explored by other candidates, most notably Joe Biden, who ended up winning. Money and name recognition can allow a candidate to explore parts of the landscape that are otherwise inaccessible. That was where Trump thrived.

More to the point, Trump’s seemingly authentic views and instinctive vitriol enabled him to naturally reach high levels of reaction very quickly. For example, one of the very first incidents in his 2016 campaign was when he suggested Megyn Kelly was having her period when she asked him tough questions in a debate. Another early example was when he said a judge who ruled against him in a fraud case only hated him because the judge was Hispanic. These incidents spurred much media coverage and signaled to voters that this was a man of genuinely high reaction.

The low virtue part of the fitness landscape also mattered. A devout candidate like Santorum could not or would not strongly appeal to voters with a taste for boorish machismo or, if you prefer, a “fighter.” They also lacked Trump’s entertaining shtick—the impossibly cartoonish narcissism, where everything is the biggest, the most beautiful, the yuuuugest, etc.

A run-of-the-mill candidate who either acted atrociously or embraced such high levels of reaction would be marginalized, starved of the money needed to campaign. Highly reactionary figures had hitherto been quite religious, and therefore unappealing to the low-virtue voters. A wealthy self-funded candidate mostly wouldn’t embrace such high-reaction views. Trump was uniquely well-positioned: reactionary and low-virtue enough to reach the fitness peak of Republican politics. His lack of personal virtue (sexual assault claims, multiple affairs and divorces, allegations of fraud, etc.) meant he could also appeal to a segment of the electorate that was looking for arrogant, politically incorrect machismo.

Is Trump a Hookian Event-Maker?

I don’t want this to be a “Trump sucks and here’s why” review. But I think for the Hook discussion to make sense, we need to briefly discuss how Trump’s purported qualities are problematic for Hook’s thesis. Hook looks for uniquely strong qualities in an event-maker. Lenin had a genius for organization. What does Trump have?

Here is the steelman case for extraordinary Trump qualities:

• His media instinct is remarkable. He dominated the 2016 news cycle so thoroughly that one study estimated he received roughly $2 billion in free media coverage.

• He has an uncanny ability to find and exploit an opponent’s vulnerability, branding them with nicknames that stick precisely because they crystallize a pre-existing doubt: “Low Energy Jeb,” “Lyin’ Ted,” “Crooked Hillary.”

• His rallies generated a fervor that no Republican candidate in living memory has matched.

• His willfulness is extraordinary. After January 6th, virtually every political observer assumed his career was over, yet he clawed his way back to the presidency.

But each of these positives is narrow, local, and context-dependent in a way that Hook’s framework can’t really deal with.

• Trump’s media dominance worked because he is willing to say outrageous things without shame. But that’s a specific trait that happens to fit the contemporary media environment, where algorithms reward provocation and cable news is desperate to fill airtime.

• His talent for branding opponents is really a talent for cruelty wielded without inhibition. In an era with less appetite for public cruelty, such branding would backfire.

• His rally energy derived less from inspiration than from the pleasures of collective transgression, the thrill of a leader who will say the thing you are not supposed to say. This is particularly effective in an era where many feel political correctness prevents frank discussion. While there is always a market for such transgression, it is not always equally lucrative.

Strip away the context and none of these qualities would make him effective in any other domain or era. It is hard to imagine him, say, challenging a Churchill for leadership of the United Kingdom. He is not a person of extraordinary ability who entered politics; he is a person whose particular set of fixations and deficiencies fit a historical moment.

This is precisely what the fitness landscape framework predicts. The qualities that look like talent are really the qualities of fit—the specific shape of the key matching the specific shape of the keyhole.

Another way to see this is to consider another seemingly superlative quality of Trump’s: his willfulness. Refusal to admit defeat can be admirable when it is paired with pursuit of some noble goal. Consider Alfred the Great’s triumph in (more or less) creating England, or Charles de Gaulle’s perseverance in forging a new French government in exile during World War II.

But Trump is not willful about policy per se—consider his frequent changes to the objectives of his Iran War. Instead, his willfulness takes the form of unwillingness to accept criticism. That willfulness is inseparable from his shamelessness, and his shamelessness is inseparable from whatever psychological baggage prevents him from processing negative feedback about himself.

Is Trump Merely “Eventful” in Hook’s Theory?

But then the question arises: could it be that Trump is just an eventful person and not an event-maker? If that is true, then anyone in Trump’s position would do roughly the same thing. That seems implausible, even unrealistically dismissive of Trump’s influence.

Trump started a massive trade war with the ostensible goal of weakening China and/or returning manufacturing jobs to the United States. He did this at a time when AI development increasingly seems likely to decimate manufacturing jobs. He then approved the export of advanced AI technologies to China, so long as the U.S. got a “cut” of the profits of the sale, much to the chagrin of AI policy commentators.

Trump does not behave in ways that an ordinary person would even if they seemed to hold the same policy views. This belies the idea that he could be eventful—he is acting in idiosyncratic ways that will have extended ramifications in the decades or centuries to come.

But where does this leave us regarding Hook’s theory? Two propositions seem clear: Trump is not “event-making” in the sense of possessing remarkable qualities, but he is “event-making” in that he is leaving his own mark on history in a way no one else seemingly would have.

My Modification of Hook: Keyhole Theory

Hook’s theory requires modification to account for figures like Trump, who have no grand qualities but who do influence history in personally irreplaceable ways.

My own proposed solution: abandon the eventful/event-making framework entirely and consider a less dichotomous alternative. I call my alternative the Keyhole Theory.

Here’s how it works:

• The complexity of broader forces and historical circumstance inevitably creates an opportunity for a specific kind of person to ascend to a position of prominence. This is consistent with Hook’s theory. I will call this kind of opportunity a “keyhole” moment.

• A person who actually does ascend to prominence in a keyhole moment—the “key” person, to continue the metaphor—does not need to have individually exceptional qualities. They only need to be warped by experience and circumstance to fit the keyhole.

• A person who successfully rises to prominence in a keyhole moment can affect the trajectory of world history in ways that are not dependent on macro trends.

This also comports with statistical reality regarding “great” men or women of history. There are currently about 8.3 billion people on Earth and 350 million Americans. Is it plausible that Trump can exert his personal influence because he has some set of qualities that score significantly higher than the other 340 million Americans? Is he more charming than the best actor, more knowledgeable than the best policy analyst, richer than the richest mogul? Of course not. But as we saw in the fitness landscape context, he had the right combination of qualities to be the key for the keyhole moment of 2016.

Keyhole Theory as Founder Effect

Keyhole Theory is fundamentally just evolutionary biology applied to a particular animal (human beings) with a particular kind of intraspecies selection pressure. Evolution is about flukish variations fitting into ecological niches and improving fitness. Keyhole Theory is about flukish variations of humans fitting into the social and political niches that emerge from the way we humans deal with the complexities of our existence in our environment.

In evolutionary biology, the founder effect refers to the genetic influence of a small group of individuals who are separated from the general population and create their own lineage. That lineage forever bears some genetic hallmarks of the founders.

Return to our ape species driven from the retreating forest onto the savanna. That particular population carried its own genetic quirks—maybe a slightly different gait, a particular digestive enzyme, a tendency toward aggression or cooperation that another population of the same species did not share. These traits had no special significance in the forest. But once the environmental shift pushed this population onto the savanna, those incidental traits became the foundation of everything that followed.

The founder effect means that the specific apes who happened to fit the moment leave a permanent genetic signature on their descendants. It did not matter that they were not the strongest or smartest apes in the forest. What mattered was that their particular combination of traits was adaptive when the environment shifted, and then all their baggage came along with them.

So too with history. Constantine was almost certainly not the greatest military commander of his age. He was probably not a unique person in any meaningful sense. But he fit the keyhole moment, and then his personal idiosyncrasies—including his particular relationship to Christianity—carried a founder effect forward. Nearly two billion people observe Christmas because of the specific quirks of one man who happened to fit the political niche of a Roman civil war.

Similarly, Trump explored a new part of the fitness landscape of Republican presidential politics, and thus he was a key that fit a keyhole moment. But once he ascended to power, his idiosyncrasies will have ramifications for centuries because it is highly unlikely other people who could have found that absolute maximum would have used power as he has.

The Napoleonic Challenge to Keyhole Theory

Napoleon is the hardest case for Keyhole Theory and the strongest case for Hook’s event-making man. He seems to transcend any single moment—a military genius who conquered most of Europe, reorganized its legal systems, and cast a shadow over the entire nineteenth century. If anyone was a Great Man in the traditional sense, surely it was Napoleon.

I think the case weakens the more context you look at. By 1799, revolutionary France had spent a decade cycling through governments that could not consolidate legitimacy. The republic needed military prestige to survive against coalitions of monarchies, but it also needed to not look like a monarchy itself. This created a very specific keyhole moment: someone who could wield authority with monarchical confidence while claiming revolutionary legitimacy, someone whose power derived from battlefield merit rather than bloodline. That is an oddly-shaped keyhole that excluded almost everyone.

Napoleon’s particular combination—Corsican outsider status, genuine military brilliance, ideological flexibility bordering on total indifference to principle, and an ego so vast it could fill the vacuum left by the Bourbons—constituted a key shaped for a particular lock.

Napoleon seems to refute Keyhole Theory because of everything that came after he took power. That would include the Napoleonic Code, the reorganization of Europe, the military campaigns that redrew the map of the continent. Surely, these were achievements that required tremendous personal qualities, not just fitting a keyhole.

But I would contend these achievements are a complicated mix of founder effects and the products of larger trends. His legal reforms drew heavily on revolutionary-era legal projects already underway. His military innovations built on French revolutionary tactics that predated him. His ultimate trajectory of overextension, invasion of Russia, and defeat simply meant that he could not defy broader military/strategic trends. Napoleon’s founder effect was real and lasting, but the keyhole that got him into the position to idiosyncratically affect history was as contingent and narrow as any other.

The strongest version of the Napoleon objection is not that he was too talented for Keyhole Theory, but that his talents were so broad they would have found some keyhole in almost any era. I doubt this is true given the extreme contingency of his ascent to power. But Keyhole Theory does not say that actual greatness is literally impossible.

To stretch the key metaphor a bit, consider that as a person increases in talent, they become more like a skeleton key, able to fit more and more keyholes. If we take Napoleon apologists’ arguments for granted, his greatness does not change Keyhole Theory. It merely implies that some people’s keys are more robust to contingency.

And there are of course hypothetical people who could skeleton-key virtually any keyhole. If Superman were a real person, he would probably have a profound influence on history regardless of class dynamics when he was around. But this is a particular application of Keyhole Theory, not the primary way history works.

Concluding Thoughts on Greatness

The determinists are correct that larger forces create conditions for historical change. But they are incorrect that individuals are replaceable in history. Individuals who reach a position of authority can leave a founder effect that forever alters the course of events. What Hook got wrong was the idea that such people must be, in some sense, Great.

Hook’s fallacy was probably a consequence of human psychology. Humans, possessed of a brain that enabled conquest of the planet by relentlessly figuring out how things work, want history to make sense. We psychologically rebel against the idea that history is just too complicated to understand. And if you want to find a cause for events, what could be more flattering than the idea that history is driven by people with, as they say, Main Character Energy?

But it is statistically unlikely that current powerful rulers like Donald Trump or Xi Jinping are genuinely “great” figures in the sense of having some nearly unique combination of incredible characteristics. Rather than seeing that as a resigned concession that larger forces are all that matter in the world, we should take it as a lesson not to look to leaders for greatness.

Keyhole Theory liberates us from the need to think the Trumps of the world are “great” because they affect history in easily discernible ways. And perhaps if we de-romanticize the people who leave a mark on history, we will get fewer adventurers and more boring people who want to make things work better for the rest of us. That is, ironically, precisely the sort of pragmatic approach Hook himself championed.

Sanders:

The founder of Buc-ee’s [] put it best: ‘Life’s too short to try to build in places that people don’t appreciate what you’re bringing’…That’s why red states like Arkansas are surging and people and businesses are fleeing blue states.”

Aplin:

We’re growing. We’re building in a lot of places. But when you find a conservative, business-friendly state with a phenomenal workforce, it makes a difference. And I’m starting to realize life’s too short to try to build in places that people don’t appreciate what you’re bringing, versus a place like this where people do appreciate what you’re building[.]

…

That leadership, that concept, that idea of conservative, business-friendly, family-oriented concept, works so much better if it trickles down from the leadership, from the very top of the governor’s office, at the congressmen, at the senators, at the mayors, the representatives. And so, what we found when we got here in Benton (Arkansas) was a business-friendly town,” he added. “They said, ‘Let’s roll up our sleeves and let’s go get a Buc-ee’s built.’”

The two things I want to zero in on in these quotes: (1) what did Benton actually do to get a Buc-ee’s built; and (2) what does it tell us about Sanders’s claim that businesses are fleeing blue states and flocking to red states?

Subscribe now

Buc-ee’s Bucks

In July 2023, Benton, Arkansas’s solicited offers for public land it had recently acquired. Only one proposal was submitted: a $2.25 million offer from Buc-ee’s real estate affiliate for 29 acres of land. The city accepted that proposal. Job well done!

Except, a year later, with a month to go before the deadline to actually sign a deal, the city dropped the price from $2.25 million to $800,000, a last-minute 64% price cut.

The stated reason for the price cut was to buy access to a Buc-ee’s private road for access to a proposed future RV park. Also, Arkansas Department of Transportation requirements had increased Buc-ee’s road improvement costs to $5.5 million (unclear how large of an increase that was). The city attorney characterized the arrangement as shifting a cost Benton otherwise would have to bear later.

The minutes from the relevant Advertising and Promotion Commission meeting also offer these rationales for the price cut:

• The purchase price of the land was originally $439,599.25, so $800,000 was still a profit (oof).

• “[T]his would keep the project moving forward.”

What Would Be Nice to Know

If the offered rationale is the real one, it would be great if Benton could release some additional information:

• a calculation of why the appropriate price reduction was exactly $1.45 million (if you’re doing the math, it’s exactly $50,000 for each of the 29 acres purchased, which seems like a very round number that is not the product of detailed analysis);

• what an alternative RV park road would cost;

• a breakdown of Buc-ee’s claimed road expenses;

• the original proposal and proposed purchase contract;

• whether Buc-ee’s threatened to abandon the project without a price cut; and

• the terms of the road deal establishing precisely what Benton received (an easement? a contract right?) and who bears future costs, liability and repairs.

Also, Benton still making money is a silly rationale. That’s not really another category of information that would be helpful, but I just want us all on the same page. Losing $1.45 million is not suddenly fine because you’re still making $400,000.

A Speculation on What Might Have Happened

I stress again: this is speculation, and I don’t know what actually happened. (I would absolutely FOIA the relevant communications to learn more, but I honestly don’t have the time.) But the following makes sense to me as a possibility.

Buc-ee’s is famous for securing lucrative property tax and infrastructure benefits from communities. Until Arkansas Ballot Issue 3 really declares open season for Arkansas towns to give away their tax money, towns like Benton might not have been able to engage in the full suite of giveaways to land their Buc-ee’s. Reducing the price by $50,000 per acre would be a way of paying Buc-ee’s to come to Benton when other options were not available.

Again, we do not know if Buc-ee’s threatened to leave or anything like that. But it would seem strange if they truly didn’t have any role whatsoever in getting a $1.45 million reduction.

Regulatory Appeal vs. Paying Businesses to Come

As should be clear from the above, I don’t think paying businesses to come to red states should be the red state model, and it doesn’t have to be. The traditional good-government conservative economic argument was that if you create a predictable, fair regulatory system, businesses will come. If businesses must constantly fear that their profits will be impounded by high taxes (a blue state attribute), and their plans subject to capricious or politically motivated regulation, they will go elsewhere.

I think that is a perfectly valid way of competing. In theory, it is a way to climb the value chain. Leverage the friendly (but not corrupt) regulatory environment to attract business, and use the tax base provided by the business to improve education and infrastructure to attract increasingly profitable businesses.

In that system, attracting businesses is a proxy for a healthy underlying economy and regulatory environment. But because everyone knows it is a proxy for good things, politicians have an incentive to just pay businesses to come because it reflects well on them.

Paying businesses to come is not a symptom of a successful regulatory regime. It is even worse when the benefits are not available to all businesses, or the benefits are allocated as the result of a non-transparent process.

I don’t begrudge Buc-ee’s for cutting deals with these towns. If towns are willing to pay, why not take the money? But I do think this sort of dealmaking tends to happen with big, flashy projects—think NFL stadiums. That is not a good sign for their actual economic value.

If you’re thinking about long-term development, Buc-ee’s and retail stores generally are not the path to actual growth. They tend to cannibalize existing sales at other stores. They don’t have an obvious complement that would enable growth. People will stop at this Buc-ee’s to buy snacks and gas. It will capture some interstate spending that might have gone elsewhere and generate some tax revenue. Some residents from Benton and neighboring towns will get jobs at Buc-ee’s. But there’s no way to climb the value chain from there.

There is something genuinely inspiring about communities that move themselves up the economic development ladder. That climb is about many parts of society working harmoniously, each contributing what they do best. It requires blue collar workers doing blistering work to provide a competitive advantage, especially in the early years. It requires a culture open to trade and innovation. It requires engineers striving to catch up to the modern state of the art. It requires politicians making intelligent decisions that will take years or decades to bear fruit—well past their next election. Taiwan basically did a version of this extremely successfully.

If any part of that fails, the whole thing fails, and you get the middle income trap. And an easy way to fail at the outset is to go for flashy projects instead of investing in long term growth.

On Blue and Red States

I hate the blue vs. red state framing on development. I hate it in part because it is needlessly antagonistic in a way that politics did not used to be. But I also hate it because it truly makes no sense.

Red states as a group are not all the same, nor are blue states. The communities that comprise these states are not uniform either. A rural town in upstate New York has a lot more in common with a rural town in Arkansas than Bentonville has in common with a rural town in Arkansas.

If you want to stereotype red states as poor and rural and blue states as wealthy and urban, then “red state economies” and “blue state economies” would logically need and want different things. Putting cultural differences aside, Buc-ee’s does not present the same value add to a town in Rhode Island as it does in Arkansas. There are not so many people traveling through all the time in Rhode Island. Land is not as cheap. The opportunity cost of having a Buc-ee’s is higher. Rhode Island is not competing for the same businesses, and both states can get what they want without either of them losing.

People have been saying for decades that businesses are fleeing blue states, but the per capita GDP rankings and business formation rates seem to stay the same. Look at per personal income over time in Texas and New York below. Does it seem to tell a radically different story (other than Texas starting somewhat poorer)? To the extent it tells any story, does it tell a story of businesses fleeing New York?

Texas currently does attract more businesses than it loses. But it’s a pretty insubstantial amount in the grand scheme of things. To quote the Dallas Fed:

Job gains due to firm in-migration averaged around 0.24 percent of Texas’ total employment, exceeding the average out-migration rate of 0.15 percent.

That’s nice! Better to gain than to lose! But the difference between a vibrant and failing economy is not 0.09% of employment.

Instead of shallow, hand-wavy arguments about red states and blue states, we should demand clarity about development plans for specific areas. What is the specific long term plan for growth? People in red and blue states both deserve candid, thoughtful answers

Subscribe now

Arkansas is tied with several states for the country’s highest required score (termed the “cut score” in the biz)—270 out of 400 points.

There doesn’t seem to be a clear red state/blue state or regional dynamic here.

• Massachusetts, New Hampshire, Ohio, North Carolina, Texas, Arizona, and others require 270.

• Utah, Minnesota, Wisconsin, Nebraska, Missouri, North Dakota, Oklahoma, and and others require 260.

• New York, Kansas, Illinois, Montana, Idaho, and others require 266.

Why do some states have higher requirements than others? What does game theory suggest about why similar states don’t land at the same number? Does 270 make sense for Arkansas?

Subscribe now

The Argument for 270

Let’s assume for the sake of argument that the bar exam is a great way to test how good someone will be as a lawyer.2 Though, quick note, the only study I could find on this found that higher cut score states did not have fewer malpractice claims or ethical issues. This is unsurprising because 270 is just a little higher than 260 or 266, and I suspect most malpractice and ethics issues don’t stem from not knowing the relevant rules tested on the bar.

The simple, naive argument for 270 is “don’t you want high standards for lawyers?”

There are some obvious holes in this argument. The “high standards” proposition doesn’t explain why 270 should be the cut score instead of, say, 280, or 300, or even higher. Judging from the clustering of states around exactly 260, 266, and 270, the true reason no state goes higher is because no one is willing to drift too far from the group for fear of incurring blame if things go badly. But from the naive theory, why wouldn’t you set the score at 271 and only have great lawyers?

Obviously, there is a tradeoff. If you set the cut score at 400 (i.e., a perfect score), you will have roughly one brilliant attorney who can’t possibly serve all of the legal needs of Arkansas. The lower the cut score, the more people can become lawyers. And not every lawyer is doing the most high-stakes work. Society needs attorneys who can handle run-of-the-mill stuff—simple wills, simple criminal matters, etc.

Striking the optimal balance is an empirical question. Do we have too many lawyers? Not enough? Just right?

The Argument Against 270: We Don’t Actually Have Enough Lawyers in Arkansas

The obvious jokey casual empiricism here would be “of course there are too many lawyers!” If you didn’t have so many lawyers, big corporations wouldn’t always be trying to take advantage of you, and the lawyers couldn’t charge so much!

What is the empirical truth? Well, there are lots of studies showing there are not enough lawyers in Arkansas. But I get being suspicious of those studies—why would anyone ever write a report that says “we absolutely have enough lawyers, don’t even worry about it!” So, here’s a less empirical way to think about it.

Decreasing the number of lawyers benefits existing lawyers far more than the population writ large. Let’s go back to the lawyer in the I Am Legend 400 cut score hypo. If she is the only attorney in Arkansas, who is going to hire her? Whoever can pay her the most, of course! And who would pay the most? Either the wealthiest corporation, or the most crooked person.

Who would pay the least? People who are not wealthy and do not think they will need a lawyer. But this creates a predatory dynamic. Whoever gets the one brilliant lawyer can use her to take advantage of those without.

The counterintuitive insight here is that the average person is the one most screwed by having too few lawyers. If lawyers are more broadly available, the “predatory” lawyers do not have as easy a time (for example) filing frivolous lawsuits to push settlements.

Reciprocity With Other States Does Not Depend on UBE Scores

A few sentences ago, I mentioned that setting the cut score at 270 decreases but does not eliminate the likelihood that X will be a lawyer in Arkansas. The reason it’s not eliminated is that Arkansas allows reciprocity with 260 states like Missouri and Oklahoma. If you practice in one of those states for three years, you can just apply to practice in Arkansas—even with your 260 UBE score.

Now, this isn’t totally crazy. The three years part could stand in for the lower UBE. But it does show that the low UBE score by itself doesn’t mean much. The same practice requirement applies to 260, 266, and 270 jurisdictions. If we think a 270 score is the difference between a competent attorney and a non-competent attorney, why do we have a practice requirement at all for the other 270 jurisdictions?

The Better Argument for 270 That is Inapplicable to Arkansas

There’s a better argument for 270 built on game theory.

Let’s say Test Taker X is someone who is expecting to be marginal on the bar exam—maybe he did poorly in law school, maybe he just knows that he is a bad test taker, whatever.

How do we expect X to choose the state where he’ll take the bar? Presumably, he’ll choose the state with the lowest score requirement where he could plausibly practice.

If Arkansas wants to deter X from practicing in Arkansas, setting a cut score that is higher than neighboring states could plausibly deter X from taking the bar in Arkansas, which decreases the likelihood that X will be a lawyer in Arkansas (we’ll discuss in a second why it decreases and does not eliminate that likelihood).

So, the best argument for 270 that I can think of is “just be a little bit higher than the surrounding states.”

The problem is that Arkansas didn’t do that. Neighboring Texas, where many students take the bar, is a 270 state. So is Tennessee. Oklahoma and Missouri are 260.

Another Argument Against 270: It Hurts Arkansas Law Schools’ Rankings

Many law school applicants consult the U.S. News and World Report rankings of law schools when deciding where to go. The methodology of that ranking system is transparent: bar passage rate of first time test takers is a significant factor, almost one-fifth of the ranking.

In 2025, the University of Arkansas-Fayetteville Law School had an 80% passage rate—92 out of 115, and the school was ranked #100. By comparison, the University of Missouri Law School (remember, Missouri has a 260 cut score) had a roughly 86% pass rate—106 out of 123, and the school was ranked #59.

How much higher would the University of Arkansas Law School’s rank be if it had a similar passage rate as Missouri? You can actually come up with a pretty reasonable estimate with a little bit of math assuming a normal distribution of scores and assuming the standard deviation in Arkansas scores is in line with the national standard deviation. I’m skipping the math in this post because this Substack is not In re Arkansas Math.

The answer comes out that the passage rate would be about 8-9% higher.3 How far would this go in the rankings? That is not as easy to calculate (you could, in principle, but it’s more work than I can devote to it right now), but it’s obviously significant.

Having a better ranking would create a self-reinforcing positive feedback. Higher ranking would lead to more students with higher test scores and undergraduate GPAs applying and matriculating. Those students would be more likely to pass the bar. That would leave to even higher ranking.

The Political Realism Argument

The above discussion suggests that Arkansas would be better off with a lower cut score. So why does the legislature not simply lower the cut score?

Presumably, the answer is that no one wants to be seen lowering standards, especially not for lawyers. But I think some political realism is called for here. No one outside of lawyers actually knows or cares what the cut score is. I candidly had no idea what the cut score was when I took the bar. If you polled Arkansans, “Would you object to having the same bar passage requirements as Wisconsin, Missouri, Minnesota, and Oklahoma?”, I think the answer would be no.

So, this is one instance where there is almost certainly no political cost to doing the smart policy move. My brilliant Hill staffer recommendation: just do the smart policy move.

Subscribe now

Turns out there’s a lot of stuff like that. And the unifying theme is: don’t name things after living people!

Subscribe now

(And do yourself a favor and read my previous installments to learn about Conway, Crittenden, and Baxter counties—I would put them on this list, but it would just be repeating stuff from that installment.)

5. Craighead County - Named as a Taunt

In 1858, Arkansas Senator William Jones campaigned on creating a new county to save residents a long trip to the county seat when they had to conduct business. Sounds like a boringly logical reason. You might be wondering, why it’s not called Jones County.

Another senator, Thomas Craighead, didn’t like this idea because the new county would take a big chunk of fertile farmland (and attendant tax revenue) out of his county. Everyone’s being so reasonable in this story so far that you can be forgiven for expecting someone to be stabbed to death with a Bowie knife on the Senate floor!

But Jones handled this disagreement like an adult. He waited for a day when Craighead was not in attendance, then called his bill up for passage. Oh and he named the county after Craighead as a taunt. (Or it might have been an attempt to butter Craighead up. But I’d much rather believe it was a taunt.)

When Craighead showed up and saw what happened, he reportedly sponsored a resolution to name the county seat “Jonesborough” after Jones. This isn’t really the best revenge though—Jones won the legislative fight and his name is now forever tied to the county he created. (And, of course, so is Craighead’s name—TOTAL LOSER).

4. Randolph County - Named for the James Dean of 19th Century American Nepo Babies

The old joke about nepo babies is that they are born on third base and think they hit a triple. Meriwether Lewis Randolph was a decent way past third base on his way to getting a county named after him.

His father came from a prominent Virginia family and was the governor of Virginia from 1819 to 1822. His grandfather was Thomas Jefferson. Ol’ Meriwether was born while his family was living at Monticello. You may have correctly guessed that he was named for Meriwether Lewis, Jefferson’s private secretary and the “Lewis” part of “Lewis and Clark”.

Meriwether was a law school student at the University of Virginia, but never practiced law. His big ticket to the highest echelons of American power was his heroic service as one of the floor managers for President Jackson’s second inaugural ball in 1833.

He worked his way up to become Jackson’s private secretary. That connection led to his appointment as secretary for the governor of Arkansas Territory in 1835. When Arkansas became a state in 1836, he used his connections with Jackson to become a land speculator and grew wealthy (well, wealthier).

This early Arkansas connection combined with his Jackson connection was sufficient to get him chosen for the honor of having a county named after him. But he objected that he’d only lived in Arkansas for a short time and had done nothing to deserve it. (Very honest of him!) He suggested to the territorial legislature that they could name the county Randolph after his relative John Randolph of Roanoke, which the legislature did. There’s no evidence that Randolph had ever been to Arkansas.

(And yes, if you want to nitpick, I guess you could say the other Randolph is technically the namesake, but let’s be honest—our boy Meriwether is the reason it is called “Randolph”.)

The James Dean part comes from the fact that Meriwether died in 1837, just 27 years old. He was in the middle of building a vast estate for himself, but caught malaria on a trip and died. His family returned to Tennessee. So, this dude was ultimately in Arkansas for two years, didn’t really do anything there, and got a county named after him. Absolute legend!

3. Clay County - Cancel Culture Conundrum!

In 1873, the legislature created “Clayton” County, named after state senator John M. Clayton, brother of Powell Clayton, the first Republican Reconstruction governor. The anti-slavery, pro-Union Republicans had adopted the GOAT Arkansas Constitution of 1868. After the Brooks-Baxter War resulted in the ascendancy of the 1874 (and current) Arkansas Constitution, the legislature wanted to get rid of the name of this Republican governor.

They just cut the last two letters off. Voila! Now they had a county named after Henry Clay, the three-time presidential election loser who was a moderate on slavery and who had been the driving force behind the compromise of 1850.

Sometimes I wonder at the political acumen of 19th-century legislators. There seriously wasn’t a single guy who was willing to buy the right to name a county? But there you go, Clay County is not just a place with ardor for a particular kind of dirt.

2. Faulkner County - Named for a Guy Who Was a Self-Reported Good Fiddler

Sandy Faulkner (whose first name also might have been “Sanford” or “Sanderson”, which gives you an idea of how poorly documented this story actually is) managed two cotton plantations that belonged to his family. He was also president of a branch of the Real Estate Bank of Arkansas that collapsed in the economic downturn of 1837, which basically destroyed Faulkner economically.

But during his financial travails, Faulkner went on a canvassing tour of the state with some politicians. He came to the hut of a settler, who for some weird reason didn’t want to just chat it up with a failed banker/plantation owner and four politicians.

But Faulkner (according to Faulkner) won this surly guy over by finishing a fiddle tune that the settler had been singing. (Sidenote: this was the basis for the famous “Arkansas Traveler” story.) But this was all based on Faulkner telling the story a million times, so I think there remains some scientific uncertainty over just how fire Faulkner’s fiddling was.

Faulkner would go on to have an undistinguished military career serving the Confederacy as commander and military storekeeper of the Little Rock arsenal. He fled to Texas after Little Rock was captured, returning to Arkansas after the Civil War. I think it’s fair to say that the fiddling story is really the only thing that could have led to him getting the county name.

I don’t know exactly how this translates to modern politics, but if some guy ever comes to visit you with four prominent Arkansas politicians, be sure to test him on the harmonica or something.

1. Garland County - Named for a B-Squad Confederate Who Really Should Have Shown Up in the Alanis Morissette Song “Ironic”

Augustus Garland was a lawyer with a problem. In 1860, he had been admitted to the bar of the U.S. Supreme Court. He was one of the most prominent attorneys in Arkansas. But he made a teensy mistake—after some sitting on the fence, he became a supporter of secession. He was appointed to the Confederate Senate during the Civil War in late 1864 (bad timing, Gus, the Confederacy was basically doneso by that point!)

The Supreme Court changed its rules after the war to prevent former members of the Confederate government and military from arguing cases before the Court. Garland litigated against the change and won a 5-4 decision. He later became attorney general under Grover Cleveland in 1885.

His tenure as attorney general was ruined by, well, corruption. He was the attorney for and owned stock in a telephone company while attorney general. His office initiated a lawsuit against Bell telephones trying to invalidate Bell’s patent (admittedly, while Garland was away on vacation, but it’s quite the coincidence!)

But his arguable biggest claim to fame came when he was arguing a case at the Supreme Court in 1899. He had been feeling a little under the weather. Fifteen minutes into his session, he stated: “This, your honors, is our contention…” He gasped and fell sideways onto the floor.

He was carried to the chief clerk’s office where he died, the victim of an abrupt stroke. He is the only person to have died at the Supreme Court (though not the current Supreme Court building—in Garland’s day, it was still in the Old Senate Chamber inside the Capitol.)

Oh and he lost the case. Unanimously.

Note for current law students: if Garland couldn’t get a sympathy vote for his case by actually dying, don’t pretend to be sick (or die) to try to sway the judges! But, look on the bright side, you probably could not underperform Garland’s “unanimous loss plus dying in the clerk’s office” performance.

Quick end-note disclaimer about Garland: in addition to being a senator in the Confederacy, he constantly lobbied against Reconstruction after the war, so if I seem somewhat gleeful about how things ended up for him, well, it’s because he was a bad guy even by former Confederate standards. (Compare Garland to James Longstreet, who supported black suffrage after the war at the cost of his reputation in the South. Also, if you’re into military history, Longstreet was absolutely right about Gettysburg, but became a scapegoat for Lee being too impulsive.)

Subscribe now

(Note: I did not include the images I’m including here in my law review article because, well, I just didn’t have the courage. That’s more of a post-tenure move.)

Subscribe now

Free Exercise: A New Front in the Ten Commandments Legal War

The Establishment and Free Exercise clauses constitute the religion clauses of the First Amendment. Broadly, the Establishment Clause prohibits government endorsement of a particular religion, and the Free Exercise Clause prohibits government from excessively burdening religious practice.1

Previously, almost all analysis for and against public school Ten Commandments displays focused on the Establishment Clause. The 1980 case Stone v. Graham applied the Lemon test (from Lemon v. Kurtzman) and held that public school Ten Commandments displays violated the Establishment Clause.

The conservative majority on the Supreme Court in 2022 explicitly disavowed the Lemon test in Kennedy v. Bremerton School District. The whole point of the current Ten Commandments statutes is to trigger court cases that will eventually reach the Supreme Court, which will duly overturn Stone v. Graham.

But at the same time the Court was curtailing the Establishment Clause, it was bulking up the Free Exercise Clause. In Mahmoud v. Taylor in 2025, the Court held that public schools effectively must offer an opt out of pro-LGBTQ reading material.2 That case reinvigorated a principle that had been sitting mostly dormant since Wisconsin v. Yoder in 1972. The state may not interpose itself between parents and the religious formation of their children—even through subtle pressure, even without compelling any specific act.

Originalism’s Struggle to Figure Out the Free Exercise Clause

To figure out if the Free Exercise Clause has anything to say about public school Ten Commandments displays, we need to figure out what the Free Exercise Clause means in this context. And since the majority of the Supreme Court is originalist, I focused on discerning an originalist argument about the meaning of the Free Exercise Clause.3

The problem with originalist free exercise arguments is that we have virtually no direct evidence of what the Founders, the First Congress, or the public at large in 1791 thought “the free exercise thereof” meant. The scholarly debate has spent decades locked in on one narrow question: whether the Free Exercise Clause requires religious exemptions from neutral, generally applicable laws. So, for example, many originalist scholars of the religion clauses wrote in the 1990s about things like whether prisons could ban the use of peyote in Native American religious rituals. But those theories are of limited use in deciding whether public school Ten Commandments displays unconstitutionally pressure the religious formation of children.

We know even less about what the Founders thought about religion in public schools because there were essentially no public schools at the time, and the First Amendment only applied to the federal government at that point anyway.

Modern litigants and courts often resort to taking isolated statements of the Founders about religion broadly and inferring a complete worldview on its basis. For example, Judge Ho of the Fifth Circuit in his concurrence in upholding the Louisiana Ten Commandments display argued that the Northwest Ordinance (which contained a general endorsement of religious education) showed that “religion was a cornerstone of American education from the beginning.” But the Northwest Ordinance tells us nothing of how the Founders would think about forced religious education in public schools, which is categorically different from private religious education chosen by the parents of the child in question.

My article takes a different methodological path. If we don’t know what people had in mind when they read “free exercise thereof”, we should look the conceptual background against which that language was written. That approach provides the best remaining avenue for understanding what an educated American in 1791 would have understood religious liberty to mean.

For this method to be an appropriate choice, and the article proposes four conditions: the thinkers must actually converge on a principle; the Founders must have demonstrably engaged with and endorsed them; there must be no record of Founding-era disagreement on the point; and the principle must be specific enough to address the actual question, rather than dissolving into “the Founders thought religion was good.”

The three thinkers who most clearly shaped the religious liberty discourse of the era were John Locke, Pierre Bayle, and Voltaire.

My Free Exercise Three Amigos: Locke, Bayle, and Voltaire

These guys were not theorizing in a vacuum. They experienced first hand the most sustained state campaign to remake children’s religious beliefs in European history.

Louis XIV spent the 1680s trying to extinguish French Protestantism (embodied in the Huguenots). He tried dragonnades—billeting soldiers in Huguenot homes to bully the occupants until the family signed an abjuration of Protestantism. That plan had an obvious flaw: adults would sign the paper and go right on being Protestant once the soldiers left. Coercing adults produced hypocrites, not converts.

So the French government figured out a workaround. Don’t try to convince parents—put the pressure on children, who are famously far more malleable in their beliefs. Huguenot children were ordered baptized Catholic and sent to Catholic schools. “[O]bstinate” parents could have their children removed and raised in Catholic households or convents. Roughly 175,000 Huguenots fled France illegally, among them the ancestors of John Jay (the least famous of the authors of the Federalist Papers).

Voltaire was a contemporary writer for the Founders. As a public intellectual in France, he led crusades against persecution of Huguenot families, most prominently the Calas and Sirven families. As part of those efforts, Voltaire drafted his Treatise on Toleration, one of the Enlightenment’s most famous works on religious liberty.

Locke and Bayle reacted strongly against trying to influence children in this way, and converged on three relevant principles, which Voltaire would later endorse in a less organized, philosophical way. First, civil government has no jurisdiction over inward belief. Second, parents, not the state, must direct children’s religious formation. Third, state-sponsored denominational instruction pressures conscience even when no specific conduct is compelled.

This wasn’t a fringe tolerationist position. Even Samuel von Pufendorf—the Enlightenment’s most enthusiastic defender of sovereign authority over external religious practice—conceded that the religious formation of children lay outside what families ever surrendered to the state.

The Founders, for their part, deeply engaged with all of these thinkers. Jefferson called Locke one of the three greatest men who ever lived; Madison put Locke, Bayle, and Voltaire on his book-buying list for Congress; Franklin credited Voltaire with disarming the bigotry of Europe; and John Adams filled his copy of Voltaire’s Treatise on Toleration with approving marginalia. Crucially, when the Founders disagreed with these philosophers, they said so, clearly and in writing. On civil authority over children’s religious formation, the record contains no such dissent.

Deriving a New Free Exercise Test

Translated into doctrine, the principle is the mirror image of a rule the Court already enforces. The Court has previously held that government cannot act with the object of suppressing religious exercise. My article’s proposed test holds that government also cannot act with the object of building it. When objective evidence—text, design, operation, official justifications, denominational specificity, audience—shows that the state is prescribing, altering, or pressuring religious belief as such, strict scrutiny applies. The inquiry intensifies where children are the audience, because their formation belongs first to their parents.

This test doesn’t require a secular purpose for every government action as Lemon did. Things like “In God We Trust” on money survives the test because nobody’s inward beliefs are reasonably targeted by a coin. Ceremonial deism before adult audiences survives. Christmas as a federal holiday survives. Religious accommodations survive. Actually teaching about religion in a history, literature, or comparative religion class survives. The test is about religious formation, not religious reference.

Applying the Test to Ten Commandments Displays

Are the Ten Commandments displays aimed at the inward beliefs of children? Several features stand out in inferring such a motive.

First, the target audience is specifically children, more vulnerable to state coercion contrary to the religious teachings of their family. Second, the sponsors of these statutes have stated plainly that they intend for children’s religious views to be influenced by the displays. The Louisiana sponsor announced on the floor that she wanted "our children looking and seeing what God's law is," and the Texas attorney general (and now U.S. Senate candidate) said he wanted "the Word of God opened" in classrooms.

Bayle actually provided one of the best observations about how laws like this pressure religious belief. He criticized “invitations to reflection” that never end. Over time, such repeated invitations to reflection amounted to coercion.

Here, the Ten Commandments displays pose a similar constant pressure via invitation for reflection with no opt out. As I described in an earlier post, there is no way for a teacher to answer a student’s question about what the Ten Commandments poster is doing in their classroom without pressuring the religious beliefs of them and their families.

If they say it’s the word of God and how we should live, that would be religious instruction, exactly what the Fifth Circuit said Texas did not authorize. If the teacher says it’s a historically important document that may or may not be from God, then it is pressuring students who believe the Ten Commandments are of divine origin. If the answer is “ask your parents,” the state has engineered a religious confrontation at home, with the state’s chosen version of the Ten Commandments as the baseline.

If the States Wanted History, They Could Have Displayed History

And then there’s the text of the displays. These statutes don’t display the Ten Commandments as Moses, Maimonides, Luther, or any Founder knew them. They display a mid-century Protestant-coded abridgment popularized as movie publicity by Cecil B. DeMille. Modern legislatures use it explicitly because a tablet on Texas’s capital grounds displaying this version survived Supreme Court scrutiny in Van Orden v. Perry in 2005.

Arkansas’s posters jettison roughly ninety percent of the Second Commandment. That edit is just the latest manifestation in a centuries-old dispute between (most) Protestant sects and Catholics about graven images. Catholics, Lutherans, Jews, and Muslims all order or understand the text differently. The traditions are mutually exclusive, which is why no committee has ever produced a “nondenominational” Decalogue and why the 1950s interfaith campaign didn’t either.

The stock defense is something like: the Founders loved religion, therefore they’d love these posters. From an originalist perspective, that commits exactly the level-of-generality sin that originalists condemn everywhere else. Nobody would accept “the Founders valued self-defense” as resolving a modern Second Amendment case about, say, whether stun guns are constitutionally protected. Yet a John Adams line about religion and morality gets stretched into blanket approval of a specific 1950s text Adams never saw, mandated by a government entity that didn’t exist, in schools the First Amendment didn’t then reach.

There are lots more fun specifics in the article, but that’s the argument. It remains to be seen what the Supreme Court will ultimately say about the Free Exercise Argument.

Subscribe now

The Arkansas Constitution currently has a provision broadly similar to the U.S. Constitution’s Second Amendment. The current version says: “The citizens of this State shall have the right to keep and bear arms, for their common defense.”

Ballot Issue 2 would do three things:

1. broaden the right to cover hunting, “recreational use”, and “any other lawful purpose” in addition to “common defense”;

2. create new rights to “the possession and use of” ammunition, “firearm accessories”, and “firearm components”; and

3. insert a declaration that “[t]he right to keep and bear arms [] is a natural, fundamental, and individual right that shall not be infringed.”

There are a lot of interesting aspects of the phrasing here. The most curious part of all is the last one, because it was originally written in a way to very clearly do one thing, and then the sponsor of the amendment arguably weakened it at the very last second. It all adds up to an intriguing constitutional law/politics whydunit, which we’ll get into in this week’s post!

Subscribe now

Theory 1: This is a Backstop For When Supreme Court Membership Changes

Let’s start with the declaration part (the third change I listed above). If you’re just glancing through, having a constitutional provision that says a certain right is “natural, fundamental, and individual” might just sound like rhetoric. And, well, the natural part is pretty much just rhetoric.1

The “individual” part does not matter presently because the U.S. Supreme Court held in District of Columbia v. Heller in 2008 that the federal Second Amendment creates an individual right to bear arms. But that was a controversial 5-4 decision, so it’s not completely impossible to imagine the Supreme Court overturning it someday. If they did, then the Arkansas Constitution guaranteeing an “individual” right to bear arms would become relevant.

Even in the blue Supreme Court Scenario, it is worth noting that a state constitutional amendment would still not be the only thing guaranteeing the right of an Arkansan to bear arms. The constitutional amendment would only actually matter if Arkansas (or local governments) started churning out gun control legislation in the wake of Heller being overturned.

It is not at all inconceivable that, say, Fayetteville could do such a thing. But obviously the overwhelmingly Republican state legislature would not. And the state legislature could stop Fayetteville from adopting gun control legislation with legislation of its own.

So, the constitutional amendment would really be guarding against the possibility of a blue tidal wave so massive that it not only flips the U.S. Supreme Court, but it also flips the Arkansas state legislature and governor’s office.

If this sounds silly, recall that Democrats in blue states have put abortion protection in their state constitutions, and the exact same logic applies there—they’d only be relevant if there was a red wave that dislodged the state legislature. Purple-ish states like Michigan have also done that, which makes sense if you think of it as hedging against a narrowly red legislature.

In both the Arkansas case and the blue states with abortion cases, the state leaders presumably think something like: “Even if it’s extremely unlikely that this will ever be useful, it’s pretty easy to do, and it will be popular in-state, so let’s just do it.”

So, “natural” doesn’t really do anything, “individual” is a fallback in case of an implausibly gigantic blue wave—what about “fundamental”? Here’s where things get very strange.

The Strict Scrutiny Mystery

If you remember your constitutional law, the word “fundamental” is something of a trigger. Courts apply “strict scrutiny” to laws that implicate rights that are “fundamental”. Strict scrutiny, in turn, is very hard to satisfy—not literally impossible, but it doesn’t happen often. It requires that the government action be narrowly tailored to serve a compelling government interest.

There has been a wave of red states adding gun rights amendments to their state constitutions. At first, they just did what Ballot Issue 2 does—declare gun rights to be fundamental so that when their courts review gun laws, they will apply strict scrutiny. But state supreme courts (most notably, Wisconsin) started holding that just declaring something “fundamental” doesn’t immediately trigger strict scrutiny. The court has to undertake its own analysis to decide whether to apply strict scrutiny.

So, red states started actually requiring courts to apply strict scrutiny (see, e.g., Alabama). They literally put provisions in their state constitutions requiring it explicitly.

And so did Arkansas Ballot Issue 2, when it was first proposed.2 But the lead sponsor amended his bill to remove that part. That is the single strangest part of this story. If the point of Ballot Issue 2 is to be a fallback if SCOTUS goes blue, why weaken it by taking out strict scrutiny?3 Strict scrutiny was one of the primary reasons the NRA has been advocating for state constitutional amendments in the first place!

This creates a problem for the constitutional fallback theory. If the primary concern is how courts would adjudicate gun rights cases, why take out the strict scrutiny part?

Deepening the Mystery: The New Categories Seem to Lean Toward Everything Being Protected…

Another layer of the purpose mystery is the weird mix of permissive and restrictive provisions. The “recreational use” protection combined with protecting “without limitation the possession and use of . . . firearm accessories [and] firearm components” seems to open up quite a can of worms.

Here’s an example: a machine gun is a portable firearm that one can lawfully shoot at a firing range. It doesn’t seem like there’s any reason it would not be constitutionally protected under Ballot Issue 2. At the federal level, Justice Scalia himself in Heller said that an M-16 (not even a machine gun) could be legally banned because it was not commonly used for self-defense.

And, of course, even if you could constitutionally ban machine guns, the amendment is crystal clear that you can’t ban firearm accessories or components, so people could just make their own machine gun through simple modifications to semi-automatic rifles.

But that’s just scratching the surface. The Arkansas Criminal Code currently defines firearm as "any device designed, made, or adapted to expel a projectile by the action of an explosive[.]”

Here’s just an off-the-top-of-my-head list of other things that would qualify because you could use them for fun recreationally:

• “Ghost” gun kits (regulation of which was recently upheld at the U.S. Supreme Court);

• Drum magazines and belt-fed firing systems;

• Armor-piercing ammunition;

• Tracer or incendiary ammunition;

• Short-barreled rifles and shotguns (famously useful for robberies);

• Portable grenade launchers and mortars (they are literally portable firearms, and their shells are ammunition); and

• Recoilless rifles or rocket launchers.

…But the “Lawful Purpose” Language Seems to Let the Legislature Decide Everything

But here’s the thing: those categories still might not be protected. Why? Because the proposed amendment says that the uses protected include “other lawful purpose[s]”, implying that the recreational or common defense use still has to be lawful. So, if the state legislature passes a law saying it is illegal to use rocket launchers, including recreationally, then maybe rocket launchers are not covered by the statute. But does that just make the coverage of this amendment subject to statutory definition for everything, including common defense? Does this weirdly allow the legislature to restrict the Arkansas right to bear arms more than the current provision?

So where does this leave the backup for the Second Amendment theory? If the amendment is meant to be expansive, why does it have the “lawful purpose” part? If it’s meant to be reasonable, why have the “without limitation” + “recreational use” hack that theoretically makes everything protected?

Theory 2: This is Just About Having Something Pro-Gun Rights on the Ballot

I can hear the more political people I used to work with on the Hill arguing at this point: maybe the supporters just wanted to be seen by their constituents being pro-gun rights, and they knew the details didn’t really matter because the limiting factor is federal constitutional law, not state constitutional law, and the Arkansas state legislature is never going to pass a law that tests the limits on gun control.

I get that things like that happen. But if that is the reason, why did the amendment sponsor get rid of the strict scrutiny part, and why did everyone else go along with it? If all you wanted to do was court the NRA, you would let them draft the provision (or at least take their advice on it), and they would put in a strict scrutiny provision, as they have supported in other states.

Theory 3: This is a Trojan Horse Amendment (No, I Haven’t Seen the new Odyssey movie)

A Trojan Horse bill/amendment refers to the idea that the flashy part of the bill is pretty popular, but there are less discussed provisions that are where the real action is.

The theory here might be that the support for gun rights in Arkansas is broader than it is deep. So, you have an amendment with a lot of nods toward popular ideas (hunting, natural rights, “without limitation”), but the real purpose is to push the boundaries in the new categories in the way I described above.

Opinion polls suggest there is 70-84 percent support for regulating “ghost gun” kits. So, if you just had a constitutional amendment ballot issue that explicitly protected ghost gun kits, machine guns, and recoilless rifles, then even in Arkansas it might not pass. But if you have an amendment that is mostly about broadly protecting the right to bear arms, but has some subtle provisions that could result in the constitutional protection of those things, it might have a better chance at passing.

Again, I could see that happening. It is basically legislative strategy 101—a tactic used by everybody since time immemorial.

But that still leaves the question: if the point is quietly expanding gun rights, why take out strict scrutiny, which is very meaningful but obscure to the vast majority of voters? Why include the lawful purpose caveat?

Theory 4: Accidental Chaos

There’s a lot of stuff in this ballot issue that is vaguely somewhere in the pro-gun rights lexicon or famous pro-gun rights cases. This could just be a bunch of different cooks in the kitchen not really working together to make the whole thing make sense.

For example, the lawful purpose thing was stressed by Justice Scalia in his Heller majority opinion. But it was limited there by Scalia emphasizing the centrality of “self defense” as the key historical lawful purpose.4 If you have “lawful purpose” in reference to current law, then you just subsume the right to bear arms into whatever is currently lawful.

The firearm accessories and components parts don’t really make sense unless you want to allow things like bump stocks or ghost gun kits. So maybe there was some vocal group that wanted that protection, but they weren’t really paying attention to the “lawful purpose” and strict scrutiny parts, which are more aimed at law nerds.

Getting rid of strict scrutiny might have been a sort of misguided attempt to pay homage to originalism, which generally eschews levels of scrutiny in constitutional law.5 But originalists don’t like levels of scrutiny in federal constitutional law because they are judicially created. Faced with an amendment that specifically directs the application of strict scrutiny, it would be entirely appropriate for an originalist to apply strict scrutiny.

I hate to end this post with the conclusion “I don’t know what the purpose is”, but that’s ultimately where I’m left on this one. Of course, it could be a combination of all the theories I’ve outlined. It’s a shame we will probably never really get a clear explanation of what the drafters had in mind.

At the very least, this ballot measure serves as a reminder of how hard (impossible?) it can be to discern original public meaning or intent, even when it’s about a constitutional provision from 2026 instead of 1791.

Subscribe now

Arkansas Ballot Issue 3 is flying under the radar. It’s even bipartisan! Not a single Democratic representative or senator voted against it when the Arkansas legislature considered it. Governor Sanders has endorsed it. It carried majority Republican support in both chambers of the legislature.

And it is an absolutely terrible idea. It’s so bad that I was genuinely shocked it has gotten this far.

I’ll get more into the logic below, but here’s a brief summary: it would amend the Arkansas Constitution to allow local governments to create economic development districts with broad powers to tax, borrow, issue bonds, and subsidize private businesses.1 It leaves almost all the important details to future legislation, but previously introduced implementing legislation, SB 647, gives us an idea of what the broader program would look like if Issue 3 passes.

I’m going to get into Ballot Issue 3 in detail in this week’s post. And yes, I have already tipped my hand a bit, but I promise I will try to give the positive case as well.

Subscribe now

A Brief Detour into the History of Bad Ideas

I worry that “economic development district” sounds so bland and boring that we might lose sight of what is really going on, so I want to lay out a historical analogy to something more colorful.

In the 17th and 18th centuries in France, when the king needed money, he engaged in a practice called “tax farming.”2 The king sold the right to collect taxes from a given area to some wealthy noble/merchant/whoever. The king got money right away, and the buyer got the taxes.

This system worked in exactly one way: it got the king money fast. It was basically what kings do instead of payday loans. But it was absolutely crippling in the long run.

First of all, the king no longer had much to gain from development. The tax farmers would reap the benefits of growth in the tax base, so the king had no reason to worry about any sort of long-run improvements. Education, infrastructure—none of that stuff helped the king, so he had less incentive to invest in them.

Second, tax farming destroyed the kingdom’s fiscal situation in the long run. A little-appreciated fact about the French Revolution: France’s system of government under the Bourbons was on shaky ground regardless of when the French Revolution happened. A state’s fiscal needs tend to grow as it becomes more developed. The only hope of keeping up with them is to capture some of the value of that development. But the French kings left themselves with a system that prevented that.

And, of course, if you separate tax collection from the ordinary political system, the tax collectors do not act in the public benefit. Adam Smith devoted a memorably scornful passage of The Wealth of Nations to tax farmers, observing that a private collector will squeeze taxpayers in ways no sovereign would dare. Every sophisticated state eventually learned to hate this arrangement.

On November 3, Arkansas will vote on whether to give it one more try.

Ballot Issue 3: The Basics

Ballot Issue 3 is all about enabling something called “tax increment financing”, which is currently legal but practically difficult for Arkansas local governments to do because of a constitutional hang-up (we’ll get to that later).

Ballot Issue 3 would enable tax increment financing (TIF) by allowing local governments to create “economic development districts” that would be empowered to carry out “projects”. Those districts would get “increment” tax revenue (i.e., tax revenue in excess of the current amount). The districts could give that money to businesses through subsidies or tax breaks in the district, pay for other projects — basically, the district can do most of the things a local government can do.

Arkansas local governments already can borrow to build infrastructure and repay the debt from taxes, including taxes generated as development occurs. Amendment 78 already permits them to pledge property-tax increments for that purpose. Issue 3’s distinctive contribution is not infrastructure financing itself, but the ability to capture a broader range of tax revenue, bypass existing debt and election constraints, and use public funds for a much broader range of public and private economic-development expenditures.

Who controls the districts and actually gest to decide how much money is raised and spent? Ballot Issue 3 does not answer that. Here, we have to look at the proposed enabling legislation to get an idea of the plan. Under that plan, business owners can propose a district, and must be represented on the district board. There is no limit to how much of the board can be comprised of the business owners.

The amendment does not define what an economic development “project” is. It does not require a finding of blight, or distress, or underdevelopment, or anything at all about the territory being carved out. It does not cap the districts’ duration, guarantee local governments a say, or promise voters at large any role in approving what the “district” does. All of that—everything that matters—is delegated to the legislature, to be filled in by ordinary statute and revised in any future session without ever going back to the people.

21st Century Tax Farming

You can see how this system is similar to tax farming. Long term benefit leaves publicly accountable hands. Money decisions and taxes are given to a group of people whose interests may diverge from the community as a whole.

Because the money at stake is an “increment” (i.e., the growth of tax revenue), proponents can say that schools will still get their baseline funding. But if circumstances change (say, the development actually works and revitalizes the area, or if inflation makes running the schools more expensive), then the schools won’t have what they need.

The game here is simple. Local governments want to entice businesses to come to the area, but local governments aren’t awash in cash. Property taxes are a big source of revenue for local governments, which spend a big chunk of property tax revenue on schooling. So, if local governments can just redirect that money to businesses, then the problem is solved.

About 25 years ago, Arkansas tried to allow TIF. But the Arkansas Supreme Court held that the Arkansas Constitution reserves a big chunk of property tax revenue for schooling. That left little money left over for TIF.

And that’s why we have this proposed constitutional amendment on the ballot this year—to effectively overturn the Arkansas Supreme Court decision and get local governments back in the TIF game with even broader authority. Schools keep a baseline amount of money, and the growth in tax revenue goes to the businesses.

All of the same bad dynamics of tax farming play out here. The municipal government loses its stake in growth—who cares if education produces more economic value in the long run, the local government doesn’t see the increased revenue? And if the businesses bring economic growth, that will inevitably increase the need for tax revenue (for more schools, more roads, bigger police and fire departments, etc.), but the revenue is already going to the businesses themselves. So, you’ll get some short-term growth at the expense of the long run.

TIF is an Economic Loser

Hey, maybe I’m just being a traditionalist stick in the mud. After all, other states do this, right? And it brings in business, right?

First, Arkansas’s TIF is far broader than normal. For example, Texas’s TIF program in theory applies only to blighted areas. There is no such limitation for Arkansas. We are entering uncharted territory.

Second, academics disagree about a lot of real-world policies, but there is little real disagreement on TIF: it does not work. At best, it can increase property values in the area where it is implemented, but for the broader economy, research suggests it’s actually worse than nothing—TIF actually hurts economic growth.

There are two reasons why: (1) TIF mostly sweetens the deal for businesses that were going to come anyway (but you lose the tax revenue that could have increased future growth); and (2) TIF tends to draw businesses to one area instead of another, leading to inefficient overall growth.

On that second point, imagine if you entice a grocery store to set up shop in Springdale instead of Fayetteville. Demand for groceries is inelastic—people basically always need the same amount. So, you didn’t create any value by luring the store to Springdale. But maybe another business (say, a bicycle shop) really wanted the location where the grocery store ended up (maybe it’s right on the Razorback Greenway). Suddenly, luring that grocery store imposed an opportunity cost in addition to the tax revenue giveaway.

By the way, I use grocery stores as an example because one extra bad feature of the proposed implementing legislation is that sales taxes can also be abated for businesses.3 That tends to draw retail stores. In the economics literature, retail stores are exactly the ones serving inelastic demand—it makes no sense to try to attract them because there isn’t slack demand for them to take up! A new grocery store is just going to redistribute the existing demand, not generate growth.

So What is the Positive Case for TIF?

Good news: TIF does increase property values in the districts where it is enacted. Of course it does—imagine if buying a home in a certain area meant you could collect your neighbors’ property taxes. It would be economically rational to pay more for your home, and the property value would go up. But it wouldn’t be a smart or fair program! The whole problem with TIF is that these benefits are concentrated and the far larger costs are diffuse.

Another argument for TIF goes something like this. Say you’re a local government and you want to invest in infrastructure to attract developers/businesses. But the businesses don’t want to come unless there’s infrastructure. TIF lets you attract the businesses first. But note that there is no actual requirement in either Issue 3 or in the enabling legislation to task these districts with some infrastructure need. So the money the district collects can just go right back out as subsidies rather than infrastructure.

TIF is also good for governors, legislators, and local government officials who want to brag about new businesses. French kings wanted short-term money; modern politicians want short-term gain for their next election.

Voters will assume that businesses are coming because those political leaders are just so dang good at their jobs. And businesses will certainly come if you pay them to—hand over a big chunk of the money that was supposed to educate your children and grandchildren, and you too may get a Best Buy in your town!

But I can’t stress this enough: the Arkansas TIF plan is simultaneously (1) sacrificing the value of having new businesses by getting rid of the tax revenue, and (2) giving money to businesses that were going to come anyway.

But Professor Thorlin, I’m Still Not Convinced, Could You Give Me Yet Another Reason Why TIF is Bad?

If you had the thought: “giving control over large amounts of taxpayer money to a non-elected group of people to parcel out to local businesses whenever they deem fit sounds like a recipe for corruption”, you are absolutely correct! That is why, unsurprisingly, studies find that TIF and similar programs are associated with increased corruption among local officials.4 Why wouldn’t a shady businessman promise a lucrative job to a local official when they leave office in exchange for years, possibly decades, of other people’s property tax revenue?

Again, this is the sort of thing people understood in the 1700s. The slogan “no taxation without representation” is pretty suggestive here—you want a politically accountable process to connect the collection of tax revenue and the decision of how to spend the public’s money.

So, in conclusion, I do not think you should vote in favor of Arkansas Ballot Issue 3. I wrote a few weeks ago about how Arkansas could develop its economy. Some of those programs require government intervention. Ballot Issue 3 is not bad because it is government intervention—it is bad because it enables funneling our money away from what will develop Arkansas’s economy in the long run (education) and into the pockets of politically-connected businesses.

I also want to give a brief shout out to the only ones who voted no when the issue went through the legislature: principled fiscal conservative Republicans who objected to handing over the public’s money to private businesses.

There were six Senate Republicans and 21 House Republicans who voted no: Senators Joshua Bryant, Jim Dotson, Bryan King, John Payton, Clint Penzo, Gary Stubblefield; Representatives Wayne Long, Mindy McAlindon, David Ray, Ryan Rose, John Carr and Robin Lundstrum (there are other representatives, but I don’t have the full list.)

Subscribe now

Texas Swinging for the Fences

The Supreme Court has weakened the Establishment Clause in a series of cases since the conservative majority took the Court in October 2020. Most notably, Kennedy v. Bremerton School District in 2022 formally invalidated the longstanding Lemon test that required, among other things, that government actions have a secular purpose.

The conservative groups backing statutes that require Ten Commandments displays in public schools in Arkansas, Texas, Oklahoma, Louisiana, and Alabama are trying to get a case to the Supreme Court regarding the constitutionality of school Ten Commandment displays. Once there, they expect the conservative majority to strike down Stone v. Graham, the 1980 case invalidating such displays and relying, in part, on the Lemon test.

This Texas statute is the next logical step beyond Ten Commandments displays in using public schools to indoctrinate students in evangelical Protestant Christianity. If the Supreme Court strikes down Stone, the justices might be willing to strike down Abington School District v. Schempp. That 1963 (pre-Lemon) case held that the Establishment Clause forbids legally sanctioned or mandated Bible readings.

The fact that Schempp is from before the time of Lemon means that it is not as immediately vulnerable as Stone. Kennedy did not have a holding that directly implicated Schempp in the same way it implicated Stone, but as a matter of vibes, Texas and other conservative states might reasonably think that the conservative Supreme Court is willing to go farther.

That is not a great bet when you look more deeply into the logic of Schempp and the specific holding of Kennedy. In Kennedy, Gorsuch explicitly said the situation was not like Schempp because the government itself was not directing the religious exercise. The majority also cited another case regarding a graduation benediction (Lee v. Weisman) approvingly.

Kennedy also endorsed a history-and-tradition inquiry for the Establishment Clause, and the history of compelled Bible reading in school is not at all favorable to Texas in this case. The Bible was mostly removed from public schools because of violent confrontations, even riots, over the selection of Protestant material over Catholic.

We will probably have a better idea of how far this Court is willing to go once a Ten Commandments case reaches them. That is far enough away (probably 1-3 years) that the composition of the court could change, which would make prognostication even more difficult.1

An Opt-Out Won’t Save This One

Last year, the Supreme Court in Mahmoud v. Taylor found that the failure to provide an opt-out for pro-LGBTQ material in public schools violated the Free Exercise Clause. Texas will allow students to opt out of the Bible material under their current plan, presumably because it would be obviously a free exercise violation under Mahmoud otherwise.

But the opt out shouldn’t have any effect on the Establishment Clause challenge. Pennsylvania added an opt out while Schempp was being litigated, and the Supreme Court still struck it down. The Court’s logic was simple: the Establishment Clause isn’t about coercion. But it’s also clear why an opt out can’t cure all free exercise problems—opting out marks a student as an oddball. Another problem: it’s not at all clear what students will actually do if they opt out of the Bible readings, and if they can’t actually get the same reading instruction, then the opt out is illusory.

How the Texas Curriculum Fails Christians: The Example of Jonah

I think the subtler problem here, and the one that should create a free exercise challenge for Christians as well as other sects, is that the curriculum pushes a shallow vision of Christianity. There might be a litigation reason for that—Texas trying to shoehorn in the Bible into a language arts curriculum as just interesting literature rather than religion. (If that’s the case, then it is also a problem because it is secularizing religious material.)

I have been reading a lot of commentary from John Locke, Pierre Bayle, and Voltaire for my upcoming law review article, and one of the most interesting notes they all have is that state religion tends to devolve into empty slogans and symbols instead of genuine religious thought.

The Texas curriculum is a perfect example. I don’t have time to go over each grade’s mandatory readings, but I will go into depth about just one that I think is representative of the general failure here.

In first grade, students will read “Jonah and the Whale” (entertainingly listed as being “by Jonah”, which is not how the Old Testament works).

The specific verses are 1:1-5, 10-17, 2:10. This boils Jonah’s story down to:

• God told Jonah to go preach against Nineveh, but Jonah tried to flee on a boat instead.

• God starts a storm to stop the boat and tells the sailors on the boat to throw Jonah overboard, which they hesitate to do, but eventually they relent out of terror.

• God withdraws the storm, and a great fish (not a whale) swallows Jonah, who spends three days and nights in the belly. Then God tells the fish to vomit Jonah back up onto dry land. The End.

This version presents a simplistic and unfavorable vision of God. If He wants you to do something, you should just do it or he’ll torment you. He wants Jonah to preach; Jonah doesn’t want to do. So God hurls a storm at him and appoints a fish to swallow Jonah until he learns to do what God says.

Here’s the rest of the story of Jonah:

• Jonah goes to Nineveh, preaches that they’ll be destroyed, and the residents change their ways. God spares them.

• But Jonah is mad about the mercy (historical note: Jews hated Assyria, and Nineveh was the capital of Assyria), and he sulkily goes off and builds a shelter to watch what happens to the city.

• God grows him a plant as a sort of bonus to shade him from the sun. Jonah does nothing to care for it, so God sends a worm to eat its roots and kill it. That leaves Jonah subjected to the elements, so he complains about that.

• God points out in a rhetorical question that the people of Nineveh earned their mercy, and Jonah did nothing to earn his.

This is a way more complicated story that actually has some moral nuance that might be helpful for children to know!

Here is a short list of stuff that’s a more important and interesting than “do what God wants or he’ll have a fish eat you”:

• God defended mercy for Israel’s enemies over the objection of God’s own prophet. Withholding mercy is not God’s will.

• Jonah is basically a narcissist. He liked the plant and missed it because it shaded him (even if he didn’t do anything to help keep it alive), but he wanted to kill a whole city out of spite. (Indeed, Jonah is understood to be a satirically self-centered character, a jerk that we are not supposed to identify with.)

• God specifically notes the fate of the animals of Nineveh, showing divine concern and pity for animals (an early animal rights activist!)

Why am I pointing all this out? These are all lessons that Texas has excised from Jonah. One could suspect these parts were excised because they don’t comport with the shallowest version of conservative policy preferences.

That is precisely why Christians should be just as angry about the Texas Bible mandate as non-Christians. State religion inevitably lobotomizes the religion, leaving an empty, bland story about a big fish to try to appeal to children without teaching them anything worthwhile.

Subscribe now

I have a hypothesis about the real reason why the Declaration of Independence and American independence mattered so much, but I think it will be more evocative if I show rather than tell you what it is. So I’m going to do a little alternate history this week.

First, the real history. The committee of five in charge of drafting the Declaration of Independence in 1776 was John Adams, Benjamin Franklin, Roger Sherman, Roger Livingston, and a 33-year old nobody named Thomas Jefferson. But there was not a clear plan for independence at that point. The Continental Congress split their work into three committees:

• The declaration committee;

• The confederation committee (i.e., the one figuring out how the colonies would collectively govern themselves); and

• The foreign treaties committee (figuring out some opening diplomatic moves).

The key guy here is John Dickinson, who was on the confederation and foreign treaties committees. He was the most conciliatory Founder.

Here’s the point of departure: moderates at the Continental Congress agree to let the Declaration process move forward only if Dickinson is put on the drafting committee to keep it calm and conditional—to keep a door open to reconciliation.

Subscribe now

The Declaration of English Heritage

Today marks 250 years since the most audacious successful gamble in the history of diplomacy. After the failure of the Olive Branch petition in 1775, no one thought peace with Britain was possible. No one, that is, until John Dickinson, a humble farmer from Pennsylvania.

He had to contend with radicals, most notably the vainglorious warmonger Richard Henry Lee and an Enlightenment firebrand philosophe named Thomas Jefferson. Though little known today, Jefferson was a radical universalist. He and his faction were ready to cast off 700 years of English constitutional history bought with the blood of their forefathers.

But Dickinson had a vision. He argued that rights come from culture, and culture from God. In a word, the English were free because of their heritage.

When he was assigned to draft a declaration of the position of the Continental Congress in June 1776, he set to the task with a passion, and his words would echo down the ages:

We do not assert a liberty unbounded by law, nor rights existing in the fancy of every man in every nation. We claim those ancient and chartered liberties which belong to us as Englishmen; liberties planted with our fathers in this American soil, secured by long usage, defended by our own industry and courage, and held by us in trust for our children.

The rights of a free people are not indulgences from ministers, nor favors revocable at the pleasure of distant assemblies. They are inheritances, descending through time, confirmed by trial, religion, law, and custom, and made effectual only among a people capable of preserving them.

We desire peace with Great Britain upon terms of ancient and constitutional liberty. We seek no conquest, no revenge, no dissolution of affection with those in Britain who remain friends to the rights of a free people. If the Crown and Parliament shall renounce the claim of governing our internal affairs without consent, repeal the statutes by which these colonies are oppressed, withdraw their armies, restore our charters, and treat with us as free communities of the empire, we shall rejoice to see harmony restored upon foundations more durable because more just.

These words reached Parliament at an auspicious moment. Edmund Burke, the great conservative philosopher, spearheaded a new diplomatic outreach: would the colonies return, acknowledge Parliament’s rightful power, and in exchange be granted more independence to govern their own affairs?

This peace feeler arrived in the colonies the day after Manhattan fell to the British. Perhaps that explained the sudden shift in the Congress toward the Dickinson faction.1 A ceasefire was declared a week later as the diplomats gathered in Montreal to talk terms. Dickinson, of course, led the American negotiation team. A deal was struck after a few weeks:

• Parliament set an overall tax amount for each colony, but each colony would govern itself through its legislature and would choose for themselves how to raise the sum required.

• The American army under George Washington and any revolutionaries fearing retaliation would be permitted to go into exile beyond the Appalachians.

• Permanent military and naval bases would be set up throughout the colonies to prevent another rising.

• Loyal militias would be raised and equipped to keep the peace.

The genius of this plan was that it gave all parties a fresh start. Loyalists were empowered to make reasonable, good faith compromises in the colonies. The rebels were free to struggle to build their Enlightenment utopia in the wilderness. The mother country had reasserted the primacy of Parliament while providing a real measure of local control to colonial subjects.

Washington, Adams, and Jefferson built their nation on the shores of the Mississippi River. In a move of strategic brilliance, Washington led a surprise attack on New Orleans, seizing control from the faltering Spanish before the British could. This allowed their new nation to survive, though it would be decades before they could decisively defeat the confederations of Native tribes (armed with barely concealed glee by the British) that opposed them.

Jefferson and his ilk continued their philosophizing, though the rest of the world paid them scant attention. They were too busy worrying about Britain.

A New Rome

Britain emerged from the Colonial Rebellion the greatest European hegemon since Rome. It had a continent of resources. It had the plantations of the southern colonies, about to explode in value from the invention of the cotton gin. It had the nascent industries of the north, which would benefit tremendously from their close collaboration with factories in Britain. It had productive farms, a world class shipbuilding industry, and a hearty and rapidly growing population to draw upon for military strength.

And, like Rome, Britain was utterly reliant on slavery. The first stirrings of an abolitionist movement cropped up in the 1780s, but the cotton boom that started in the 1790s ensured it would remain a peripheral moral cause.2 Even if Britain had wanted abolition, it had ceded most internal control of the colonies to local legislatures as part of the Colonial Rebellion settlement.

Each colony governed its own affairs, and there was no reason they all had to agree on slavery—an issue that surely would have provoked tension in a United States of America had it survived the war.3 The northeast colonies continued to industrialize and profit off cheap cotton. The southern colonies continued to rely on slavery. The colonies produced no Great Emancipator, because the colonies didn’t have to agree with each other, and London merely had to let each colony follow its conscience.

Dickinson’s theory of rights through heritage fit comfortably into the philosophy of slavery. If rights are a tradition passed down among Englishmen, then what rights do Africans have?

Eventually, in the late 1800s, mechanization rendered slavery less economically beneficial. Britain and its southern colonies clung to it the longest, completing a gradual, 20-year abolition program in 1905.4

The World of British Autarky and English Rights

Continental European nations attempted to orchestrate various alliances against the British hegemon, but never managed to slow the spread of its empire. At most, they succeeded in preserving their own sovereignty, their own right to press their own nationalist claims against their neighbors.

The world industrialized. Economies grew, slowed only by the insistence of Britain on ruling all trade for the benefit of its own people (and consequently inviting a thousand forms of corruption).5 A small Britain might have been pro-free trade, but with the colonies, Britain could perpetuate an autarkic mercantilism.

British dominion on the Edmund Burke/John Dickinson theory of rights created a world of nationalism and real politik. Far from launching a governing philosophy of universal rights and consent of the governed, the Royal Colonies of North America provided a model for further British expansion. And the success of Britain became the model for the rest of Europe, and then the rest of the world.

Indians and Chinese joined Africans in the ranks of colonial victims. And in each case, the victims had no shared language of rights. Their colonizers were not even hypocrites, preaching universal rights but not granting them to all. They thought their rights came as an inheritance, and so those outside their clan deserved none. When the colonies eventually rebelled again, they did not unite behind philosophies. Nationalism came to dominate, as philosophies based on blood and inheritance sparked a hundred jostling claims for supremacy. After a violent uprising, India fragmented into a half-dozen ethno-states. China shattered into many warring factions.

Back to Reality

Jefferson’s ideas were not new in 1776—he didn’t even claim that they were. I don’t think the Continental Congress even really knew what it was doing when it adopted his language.

We have a hard time understanding just how profound the Declaration was because we live in a world that is pervaded by the idea of universal rights and consent of the governed. Yes, Jefferson himself was a hypocrite, a slaveholder, as were many of the Founders. But almost everyone ended up buying into his basic ideas. The French pretty much poached the Declaration in 1789. The 1948 UN Declaration of Rights followed the formula. Ho Chi Minh followed it.

But the ideas wouldn’t have had the same attraction to the French, and after them everyone else, if they weren’t so firmly, permanently affixed to the United States. The United States had a diverse citizenship who had more or less just arrived, strong rule of law, a continent of resources, and few plausible strategic threats. The conscious embrace of universal rights and the success of the country provided an example for the world.

The U.S. being forced to deal with slavery freed Britain to pioneer abolitionism. And the constant, grating contradiction between the Declaration and slavery ratcheted up the pressure to do something about it in the United States. That friction also produced Lincoln, and a whole new wave of universal rights (a new birth of freedom, to steal his phrase).

Whatever happens now or in the future to the United States, the Declaration probably irreversibly set the world on a path to universal rights. I don’t know if there will even be a United States in a thousand years, but if there are still people, they will still be talking about inalienable rights and the consent of the governed.

Subscribe now

I’m going to do the 250th anniversary next week (so, a few days after the 4th instead of five days before it), and Texas’s blunder will have to wait two weeks. But don’t let that ordering fool you, the Texas thing is absolutely disastrous, and I can’t wait to get into it more.

Subscribe now

This Report Doesn’t Care About Getting Things Right

I’m going to do a lot of nit-picking of this report. But I’m going to start with the one central argument that pervades the whole thing: the idea that the "the wall of separation” between church and state "took off in the mid-twentieth century, when the Supreme Court began citing Thomas Jefferson’s 1802 letter to the Danbury Baptist Association, written 10 years after the First Amendment was ratified.” That’s from the Executive Summary, page 21.

The report part of the report (i.e., the part no one reads, but which serious reports put effort into nonetheless) repeats this idea on page 42, which describes Justice Hugo Black as “repurpos[ing] a metaphor coined by Thomas Jefferson in an attempt to legitimize the new critical philosophy by placing it in the constitutional tradition” in a 1947 Supreme Court case.

This argument has one big problem: the Supreme Court unanimously endorsed Jefferson’s wall metaphor in an 1879 opinion: “Coming as this does from an acknowledged leader of the advocates of the measure, it may be accepted almost as an authoritative declaration of the scope and effect of the amendment thus secured.”

This particular screw up matters because to the extent the report tries to actually engage in history, it depends on the 1947 timeline. The report offers a pseudo-intellectual thesis about how Nietzsche, Michel Foucault, and Jean-Paul Sartre “[i]n the 1900s” (BUNCH OF LEFTIST WEIRDOS!1) concocted the wall of separation idea. Then a bunch of woke mid-20th century woke Americans like Justice Hugo Black (former KKK member and proto-originalist who voted against finding a constitutional right for married people to use contraceptives in Griswold v. Connecticut, that woke justice) imported it into the United States.

But no one who wrote the report bothered to read anything about it, so unless Sartre had a time machine, the Supreme Court was already all over the wall of separation thing (which, it must be stressed, was used by Jefferson in the lifetime of people then on the Supreme Court).

Also, in the case where WOKE Hugo Black supposedly smuggled the wall of separation into constitutional law (Everson v. Board of Education, 1947), he wrote the majority opinion upholding a state law against constitutional Establishment Clause challenge.

Oh also, I know Nietzsche is hard to spell, so I can’t mock the writers too hard for misspelling it in the footnote below, but I am going to mock them for citing the Kindle eBook version of one of the most famous books of all time, readily available for free online. I’m also going to mock them because Nietzsche died in 1900 and all of his famous works are from before the 1900s.

This Report Argues So Hard Against a Position No One Holds

Over and over, the report indignantly insists that people want to “exile the practice of religion from public life” or that the wall of separation means “the church and state are opposed to one another” (both on p. 43). If you want another version that is even worse written (p. 13): “far too often in our national life, religion is treated not as a protected and valued contribution to public life, but as a problem or annoyance to be managed, restricted, or sidelined.”

The writers never really explain what they mean by “public life” because they want to make it seem like people are walking around forcing Christians not to wear cross necklaces in public or something like that.

The religion clauses really boil down to something like this: the government (federal, state, or local) is not allowed to pressure you to hold any religious views. That’s it! You personally can do whatever non-violent, non-coercive thing you want to spread your religion, even in public! But there are things the state can’t do to promote religion. You might love your religion so much that you want to use government to force people to pay attention to it (see, e.g., Texas’s new Bible curriculum), but we don’t allow that.

We don’t allow government to pressure people about religion for two big reasons: (1) it inevitably stirs up violence and conflict in society, and (2) it inevitably corrupts the religion it sets out to spread.

The second part is subtle, but clear in the historical thinking on the subject. This report fatuously cites Martin Luther King Jr.’s “A Knock at Midnight” for the proposition that “[r]eligion offers moral standards to which government can be held accountable, keeping tyranny at bay.”

The writers should go back and actually read that work.2 King is not praising “religion” in general and saying government should promote it—he is excoriating churches that cozy up to conservative governments in pursuit of power. Read this paragraph and see if you think King wants the state endorsing churches:

Then there are those who have gone to the church for the bread of economic justice. How often has the church left men standing in the frustrating midnight of economic deprivation. In so many instances it has so aligned itself with the privileged classes and defended the status quo that it found it impossible to answer the knock at midnight. We must never forget the lesson of the Greek Church in Russia. This church allied itself with the status quo and became so inextricably bound with the despotic czarist regime, that it was impossible to get rid of the corrupt political and social system without getting rid of the church. This is the fate of every ecclesiastical organization that allies itself with the status-quo.

Churches that ally themselves with government have their beliefs dumbed down, softened, and compromised. This is why James Madison and John Adams, both extremely devout, very much supported the idea of a “wall of separation” even if they didn’t use that same phrase.3

One of the things I have said repeatedly about the Arkansas Ten Commandments law is that it violates the free exercise rights of Protestant Christians just as much as it violates the rights of Catholic, Jewish, Muslim, or non-believer students and families. The state legislature rewrote the commandments many Protestant Christians deem sacred in order to try to make them more broadly appealing. That is precisely the dynamic that motivated the religious Founders to oppose government involvement in religion even when it was their own religion that was being promoted!

This report is so full of political policy statements that have nothing to do with religion or religious views—criticizing COVID vaccines, criticizing transgender rights, criticizing a bank that closed former Kansas Governor Sam Brownback’s bank account (not making this up, it’s on pages 166-67)—that the whole thing is demeaning and disrespectful to people who actually fought and argued for religious freedom.

A Million Little Stupidities

There are so many little things in the report that don’t make any sense. I’m not going to flag them all because I don’t have time, but here’s a sampling.

The stereotype about reports like this is people only read the executive summary. A good dividing line between a legitimate report and a shallow political exercise is what the actual report says. Is it well reasoned? Did people try to get the arguments right?

Here is the very first page of the report, in which the line spacing randomly changes in the fifth paragraph, perhaps because “Commissioner Dr. Ryan Anderson” rocked his completely incomprehensible quote so hard:

Commissioner Kelly Shackelford (p. 29): “Our Founders called this our first freedom because they understood: if you lose this freedom, you will lose all of your freedoms.”

First, of all, the Founders didn’t call the religion clauses that. The stuff that became the religion clauses were in multiple parts of Madison’s bill that became the Bill of Rights—none of those parts were in his proposed First Amendment.

Second, the Founders thought the First Amendment as a whole (much less just the religion clauses) was less important than, say, the right of habeas corpus. How do we know that? Well, they put the right of habeas corpus in the Constitution itself at the Philly convention. The amendments came about later as an attempt to appeal to voters who were somewhat persuaded by Anti-Federalist arguments. A bunch of Federalists, including Madison, were skeptical that the amendments were needed at all.

“Repeal the Johnson Amendment which purports to give the government authority to regulate religious leaders’ sermons and spiritual guidance to their communities.” (p. 23)

The Johnson Amendment says that if churches endorse political candidates, they can’t claim non-profit tax exempt status. That puts them at exactly the same standard as every other non-profit. Obviously, if you let churches endorse political candidates and then claim non-profit status, every single candidate and every single non-profit is suddenly going to declare itself a religious organization, and all political spending will be tax-exempt.

It is “regulating” religious leaders’ sermons in exactly the same way that, say, rules about the practice of law regulate sermons—after all, a pastor can’t say in a sermon, “I am acting as your lawyer, and you should go ahead and lie on your tax returns!” The Johnson Amendment is a completely religiously neutral law, and repealing it would be a huge catalyst for political spending (which is, of course, why hyper-political operators want it gone).

“Jefferson [] coined the phrase ‘wall of separation’…” (p. 42)

Nope. Roger Williams, a 17th-century minister, take it away:

“[W]hen they have opened a gap in the hedge or wall of Separation between the Garden of the Church and the Wildernes of the world,” he warned, “God hathe ever broke down the wall it selfe, removed the Candlestick, &c. and made his Garden a Wildernesse.”

I am out of time to single out more of these errors. But take my word for it, if I had all day, it wouldn’t be enough time to get to them all.

On Creators and Unalienable Rights

The writers of the report, like so many religious conservatives, are obsessed with the idea the entire edifice of American liberty is built upon worship of the Christian God. The cornerstone of this case rests on the phrasing of the Declaration of Independence:

We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights…

I’ll pick just one overheated, silly quote from the report about this, the one from earlier that caused the line spacing to change:

It’s important to understand that religious liberty matters because religion matters—because we are creatures, and we owe obligations to the Creator… there’s a duty that we, as creatures, owe to the Creator. And that’s what then creates the rights amongst men, which governments are then instituted to protect[.]”

Jefferson’s rough draft didn’t mention a creator at all. The phrasing got added later; most historians think Ben Franklin did it. The reason is that the logic of the argument has nothing to do with how we ended up as people deserving rights.

The idea is that we are created equal. Nobody has a higher claim to particular special treatment than anyone else. It is actually irrelevant how we were created equal—evolution, God, whatever.

Because we are equally deserving of rights, that implies that government should not do certain things that would ordinarily only be acceptable if you thought some people had higher rights than others (for example, tax them without allowing them political representation!). The rights are self-evident because we can immediately perceive them through reason. That is the exact opposite of a divine grant through revelation—there’s a reason there’s a Sermon on the Mount and not the “Self-Evident List of Facts Stated by Jesus”.

Subscribe now

Discussing state economies rapidly devolves into domestic politics, with conservatives arguing that blue states have stifling regulations and taxes and progressives arguing that red states have terrible infrastructure.

As a thought experiment, I want to talk about the economy of Arkansas as if it were a foreign country. What would we advise Arkansas leaders to do if we could completely ignore politics?1 How can we develop a better economy?

Subscribe now

How the Resource Curse Works

There’s still lots of academic debate over exactly how the resource curse works, but there’s a general consensus around something like the following:

• When the resource in question is discovered, its exploitation immediately becomes the most valuable thing to do in the economy. It crowds out other investment opportunities in the local economy—who would invest in, say, an uncertain tech startup when you can just become a gold miner and get solid, certain pay?

• The most efficient arrangement for foreign investors is to use cheap local labor, but retain all of the higher corporate functions and planning. This is especially true if the local education system is not strong.

• Centralized control over the resource breeds corruption. To stay in power, the leader allocates the profits from resource exploitation to politically connected groups to cement control.

• When the resource is depleted, loses value for other reasons, or the staggering weight of corruption finally consumes all of the profit, the country is left with an under-educated workforce, an authoritarian political system, and few other economic opportunities.

There are lots of sad examples. Russia and Venezuela definitely qualify. Nigeria and the Democratic Republic of the Congo qualify. Argentina historically is another example, with resources closer to Arkansas’s (agricultural).

The single most important thing to note here is that this is not an inevitable process. At most, it is how things tend to develop when there is not a conscious attempt to avoid it. So, it’s not really a curse, it’s a failure of government to address a predictable problem.

How do countries avoid it?

The key is to move up the value chain. The value chain is basically: raw material/routine service —> processed product —> engineered components —> integrated system —> software, design, and intellectual property —> brand, distribution. Think of, say, the United States going from mining bauxite to forging aluminum to building airplanes to designing airplanes the whole world uses.

The resource curse is especially pronounced in countries that were poor before the resource was discovered because they lacked the institutional capacity to (a) resist corruption and (b) use the resource boom profits to move up the value chain. But even some formerly poor countries figured out ways to beat the curse.

Taiwan is a great example that I’ll walk through a little, but there are lots of others. South Korea and Botswana are also good examples. China also fits this mold, of course.

Countries that successfully moved up the value chain did so over time with a ferocious commitment to diversifying.

Taiwan started with a very concentrated agrarian economy. It carried out land reform, producing a relatively equal, literate society (there was also cultural pressure in favor of education).

It started out with cheap labor as a sort of natural resource. That resource was tapped in labor intensive industries like garments and simple electronics assembly.

At the same time it was cashing in on the cheap labor, Taiwan coordinated major construction projects while expanding technical and engineering education.

The government made a deliberate bet on semiconductor technology. It funded an institute to study the technology in the United States, centralized the lessons, then spun the capability out to private firms like TSMC. If you’re not familiar, TSMC is now, well, one of the most important and technologically advanced companies in the world. We very well might fight a war against China some day in part to protect TSMC!

If you want a more tangible resource than cheap labor, consider Finland. It started out with a forest economy—lumber and the like. It built up engineering and research capabilities, and now is a world leader in industrial machinery, electronics, and a wide variety of tech.

Being Cheap is One Step in the Plan, Not the Whole Plan

Note that Taiwan competed on cheap labor, but that was just to kickstart the development engine. The plan can’t be cheap labor forever because when cheap labor starts developing the economy, it stops being cheap labor.

This is a crucial point in the Arkansas context. Obviously, Arkansas can compete by being cheaper than other states in things like housing—that was the premise of the state plan to lure remote workers during COVID. To be clear, that might be a clever plan, but that advantage is inherently fleeting. If you succeed in attracting people with low costs, demand for housing and everything else increases, and prices rise. By the time they do, you need to be prepared to move up the value chain, or you’re stuck.

So How Do We Make Arkansas Into Taiwan and Not the Congo?

Arkansas has a bunch of economies stitched loosely together. Northwest Arkansas is doing great thanks to Walmart, Tyson, JB Hunt, and the University of Arkansas. The Delta in the east is one of the poorest parts of the country. South Arkansas has oil and gas and, soon, lithium mining. Central Arkansas is a kind of government and services hub. The Ozarks run on tourism, retirees, and some manufacturing.

All of these areas can climb the value chain further. I would summarize the formula as: (1) invest in education broadly at K-12 and in targeted industries for technical schools/universities; (2) when allowing outside investment to extract Arkansas resources and profit from cheap Arkansas resources, use those agreements to build up the next step in the value chain; and (3) invest in specific steps to help Arkansas companies climb specific value chains one rung at a time.

• In Northwest Arkansas: Climb the innovation value chain connected to retail, logistics, and transportation. We eventually want Northwest Arkansas companies selling technologies that Walmart and JB Hunt can use—demand forecasting, warehouse robotics, etc. Similarly, we can climb the value chain in poultry and food processing—food safety, animal health and nutrition, production automation, consumer brands. There’s already a technology accelerator program, but it needs to be a much bigger priority. Retain University of Arkansas graduates to start innovative businesses here. Fund industrial engineering, computer science, and related programs. Make the U of A a national star in those areas—throw money at it. Also, to keep talented people here and attract entrepreneurial outsiders, we need good housing supply. We need to let the major cities here densify and build.

• Central Arkansas: current big industries are healthcare and some nascent manufacturing, especially aircraft. The aerospace market could be lucrative, climbing the value chain from completing aircraft to making specialized components and avionics, and eventually managing maintenance cycles.

• Northeast Arkansas: climb the value chain associated with steel (lots of steelmaking plants there now). For example, you can build metallurgical expertise in parallel with the plants (and also drawing on local educational infrastructure).

• The Lower Delta: This is where you try to move up agriculture value chains. Without going into each one individually, you can see how this goes: the crop (rice, for example), the finished product (processed rice, including specialty varieties), the stuff made from the finished product (things like frozen dinners), the branded version of the stuff made from the finished product (high end frozen dinners).

• Southern Arkansas: Climb the value chain associated with lithium brine extraction. Develop local businesses that produce lithium chemicals, cathode and battery materials, and then battery component manufacturing. Each step needs to be paired with workforce building—community college and university programs, supplier development, etc. Don’t just offer incentives for outside companies to build a plant here, require them to do local hiring and training.

There’s a lot more detail I could go into about other regions, but I’ll save them for a later post. You get the idea.

I haven’t been saying this for each region, but it should be obvious by now: all of this depends on education. And I know, I’m an academic, but the important part is focused educational spending. The U of A is never going to be the world leader in every academic category—it’ll probably never have the best medical school in the world. But we should consciously try to make it the system with the best industrial engineering for specific fields. We need K-12 to be effective in creating a population of literate people who are ready to take the next step.

Technical schools and apprenticeships are part of this story, but they can’t be the whole story. Remember the Taiwan scenario: you always need to be building the population’s skills so they can take on the next step. We need technical schools and apprenticeships to go from extracting resources to working with them, but we need university-level education to go beyond that. In 10-20 years, we need the next generation to be ready to hit the next part of the value chain running.

Stepping Back into Politics (But Being Cool About It)

If it were easy to make these choices, then every country would do it. It’s easy to say “just climb the value chain.”

But as we saw with Taiwan, it involves simultaneously funding research and helping industries grow, but also subjecting them to world competition so that they get better. It’s not easy politically to ensure that governments keep making the right decisions for decades at a time.

For example, if you’re trying to seed local industries to move up the value chain, one way to do that is protective tariffs. A lot of Latin American countries tried that. The problem is tariffs weaken competition. So, you end up with inefficient companies that serve a captive market and never become viable internationally. That was the path Taiwan was on until it switched gears and became much more open to trade in the 60s.

But think about how tough it is to take tariffs or subsidies away from a politically powerful industry. Now try to imagine doing it in a politically divided electorate. The same idea happens if the industrial policy gets captured by politics. Money goes to people with political connections and the failures are propped up forever.

That, in a nutshell, is the argument for libertarian economics—that government is doomed to fail in the effort to consciously move up the value chain.

But we know that it’s possible! And the bummer is that if you don’t try to climb the value chain, you also generally don’t succeed—just ask every weak state in Africa and Central America.

So, all we need is Arkansas politicians to be willing to make wise choices, not act dogmatically (either supporting investment in everything or never supporting investment in anything), and resist corruption. Easy peasy!

Subscribe now

Fayetteville, Arkansas is having its own data center debate right now. The city has proposed new regulations on data centers. They exist within a tricky legal framework: the state passed a law in 2023 that may or may not prohibit the regulations in question. And if they don’t, the state might just take another crack at it.

On their face, the proposed Fayetteville regulations address the issues we all have heard about regarding gigantic (“hyperscale”, to use the industry phrasing) data centers: water use, electricity use, and noise. But when you dive into the details, Fayetteville was basically already a terrible place for hyperscale data centers that actually incur serious water, electricity, and noise problems.

The new regulations would prevent small and medium-sized data centers in Fayetteville, the kind that aren’t really associated with AI. I have no idea if that was the specific intent of the regulation, but the debate does not seem to be tracking that point, so I will expound on it a bit.

The bottom line is that it could be rational for Fayetteville’s planners to decide data centers aren’t as lucrative or job-creating as, say, a drone manufacturing plant. But that should probably be a conscious decision, not the unspoken or accidental byproduct of a regulation that seems on its face to be aimed at a different problem.

Disclaimer: I have done legal consulting work for and about AI companies, and obviously AI companies need giant data centers, so that is a potential source of bias. But also, I do pay for electricity and water in Fayetteville, which could become more expensive with data centers, so I also have a bias on the other side. Two biases make a right, maybe?

Subscribe now

Data Centers Don’t Have the Same Problems at all Sizes

First, when you hear “data center”, I want you to just think “a bunch of computers in one place.”1

There is more than one reason to have a bunch of computers in one place. The various reasons also call for different numbers of computers—different sized facilities.

AI companies use crazy large facilities because they can pool all the big fixed costs (cooling plants, security, staff, etc.) in one place to push down the marginal cost of computer power.2 There’s also a more technical reason—keeping all the computers together decreases the time it takes to communicate between them, which is helpful when you need tons of them working together to do AI magic.3

But not everything you use computers for requires that many computers.

The typical way to measure data centers is by the amount of electricity they need. The absolutely gigantic ones take 1 GW (a billion watts) or more to run—the one I linked to takes 7.5 GW. The more prosaic “hyperscale” facilities might take a few hundred MW (megawatts, so a few hundred million watts).

The data centers we all hear about in the news are the monstrous “hyperscale” facilities (or the even bigger ones that are variously called “frontier”, or “titan” campuses). Those are for AI because they need so many computers and it’s more efficient to have them in one space.

But there’s also an economic need for not so gigantic data centers (remember: just a bunch of computers in one place). For example, if you generate and use a lot of data locally (say, if you’re the headquarters of the world’s largest retail company), then it’s expensive to send it all over the world for processing and then back to you. You should just have a data center (or a few) right there. But they don’t need to be monsters. A small data center is ~1-10 MW, and a medium might be ~10-50 MW.

Bigger facilities tend to use resources more efficiently, but smaller data centers are just so much smaller that they use a comparatively tiny amount of resources. A 1GW hyperscale data center uses almost as much electricity as a million homes. There’s simply no way to add that to an existing electrical grid without reworking the whole thing at large expense. A 20 MW data center uses the same amount of electricity as a few thousand homes. That’s still a lot, but it’s not necessarily going to require major changes to the grid.

Water use scales similarly—big facilities use a lot more, small and medium facilities do not necessarily have a significant impact.

Fayetteville Was Already a Bad Place for Hyperscale Data Centers

Fayetteville (and Northwest Arkansas generally) is a terrible place for a hyperscale data center because of how we get our electricity. We have cooperatives (co-ops). Those are nonprofit entities owned by the members they serve. They are a historical holdover from when rural electrification was happening—it wasn’t profitable to serve rural areas, so coops banded together to make it happen without outside investors looking for profit.

Consistent with the plucky little guy motif, our coops are distribution only. They don’t even generate their own electricity. They buy electricity from generation-and-transmission coops.

The downside is that coops don’t have that sweet, sweet investor capital waiting to, say, build a new gigantic power plant to serve a hyperscale data center.

Compare that to, say, Entergy, an investor-owned utility that serves central and eastern Arkansas. If you look where most of the big data center projects are in Arkansas, they are in areas served by Entergy.4

Investor-owned utilities can be very annoying. They basically have a monopoly, and they earn a regulator-approved profit, so they have every incentive to keep building more infrastructure even if it makes electricity more expensive for existing customers. But more infrastructure is exactly what the big data centers require, so investor owned utilities and data centers are a match made in heaven.

Stamping Out Small and Medium Data Centers

So, Fayetteville was never going to have hyperscale data centers. The new regulations obviously would make those centers even less feasible. There’s no way a hyperscale project could avoid having an impact on water and electricity rates. But small or medium data centers could do that.

The killer, and less flashy, part of the regulation is a setback rule. Sound and heat are far less important issues for a small or medium data center, but under the Fayetteville regulations, they too must be 1,000-feet back from any residential zone.

As the news articles on the regulations have pointed out, only 2% of Fayetteville is zoned for industrial use. If you add a 1,000-foot buffer from residential areas, you end up with very few possibilities for a data center—basically, you’re already deep into industrial parks which might not be contiguous, accessible, and actually developable.

Probably Not the Best Way to Run a Railroad

So, the regulations probably kill the possibility of small or medium-sized data centers. Is that a good thing?

I don’t love a regulation where you need a Substack post to figure out the actual impact. Here’s what my request would be to the proponents: please explain why we don’t want small or medium data centers.

There could very well be a good reason! Maybe we think we’re higher up the industrial value chain—why even allow small or medium data centers when we could be getting drone manufacturers? But if a higher value industrial concern wants to move in, presumably it would outbid a lower value concern, so I’m not convinced you need regulation to stamp out the lower value concerns.

Alternatively, keep the water and electricity parts, which would pretty much only rule out hyperscale facilities, but get rid of the setback rule. Fayetteville actually already has more nuanced noise regulations (something any SEC college town already knew about pre-data centers)!

Arkansas Also Might Preempt the Whole Thing

Arkansas passed a law in 2023 that, despite being named the “Arkansas Data Center Act of 2023”, might not actually be relevant. The law actually only specifically covers “digital asset miners”—i.e., cryptocurrency miners. The name was presumably chosen to make it sound more likable and anodyne, since crypto is, well, generally not as likable and anodyne. The law prevents Arkansas towns from banning crypto miners, but by its text (other than its title!) does not really affect data centers per se.

It is possible but unlikely that a court would find this act applies to non-crypto mining data centers. If it doesn’t, then it is possible Arkansas could also intervene with another bill that actually does preempt municipal regulation of other kinds of data centers. I’m not sure I would bet on that though—there seems to be quite a lot of ground-level antagonism toward data centers, making the issue tricky politically.

AI as Final Exam for Humanity in Creating Pareto Efficiency

As much as politicians pretend otherwise, it is almost never the case that an event or policy is good for absolutely everybody. If someone discovered a giant gold deposit in Fayetteville tomorrow, it would make a lot of people rich, but it would also be disruptive for lots of people. Locals would complain about the increase in traffic and the strain on city resources from all the people coming to work in gold mines. Housing prices would go up. There would be noise from construction. You get the idea.

Economists have thought about this problem. Having a gold mine is good for the owner, good for the people who buy gold, but it has some bad effects on third parties. Economists call that a negative externality.

But economists have a solution (in theory)! If the good of some enterprise is worth more than the negative externalities, it should be possible to redistribute some of the good to the people who suffer the negative externalities so that everybody is at least as well off as they were before. That concept is called Pareto efficiency.

AI is basically a giant test in whether humanity can figure out how to create Pareto efficiency in a tricky situation. AI can already do amazing things, even beyond making graphics for this Substack. It’s useful in serious ways.

But it has lots of negative externalities. It makes academic dishonesty a nightmare to detect. It is going to put a lot of people out of work, probably. And data centers, utterly necessary to running AI, are a pain.

The real question about AI, as many people have pointed out, is whether we can redistribute the vast winnings to people who lose their jobs, or live next to data centers, etc. This absolutely should be possible in theory, and every time an AI problem emerges, we should be thinking about how we can make it a win for everyone.

This part of the debate gets lost sometimes between people who are blind to the problems data centers pose and people who are hostile to AI generally and oppose all data centers on those grounds. In the context of municipal AI regulation, I would prefer an approach more tailored to “how can we make sure everyone wins”.

Subscribe now

Subscribe now

#4, The Sevier-Newton Duel of 1827: The Fanboy Duel

If you’re familiar with dueling from history or literature, you may know that each of the duelists has a “second”, who is generally an assistant. They negotiate before the duel, help with set up, make sure the rules are followed, etc.

Ambrose Sevier and Thomas Newton were not even seconds.

They were just kind of fans of two more important people who were dueling (we’ll get to them later in this power ranking). They dueled right before their guys in the main duel, and only for the honor of the main duelists.

But the crazy thing is that Sevier and Newton are both famous in their own right—so famous that they would each end up having a county in Arkansas named after them!

Sevier was the grandnephew of the first governor of Tennessee, and he would go on to become the head of an influential group known as “the Family” that would dominate Arkansas politics for decades. He would also become one of the state’s first two U.S. senators. Also, in 1842 the Arkansas legislature censured him for financial malfeasance tied to bonds of the failed Arkansas Real Estate Bank. Just remember that institution; it’s going to come up again.

But Sevier had a tragicomic end to his story and, well, life. President Polk sent him to Mexico to help finalize the peace treaty that ended the Mexican-American War. but while there he fell ill. While he was away, a guy named Solon Borland took his Senate seat as a supposed placeholder, but then stunned Sevier by beating him by four votes. Sevier never recovered and died on December 31, 1848.

Newton wasn’t quite as successful, but he also has a crazy story. He studied law under one of the duelists in the main duel, so felt honor-bound to fight on his old teacher’s behalf. (Note to past and future students: you do not need to fight in duels for my honor until after I get tenure.) After the duel, he became a bank cashier (I guess the law didn’t work out for him?), became a state senator, and then became a U.S. representative.

I guess I spoiled the outcome by telling you they both survived, but maybe they had some mega cool fight beforehand where they just fought to a standstill and ended the fight out of grudging respect?

Nah, this is #4 on the list. They both fired in the air and their own seconds called off the duel. I have no idea how that constituted defending the honor of the main duelists. I’m increasingly understanding why most of the Founding Fathers thought dueling was stupid.

#3, The Brooks-Baxter War

This is going to sound like Harry Potter very quickly, so bear with me.

The gubernatorial election of 1872 pitted carpetbagger (i.e., a guy who moved into the state just to run for something) Joseph Brooks against scalawag (a pejorative term Southern Democrats used against white men who supported federal reconstruction) Elisha Baxter. Brooks led a faction of radical Republicans known as the Brindletails. Baxter was a representative of the mainstream Republicans, called the “Minstrels.” (Yikes.)

Their platforms? Virtually identical. Who actually won the election is unclear, but Baxter’s people controlled the election machinery and declared him the winner. Seems legit.

Brooks tried to contest the election in court for two years and got nowhere—until Baxter made a teensy little mistake. He opposed a railroad construction bond. Party bosses met behind the scenes and resolved to remove him. Suddenly, a Pulaski county court awarded the governorship to Brooks.

Brooks stormed into Little Rock with armed men and the Pulaski county sheriff. He demanded Baxter leave. Baxter, with no armed men of his own, duly complied, but then assembled his own armed force outside Little Rock and wired President Grant for assistance. He marched right back in with his men.

Meanwhile, Brooks’s men had broken into a state armory and stolen two six-pound cannons. Not to be outdone, Baxter’s men dug up a bigger Civil War-era cannon and dubbed it the Lady Baxter. The two sides fired shots at each other, prompting a federal arsenal commander to build barricades between the two forces.

But outside Little Rock, a real conflict brewed. There were various ambushes, and an estimated 200 people became casualties. In case you were wondering, the Arkansas Industrial University (which became the University of Arkansas, Fayetteville) sent munitions to arm Brooks’s men.

How’d this crazy thing end? Well, Baxter’s attorneys in Washington (including the brother in law of Sevier, who inherited Newton’s congressional seat—small world) successfully persuaded Grant’s attorney general to declare Baxter the winner. Brooks folded.

So why is this only #3? Well, Baxter stayed in his room heavily medicated throughout much of the war. It’s one thing if the loser of a war does that, but when the triumphant victor does, it makes the story a little less impressive.

#2: The Conway-Crittenden Duel: The Main Event

This is the main event that the Sevier-Newton duel immediately preceded! And like the undercard duelists, both of these participants also became the namesakes for Arkansas counties, so you know they were big deals.

Given that such big figures were involved, surely there was a good reason for the duel, right?

Well, you decide.

Robert Crittenden and Henry Wharton Conway were good friends. Conway was basically a patron of Crittenden, who was a very powerful man in Arkansas. His older brother was a U.S. senator from Kentucky, and older bro used his influence to get Crittenden appointed the first secretary of Arkansas Territory in 1819. Crittenden served as acting governor while the real governor dawdled on his way to Arkansas, so Crittenden appointed basically everyone and consolidated power. He also pushed to move the capital from Arkansas Post to Little Rock, where he owned real estate that became quite valuable.

Conway was a territorial delegate to the U.S. House of Representatives. But in 1827, he ran against a challenger, Robert Oden. Oden alleged that Conway had misappropriated $600 meant for the Quapaw (a.k.a., the Arkansas tribe). Conway’s defense was that Crittenden had told him he could take the money. Crittenden denied saying that. (Also, bro, you can’t just do something obviously illegal because a cool guy told you you could!)

Duel on!

Unlike the undercard duel, this one got real.

Conway fired first, and allegedly hit a button on Crittenden’s coat, causing the shot to carom away harmlessly. Crittenden then fired back and hit Conway in the chest. Conway died in agony 11 days later. The Arkansas Gazette originally described his wound as “severe, but not dangerous.” Oops!

But the weird thing is that Conway’s family became, well, The Family of Arkansas politics for the next several decades because Sevier—he of the aforementioned fanboy duel—ended up winning the special election to fill Conway’s seat. So Conway lost the duel, but kind of won the war.

#1 (The GOAT), The One Where the Speaker of the House Stabs a Representative to Death on the House Floor

This is the reason I started writing this post. Where to begin?

December 4, 1837. The news is good: the new state of Arkansas has a predicted tax surplus! The governor has called a special legislative session to figure out what to do with it.

One Senate bill proposed to pay bounties on wolf pelts. But a practical question arose: how do you verify that a particular pelt presented to a county justice of the peace is actually from an Arkansas wolf?

Representative Joseph J. Anthony was there, and while he seemingly didn’t really care about the wolf bill, he did care about the Arkansas Real Estate Bank. And he had an idea for a fateful bon mot.

He offered an amendment to the bill asking that “the signature of the President of the Real Estate Bank be attached to the certificate of the wolf scalp.” Sadly, given what happened, the joke isn’t even good enough that it stands without explanation—he meant that the word of the bank was worth as much as a dead wolf.

But the Speaker of the House, John Wilson, was the newly elected president of the bank. And he was not in a joking mood. He took the entirely joking amendment as a personal insult.

He descended from the speaker’s platform, Bowie knife in hand. One representative tried to throw a chair in his path to stop what was about to happen. But instead, Wilson stepped up on the chair for additional leverage and stabbed Anthony dead on the spot.

The Arkansas General Assembly expelled Wilson, the guy who threw the chair became the new speaker, and Wilson was tried five months later. But Wilson’s lawyer was a savvy litigant. The lawyer argued Wilson was merely protecting his honor. He and Wilson bought the judge’s meals during the trial, and had a crowd assemble and talk over the prosecution’s final argument. The jury duly returned a verdict of “guilty of excusable homicide” (so, basically not guilty.) Wilson celebrated by buying them all drinks. A drunken crowd even gathered around the lodging of Anthony’s relatives to mock them further.

This was a pretty embarrassing incident for Arkansas, which already had an iffy reputation. The Arkansas Gazette said it was “another example of the barbarity of life in Arkansas.” People latched on to the joke of calling the Bowie knife the “Arkansas toothpick”. A Bostonian of the era (two years before the duel!) wrote that good people “absolutely shudder at the bare mention of Arkansas.”

But I think nearly 200 years later we can all agree: that’s a pretty good story, and certainly the GOAT of absurd 19th-century political violence in Arkansas.

Subscribe now

Changing gears, the post this week is about an issue Arkansas had been leading on, then abruptly decided not to care about: federal preemption of state AI regulation. We’ll get into the substance, but we’ll also talk about the seeming disparity between the idealistic rhetoric used by public officials and political reality.

Subscribe now

State AI Regulation

State AI regulation has (predictably) diverged between red and blue states. It’s a complicated topic, but to give a 30,000-foot summary, red states have regulated specific AI issues—creating sexual deepfakes or child pornography, clearly allocating intellectual property rights on AI-products to the prompt writer, etc. Blue states have done that as well, but have also required deeper disclosure of information on the models themselves, reports of incidents, and testing.

The Trump Administration has been broadly hostile to AI regulation—federal preemption would be a way of stopping states from regulating AI more tightly than Trump wants. Trump himself and various surrogates have offered varying explanations for his opposition, and critics have suggested other, more corrupt motives. A brief rundown of the possibilities:

• Ideological opposition to regulation generally. It’s hard to gauge how likely this is. On one hand, obviously, Republican policy has been against regulation for decades. But on the other, Trump and his administration love regulation when it hurts political opponents or furthers other culture war objectives.

• Maximizing the speed of development to stymie China. This is a plausible explanation, and, in my opinion, the best on substance. Perhaps even good-faith regulation would slow AI companies to the point that Chinese competitors less concerned with safety could catch up. There are two big problems with this argument: (1) Chinese companies do not seem to be competing to make the same kinds of AI American companies are (they’re making cheaper AI); (2) Trump has blessed the export of NVIDIA’s second-best AI chips to China, which suggests his administration is not going all-out to stay ahead of China.

• Corruption of various flavors. There are lots of angles on this—Trump’s personal dealings with Gulf countries making major AI data center investments, for example. More prosaically, David Sacks, the “AI and Crypto” czar of the Trump Administration, has donated a lot of money to Trump and stands to gain tremendously from AI acceleration.

Whichever combination of explanations is true, the fundamental tension is that both red and blue states have regulated in this area, and the Trump administration has (until recently, which we’ll get to) broadly disfavored AI regulation.

Big, Beautiful Preemption

A little less than a year ago, there was widespread debate on whether Congress should preempt state AI regulation efforts in the one major legislative act of President Trump’s second administration—the One Big Beautiful Bill. Governor Sanders and her administration successfully led an effort to stop that preemption from happening.

They appealed to the ideals of federalism, the Founders’ idea that states should retain control over many areas of law. Here are some quotes from Governor Sanders’s Washington Post op-ed on the subject:

Attorney General Tim Griffin joined in:

This effort was successful: the preemption of state AI regulation was stripped from the bill before final passage.

That dust had barely settled when the Trump Administration issued an Executive Order that had the explicit aim of stopping states from regulating AI. The President’s stated concern in the order was preventing a patchwork of different regulations that would make AI development more cumbersome. But since the President can’t actually legally preempt states, Trump threatened instead to withhold broadband internet funding for any state that regulated AI in ways that it disapproved of.

I couldn’t find any statements by Sanders or Griffin criticizing this Executive Order. In the run up, Sanders reportedly reiterated that she had opposed the congressional effort. But since then, nada.

One big reason for Natural State silence might be that section 8 of the Executive Order has a safe-harbor provision that pretty much exactly tracks Arkansas’s current AI regulation laws. More expansive AI regulations, like those found in blue states, are not so lucky.

Who Cares About Federalism?

Federalism had been relatively fixed in politics for a long time—Republicans liked it (generally, but they would ignore it when convenient sometimes), Democrats did not (unless they occasionally opportunistically supported it when Republicans were in power). But increasingly, both sides are cohering around opportunistically name-checking federalism while ignoring it whenever it impedes policy goals.

The Founding-era logic of preserving state power so that states could test out different policy ideas makes sense, and it really does work in a variety of policy areas. Sometimes red states come up with policies that improve people’s lives, sometimes blue states do.

From a law and economics perspective, federalism is a public good. It would often be in everyone’s collective interest to allow states more authority to direct their own affairs, but the general good of federalism comes at a specific cost. It means the party in control of the federal government at a given moment has less power to enact its agenda nationwide. If either party unilaterally gave up that power by deferring to states, its partisans would view it as weak.

Republicans had defied this law and economics perspective in the Reagan era. President Reagan genuinely refrained from pushing the boundaries of federal power in at least some cases out of stated respect for federalism. Even then, that respect was inconsistent—Reagan also supported establishing a national drinking age, for example. But there were at least some concrete steps to allow states more latitude to try new things.

The AI regulation debate has shown federalism to be a vestigial organ in the GOP. It is still name-checked, but it serves no actual role. Meaningful respect for federalism would entail red-state leaders opposing attempts at unilateral preemption by President Trump—even when the preemption is specifically tailored to only affect blue states. But this is just another way in which the modern GOP’s principles have shifted. And if no one cares about federalism, then we shouldn’t be surprised when we stop reaping the benefits—no more Mississippi or Connecticut miracles.

Subscribe now

This week, I’m going to focus on a column in last week’s Arkansas Democrat-Gazette that said, among other things:

The most pressing question for th[e] forthcoming interim dean [of the University of Arkansas-Fayetteville School of Law]: whether he intends to move legal education in a direction that restores tolerance of viewpoint diversity.

…

In virtually all law schools, there is broad (or is it blind?) agreement on social and political questions that are contested in the real world--race, gender identity, DEI policy, speech norms, and, indeed, the purpose of legal education itself. Left-of-center political and social positions have been institutionalized as professional orthodoxy rather than treated as open questions for debate. Hiring patterns, faculty culture, and administrative leadership artificially imposed, and continues to enforce, that, uh, consensus.

The column, written by University of Arkansas-Little Rock Professor Robert Steinbuch, has some other claims in the same general vein that warrant discussion.

Subscribe now

Why People, Including Me Circa 2011, Have Been Claiming Legal Academia is a Monoculture

Today, I’m a self-proclaimed neoliberal, but when I was a 2L at Harvard Law School in 2011, I and a likeminded friend started a campus organization called “The HLS Tea Party“.1 I don’t want to speak for him, but I certainly felt there was an ideological monoculture on campus of chummy institutionalist progressivism. The Federalist Society was the largest group on campus, but it functioned mostly as a way to schmooze with law firm attorneys and judges visiting campus. My contracts professor Elizabeth Warren was about to run successfully for the Senate in Massachusetts, and everyone seemed to be on Team Warren. Meanwhile, I had cheekily brought in donuts for my 1L section when Scott Brown won his January 2010 special Senate election.

We invited some interesting people to campus to talk about issues, hosted a debate about the future of unions, things like that. We tried to get then-presidential candidates Newt Gingrich, Rick Perry, and Mitt Romney to come to campus to speak, figuring that New Hampshire wasn’t that far away and they were all there campaigning, though none took us up on the invite.

I’m sure I complained at the time about HLS being a progressive haven, “Moscow on the Charles”,2 an ideological monoculture, etc. But I never thought the issue had anything to do with how HLS was run. I never wished that the Massachusetts state legislature would threaten Harvard if it didn’t hire more conservative faculty. I thought most people on campus were progressive, and I wanted conservatism to win through persuasion. I wanted a fair chance to do that, and HLS allowed me that chance.

Is Monoculture the Biggest Problem Facing the Law School?

Today, conservative students at the University of Arkansas School of Law have pretty much the same chance I had at Harvard, perhaps even a better one. The students as a body are probably more conservative in Arkansas, though not outrageously so—I actually knew quite a few conservatives at Harvard. Several are now prominent figures in the Trump Administration. I have not noticed any disparity in guest speakers. I had a Trump appointee speak to my Constitutional Law class this spring!

Steinbuch complains that “virtually all law schools” have “broad” agreement on “social and political questions” that are “institutionalized as professional orthodoxy rather than treated as open questions for debate.”

The only evidence he cites for that proposition is that he personally attended “a forum” (where?) in which “an academic” (who—and the phrasing makes me wonder if they are even a professor?) said that “a law” (which law?) was unconstitutional because a trial court ruled that way. A “higher court” (which?) had upheld the law on appeal, so the only possible explanation was a “politically biased refusal or inability to present a complete picture of the law.”

It is simply nonsense to claim that this forum and the ideological bias one presenter may have had shows that ideology is the most important issue facing any law school, let alone Arkansas. Law schools have actual problems! Here are a couple that don’t have easy answers:

• What do we do about teaching AI? Should we have students draft memos with AI assistance to more closely mimic what many lawyers do in the real world, or do we just forbid it entirely so they develop independent judgment?

• How should law schools balance black-letter law with teaching students to think like a lawyer?

• How should law schools balance real-world experience (e.g., externships and clinics) with ensuring at least a solid baseline in more subjects?

If ideological diversity will only be sufficiently obtained when there is no forum anywhere in which a law professor says something stupid, the goal is unattainable. And, of course, if you want intellectual diversity, you’re going to get a wider spread of ideas presented, including bad ideas.

Diversity and How You Get It

Did you know there are far more animal species on land than in the ocean? The reason is that the ocean is far more interconnected—you don’t get as many isolated genetic niches for a new species to emerge in. In the ocean, a species emerges that does a superlative job of occupying a niche, then it spreads and dominates that niche over a wide range.

As the country and the world grows more interconnected (like an ocean instead of an isolated ecosystem on land), the optimal law school candidate has become a clearer and clearer niche. That naturally suppresses diversity of all kinds. If you want to maintain any kind of diversity—racial, ideological, geographic, etc.—you need to preference it in the admissions process.

The legal education system has had at least several decades to establish a competitive paradigm: grades + LSAT score.

In retrospect, to the extent there is/was an ideological monoculture, I think it was actually the result of admissions being largely focused on meritocratic measures. You basically needed to have close to straight A’s and a very high LSAT score to get in. To do that, it helps tremendously to have access to great schools and teachers, and study a major without curved classes (i.e., a non-STEM major). Also, you need to be someone who deals extremely well with structured learning and tests.

Not all accomplished or intelligent people fit that mold. If you want more ideological diversity, you basically need to start focusing on non-meritocratic measures of worth—maybe give preference to entrepreneurial experience or being someone who thinks outside of the box. But note that if you try to implement that system, it puts you into a more holistic, hard-to-pin-down admissions process—exactly the sort of thing the Supreme Court recently ruled unconstitutional in Students for Fair Admissions v. Harvard.

Meaningfully Promoting Intellectual Diversity

As a constitutional law professor, I am probably on the front line of how to tolerate and manage “viewpoint diversity.” If there was a plot to brainwash students into becoming Alexandria Ocasio-Cortez-niks, my class would be the place to start.

But no administrator or fellow professor has ever made a suggestion on how I should substantively teach the course, let alone pressured me about it. Nobody slipped me $100 and said, “Maybe tell them Roe v. Wade was correctly decided.” Nobody has said, “Tell them Lochner was an abomination or you’ll never get tenure.”

As anyone who has taken one of my classes will tell you, I bend over backwards to avoid expressing my views precisely because I want the students to express theirs. Nothing makes me happier as a professor than when students have a hot take on something, because it means that they care about the material, and caring about it is the first step toward learning it.

Because of that teaching philosophy, I would absolutely advocate for hiring an intelligent person of pretty much any political persuasion who was good at guiding and moderating thoughtful classroom discussions.3 I would advocate for such a person over an intelligent person who shared my views but who was not good at teaching.

I think what people on the outside of academia don’t get is that we simply do not have that choice very often. The world of people who are (a) qualified, (b) want to become law professors, and (c) want to do so in Arkansas is not huge. So the choice is pretty much never between two candidates, both perfect, one of whom is conservative and one of whom is liberal. It’s like buying a house—there are many variables that differ between candidates, and there are not enough candidates that fit all the requirements so that you have a full spectrum of options.

The only way we could guarantee having more conservative professors is to have affirmative action—making conservatism a preference in hiring sufficient to overcome a deficit in teaching experience, past published research, or professional experience.4

My conclusion is thus that the only acceptable way to promote intellectual diversity is to hire good professors who help students cultivate the best version of their own beliefs—nothing more, nothing less.

Structural Bias is Complicated

I think it is true that law professors as a group are more liberal than the country writ large—the median law professor is almost certainly more liberal than the median American voter. But that is an inevitable consequence of current politics separating out by education.

Voters with post-graduate degrees (e.g., every law professor) voted Democrat in the 2024 election by a roughly 2-1 margin—65-33. Consider also that after starting with twice as many Democratic options, a large number of possible candidates do not want to become law professors—they want to get rich working at law firms. For various reasons, such people tend to be more conservative. So, by the time you get to the pool of people actually applying for legal academia jobs, I would guess you are looking at more of a 4-1 margin.

But it’s even harder than that! A lot of legal academics are not super political. If you’re hiring someone to teach civil procedure or patent law, do you specifically care what their politics are? If someone doesn’t have any obvious ideological identifiers on their CV, should you ask them about their politics during the hiring process? Should you preference conservatives even if the students will never know what their politics are because they are too busy learning about nonpolitical stuff like the Erie Doctrine?

So What Should the Next Dean Actually Do?

To bring this all back to Steinbuch’s claim, I don’t really think the next dean (interim, permanent, “for life”, whatever) can do much to change intellectual diversity per se. All that person can (and should) do is ensure the law school is a place where faculty and students can express themselves professionally and reasonably without fear of punishment. I think that’s a pretty low bar that the law school already clears. The only prominent case I can think of where it was violated was the hiring and firing of Emily Suski.

I can’t resist noting one more part of the Steinbuch column on this point: “History shows that the healthiest intellectual environments are not those in which everyone agrees, but those that accept--or even welcome--open disagreement without exile.“

I absolutely agree. Why could we not have a dean who disagreed with Arkansas state legislators regarding the age at which trans girls should be allowed to participate in women’s sports?

The dean doesn’t even get to unilaterally decide on whether we embark on a conservative affirmative action plan to hire more conservative law professors. The faculty would have to vote for the candidates in question.

The best way to make law school students more conservative (or liberal, or neoliberal, or progressive, or MAGA) is to be a good example. If we want legal academia to do better, we have to exemplify rational debate. I would urge any professor, including Steinbuch, to argue about ideas and discuss evidence. So much mainstream political debate is about finding the weakest opponent, punching down at them, and then declaring that you DESTROYED them. The only way that’s going to get better is by being the change we want to see.

Subscribe now

The statute is in the news right now because Benton County recently charged a defendant with attempted capital rape, the first instance of such a prosecution in Arkansas.1

The whole issue raises tough and interesting questions in a variety of fields: constitutional law, law and economics, and criminal law for starters. So, even though it’s about as sensitive an issue as you can get, I’m writing about it this week.

Subscribe now

The Ugly Parts of Law and Economics

First, a warning: discussions of criminal incentives get gruesome, so please skip if you don’t want to deal with dark, sad stuff. The constitutional issues are related to this incentives discussion, but they’re distinct enough that you can skip this part and still get the gist of the constitutional debate.

I think we all get the moral intuition of a capital rape statute. Many, probably most, people wouldn’t shed a tear for the death of a child rapist, regardless of their opinion on capital punishment.

But making child rape a capital crime creates difficult incentive problems that at the very least suggest that we need to be careful in crafting a statute actually imposing a death penalty for child rape. As we’ll see, other states have taken a markedly different path on the same general issue.

The deterrent effect of the death penalty must be considered. This is an empirical question, one that I think it is fair to say is unsettled. Social scientists have been arguing about it for decades. And there aren’t many studies specifically about capital rape, which is of course defined differently in different states. That means you have to be very careful about jumping to conclusions. Even setting aside the well-known replication problems in social science, conventional significance thresholds accept a 5% false-positive rate by design — meaning roughly 1 in 20 'statistically significant' findings would arise by chance even if the underlying effect were zero. With only a handful of studies in the field, that's not a reassuring base rate

That said, a 2016 study found that adding a death penalty for child rape was associated with a 37% reduction in reported child rape overall. That is consistent with the idea that increased deterrence from a harsher penalty decreases the crime.

But there’s a caveat. Reported rape might be expected to decrease with the death penalty for a few reasons unrelated to deterrence.

The classic problem with imposing the death penalty for something short of murder is that there is then no marginal deterrence for adding murder. If we think child rapists are actually rationally calculating the deterrent effect of capital punishment, we also need to consider that there is no reason at all for them not to kill their victim if they are getting the death penalty anyway. Killing the main witness makes it less likely that a rape would be reported and a case brought against the perpetrator. So, for the criminal, there is much to gain and nothing to lose.

Sadly, there is some (limited) empirical work on this too that strengthens the incentives-based argument. One study found that mandatory 25-year minimums for child sexual abuse increased child homicides by approximately 80%.

Another terribly sad reason for fewer reports is that a capital case often requires the victim to testify multiple times. The trauma associated with suffering rape as a child is, obviously, awful. Being forced to recount it multiple times in court would be terrible. And the defense generally gets more latitude for cross-examination in capital cases during the penalty-phase of litigation, so not only would the victim have to testify multiple times, but they would be subject to harsher questioning than in a non-capital case. It is possible some people do not report for that reason.

What is the upshot of all this? One obvious takeaway is that you have to be careful about how you implement a death penalty for child rape, even if perpetrators deserve to die.

The Eighth Amendment and Capital Rape

If you wanted to skip the crime details, you can start reading here again.

Imposing the death penalty for child rape has been unconstitutional since the Supreme Court’s decision in Kennedy v. Louisiana in 2008. Understanding that case requires a little dip into Eighth Amendment jurisprudence.

The Eighth Amendment prohibits “cruel and unusual punishments.” The Supreme Court has developed a two-part analysis to determine if a punishment violates the Eighth Amendment. First, it looks to “objective indicia” of society’s standards—basically, is the punishment something most states do, or is it an outlier? Second, the Court offers its own subjective take on whether a penalty is cruel and unusual. The Court has not explicitly said whether either or both prongs are necessary or sufficient on their own.

In Kennedy, the majority held that there was a national consensus against the death penalty for child rape because only one state (Louisiana) had actually sentenced an individual to death for child rape since 1964, though six states had made child rape a capital offense (Louisiana, Georgia, Montana, Oklahoma, South Carolina, and Texas). Justice Alito’s dissent complained that the list was artificially short because more states had repealed their statutes after an earlier case had ruled the death penalty unconstitutional for ordinary rape cases.

The Court also concluded it was difficult to ensure that the death penalty would only be imposed in “severe cases” of child rape and not arbitrarily. The majority also expressed concern over the potential incentive to murder witnesses. The Court held that if the victim’s life was not taken, the death penalty could not be constitutionally imposed outside of treason, espionage, terrorism, or “drug kingpin activity”.

Alito wrote a dissent for himself and three other justices that is probably closer to what a majority of the Court believes today. He stressed that Americans probably believed that the “very worst child rapists . . . are the epitome of moral depravity,” suggesting it would not violate standards of decency to execute them.

What has changed since 2008? The composition of the Supreme Court. With a more conservative court, it is reasonable to think the 5-4 outcome of Kennedy could be overturned.

The overturning of Roe v. Wade in 2022 has set off a series of attempts by states to pass laws similar to those previously ruled unconstitutional by the Supreme Court in hopes of spurring reconsideration of those precedents. (This dynamic should be familiar to readers who have kept up with the Arkansas Ten Commandments statute, which mirrors the one ruled unconstitutional in Stone v. Graham in 1980.)

The capital rape statutes seem related. Florida in 2023 set off a wave of conservative states adopting capital rape statutes: Tennessee in 2024, and Arkansas, Idaho, and Oklahoma in 2025. Also in 2025, the attorneys general of fourteen states (all very much red states, with the exception of Virginia, which merely had a Republican attorney general at that time) sent a letter to then-Attorney General Pam Bondi asking the Trump administration to support state efforts to challenge Kennedy.

One legislative finding (basically, non-binding language at the beginning of a bill expressing the intent/will of the legislature) in Arkansas’s version specifically says: “the outcome in Kennedy v. Louisiana may have been different if more states had statutes providing that child rapists could receive the death penalty.” That same language appears verbatim in Florida’s legislative findings.

One strange point that I’d love more information on: if all these states are acting in a coordinated fashion to get the Supreme Court to overturn Kennedy, why do the statutes themselves work in significantly different ways?

Testing Arkansas’s Version

Arkansas’s statute created a new crime, defined as capital rape. The statute defines “capital rape” in a fairly complicated way. The first requisite element is engaging in “sexual intercourse or deviate sexual activity” (the latter is basically oral or anal sex) with a person 13 years or younger.

Then, the perpetrator needs to have at least one of four aggravating factors:

• Serious physical injury or threat thereof;

• Committing one of thirteen enumerated felonies “during the course of or in furtherance of” the sexual intercourse/deviate sexual activity (we’ll return to this one, but it includes things like residential or commercial burglary and controlled-substance delivery);

• Using or threatening to use a deadly weapon (even if not used to coerce sex);

• Prior rape conviction (including statutory or non-capital rape).

The punishment part of the statute takes the death penalty off the table for a perpetrator under the age of 18, imposing instead life in prison with the possibility of parole after 20 years. A perpetrator over 18 is punished with death or life without parole.

A careful lawmaker might look at that list of aggravating factors and consider: what is the “least bad” set of facts that would qualify as capital rape and thus carry the possibility of the death penalty? Here is what I came up with (and please keep in mind, I do think these are still terrible things, and all of these scenarios involve statutory rape in which the 13-year-old cannot meaningfully consent):

• An 18-year-old man goes to the home of a 13-year-old. They sneak into a parent’s locked bedroom and have sex there. This would constitute residential burglary (unlawful entry of a residence with intent to commit any offense punishable by imprisonment), one of the enumerated offenses that when combined with statutory rape of a 13-year-old constitutes capital rape.

• An 18-year-old man is a marijuana dealer. He has had a series of previous convictions for dealing marijuana.2 He delivers a small quantity of marijuana to a friend (say, half an ounce) who is having a party. He stays the night there, and while there has sex with a 13-year-old. Note here that the statutory text says “during the course of or in furtherance of the offense”, so the felony offense can be unrelated to the rape.

You get the idea. The point is not that these scenarios are acceptable. They are not. The point is that capital punishment doctrine is supposed to distinguish the worst cases from other serious crimes, and the Arkansas statute’s structure may not do that cleanly.

The enumerated felonies part opens up the possibility of behavior that is basically just ordinary statutory rape (still very bad!) being punishable by death because of substantively unrelated behavior.3 Consider also that if the 18-year-old in the scenarios above is, say, 14 (or even 13), he could still end up in prison for the rest of his life. That, I think, is not what the average person would think appropriate punishment for the crime.

Comparing Arkansas and Florida

As I mentioned, the various states’ versions really differ significantly state-by-state. They all apply to a subset of rapes, but they differ in what subset. And each state has seemingly random differences that make either statute significantly more or less harsh.

Florida, for example, doesn’t define a new crime; it has a list of “aggravators” (i.e., aggravating circumstances) at the penalty phase for perpetrators of sexual battery against a person 12 or younger (compare to 13 in the Arkansas statute). A perpetrator has to have at least two aggravators from a list of 11 possibilities to warrant capital punishment. Florida seems to have done a better job of ensuring only the worst of the worst qualify for the death penalty.

But Florida is also the only state that allows a death penalty recommendation on less than a unanimous vote of the jury—only 8 of 12 are needed. This has not been ruled unconstitutional outright by the Supreme Court, but pretty much—Ramos v. Louisiana in 2020 held that there must be a unanimous jury verdict on serious offenses. Florida’s law sidesteps that by applying the 8 of 12 rule at the penalty phase rather than the verdict phase.

The Alito dissent in Kennedy was focused on the idea that surely the worst of the worst child rapists could be executed without violating the Eighth Amendment. In theory, the law that would be most likely to get the Supreme Court to overturn Kennedy would apply the death penalty to only the worst of the worst child rapists. It will be intriguing to see if Arkansas or these other states have sufficiently narrowed the world of potential defendants.

Florida arguably has the better statute for getting the Supreme Court to overturn Kennedy because of things like the requirement for two aggravators rather than the one in the Arkansas statute. But its 8 of 12 jurors requirement could create a different avenue for constitutional challenge.

Concluding Thoughts

This is such a terrible topic that I mildly regret writing about it this week. But it is also important to think carefully about it and get it right. The knee-jerk reaction to criticism of this sort of statute is something like: “Why should we care about being too harsh to child rapists? They deserve whatever they get!” And, of course, they do.

But the reason to be careful is not solicitude for child rapists. If you could magically push a button and kill every terrible child rapist or would-be child rapist with absolute certainty of guilt and no marginal cases like the ones described in this post, I think most people would push that button, including me.

But the actual legal definitions and the system matter tremendously in determining whether a real-world capital punishment statute is a good idea. There are, of course, people who object to the death penalty on philosophical grounds unrelated to the practical problems of its administration (i.e., even if we could determine guilt perfectly and painlessly administer the penalty, some would still not support it). I think almost everyone else has some sliding practical scale. If they had the choice, few would support the death penalty in Putinist Russia, where the legal system is very unlikely to determine guilt correctly and authorities will absolutely abuse the power to kill.

One of the many terrible parts of this area of the law is a bad statute can get children killed by being too lenient or by being too harsh. A law that is too lenient does not deter perpetrators. A law that is too harsh means that the perpetrator has nothing to lose by going further.

I think at a minimum, the state legislatures involved, including Arkansas’s, should conduct and/or pay for a more in-depth empirical investigation of the effects of these laws. Does increased deterrence outweigh the increased incentive to kill witnesses? That is not (just) a philosophical question—it is susceptible to scientific study.

It is certainly not a good idea to rush a law like this just to take advantage of a conservative Supreme Court bench. I would feel a lot more comfortable if the legislative findings were focused on real-world effects rather than the state of constitutional jurisprudence.

Subscribe now

Subscribe now

Act 975: Struck Down!

First, a little more of a victory lap note regarding the Arkansas Supreme Court unanimously striking down Act 975, which assigned original jurisdiction to the Arkansas Court of Appeals for facial constitutional challenges to state laws. I had written a few posts on the Act’s unconstitutionality earlier this year, as well as the general problems with creating a parallel court system for politically sensitive cases.

I don’t have a lot to add substantively beyond what I wrote in my earlier posts, which outlined the various arguments that ultimately proved persuasive to the Court. But the fact that this was a unanimous decision was heartening and speaks to the resilience of the state’s legal system in the face of partisan pressure.

The Franklin County Prison Project: Seemingly Abandoned!

I had written on the predictable problems that arose after the state’s secretive effort to place a big new prison in a rural, seemingly politically weak county (Franklin County). I recommended that the state start over from scratch with a more inclusive process in recognition of the law and economics problems of new prisons.

In the face of legislative determination to kill the Franklin County Prison Project, Governor Sarah Huckabee Sanders’s office announced that it was exploring other alternatives.

In my experience, the second the door is opened a crack to abandoning this sort of effort, the project is kaput. The only way the governor was going to get this project done in Franklin at this point was the Trumpian approach of using all the power at her disposal to threaten and crush dissenters. That still might have failed, of course—it often fails for Trump in his own legislative initiatives, whether Obamacare repeal (recall my old boss’s last big moment) or the effort to nationalize election administration.

There is a tendency to glamorize politicians who never give up, as if it is more principled to eternally fail rather than to figure out a way forward. But I like to think this is the rare case of a politician realizing that a project is a sunk cost and changing course, hopefully using the experience to do better the next time. We should applaud when politicians stop pursuing a bad idea to incentivize them to do so more often. So, kudos to Governor Sanders on this one!

The Iran War: Still Not Legal!

I wrote at the beginning of the war that it was probably not legal (the President couldn’t initiate hostilities without Congress) and that there didn’t seem to be a clear strategic objective.

The war remains a vast strategic waste. U.S. intelligence reportedly assesses that Iran’s breakout time to acquiring a nuclear weapon has not changed as a result of the conflict. No one is under any illusions about how long it will take Iran to rearm its missile and drone supply after the conflict. We have seemingly given up on the idea that we will demand that Iran stop arming militias in neighboring countries.

Iran has shut down trade through the Strait of Hormuz, harming the U.S. and world economy. We have set up a retaliatory blockade in hopes of driving Iran to the negotiating table, though it is not clear whether we or they are more desperate to open things back up. Every few days, there’s a new initiative that seemingly goes nowhere. Remember Project Freedom, the escorting of tankers and merchant ships, that started earlier this week and was called off reportedly because the Saudis refused to allow use of our bases on their territory for it?

The legal state of the war has also turned screwy. The War Powers Resolution requires the President to withdraw U.S. military forces engaged in hostilities or the imminent threat of hostilities after 60 days if Congress hasn’t authorized military action or granted an extension. The Trump Administration has argued that the timer on the 60-day clock pauses when there is a ceasefire.

This is nonsense.

Here is a quick primer on the structure of the War Powers Resolution. Section 4 has a requirement that the President submit a report in specific circumstances, the most relevant to this situation being:

in any case in which United States Armed Forces are introduced--

(1) into hostilities or into situations where imminent involvement in hostilities is clearly indicated by the circumstances; [or]

(2) into the territory, airspace or waters of a foreign nation, while equipped for combat, except for deployments which relate solely to supply, replacement, repair, or training of such forces[.]

Then section 5(b) requires the President “terminate any use of United States Armed Forces with respect to which such report was submitted (or required to be submitted)” after 60 days unless Congress approves hostilities or an extension. So, the thing that triggers the 60 day clock is when the President had to file the report, and he had to file the report when he introduced U.S. Armed Forces into a situation where hostilities were “imminent.” After 60 days, U.S. forces must be withdrawn from the situation that triggered the reporting requirement.

Obviously, the current Iran situation fits that. Even though there is a nominal ceasefire, U.S. forces are deployed for combat and are routinely engaging in it. U.S. forces have been in a situation that triggers the reporting requirement for the entire war. They have been in actual and imminent hostilities the entire time. Had they been withdrawn, that might arguably stop the clock entirely, and then trigger a new 60-day limit when they are reintroduced. But by the clear letter of the law, the 60-day limit doesn’t pause.

If you think about it, there is no way to create a law like this that would incorporate ceasefires. If no one is firing at a particular moment, does the clock pause? Is it only when a ceasefire is officially agreed on all sides? What about a situation like Ukraine, which recently had a three-day ceasefire to commemorate Victory Day (the end of World War II)?

The Arkansas Anti-Sharia Bill: Voted Down in Committee!

Not only was the weird anti-sharia bill shot down in committee, but at least some representatives who opined on it made similar arguments to mine!1

GOP Rep. Jeremiah Moore of Clarendon said he had concerns the proposal would not apply to a white supremacist compound in northeast Arkansas, and that membership agreements for hunting clubs could be affected.

“If we’re trying to worry about disclosures for Sharia Law, I would also like to cover the white supremacists up near Ravenden,” Moore said. “I’m pretty sure they’re trying to do the exact thing that you’re describing here, which is you purchase into an LLC or to some sort of corporation that owns the actual real estate, and you’re bound by their laws or their rules and regulations.”

I will be curious to see if this one comes back in some form, but for now, this is good news!

Subscribe now

Subscribe now

If Rule of Law is So Good, Why is it Rare?

I teach a course on law and economics, which was one of my favorite courses in law school. The best part is ignoring all the moral or political baggage around a particular issue and just looking at it in terms of incentives. In my classes, I will sometimes do this by asking students to “put on their Machiavellian hats” or by formally ordering morality to leave the room. Morality and politics are still important, of course, but we hear about those all the time. Sometimes you can get a more interesting insight by looking past what we think is good or fair.

The rule of law is always presented as an obviously good thing—why would anyone want a system that is not fair and impartial, subjecting the powerful and powerless to the same rules? Don’t people get that if you have corruption, then you can’t have investment, innovation, contracts, and all those great things that enable people to create wealth?

There’s a famous old economics joke. Two economists are walking down the street. One of them spots a $20 bill on the sidewalk and says, “Hey, look — a twenty-dollar bill!”

The other replies, “Don’t be ridiculous. If there were really a $20 bill on the ground, someone would have picked it up already.”

The rule of law seems like a $20 bill sitting on the ground, but almost no civilizations in human history (or even today!) actually had the rule of law as we understand it. Were they stupid?

Evolution of the Rule of Law

Step 1: Kinship

Humans aren’t the only species that cooperates or has some kind of social order. It is not surprising that as best we can tell, human order started like the animal order we know of, with evolutionarily beneficial kindness toward family and close genetic relatives.

Humans broadened that to clans and kinship networks. Kinship networks work in low-trust environments because you can’t really cheat a family member in the long-run. You know all the same people, and you can’t work with them if everyone knows you’re a cheat or a criminal. Afghanistan is the quintessential clan-based society.

The downside to kinship as an organizing principle is you are committed to working only with a relatively small group of people—certainly no more than a few hundred. There is little opportunity for specialization of labor, long distance trade, etc. It is no coincidence that clan/kinship societies are among the poorest on Earth.

Step 2: Patronage

The next step from there is patronage loyalty. Think the Sopranos, or Putin’s Russia. Big strong man collects and redistributes resources in exchange for loyalty. No strictly observed rules necessary beyond loyalty.

Patronage tends to emerge as the first alternative to clans because it preserves the cognitive and emotional structure of kinship while detaching it from biology. It's still personal, face-to-face, but just voluntarily entered. A society can extend cooperative networks beyond the clan without developing the cultural and institutional machinery to trust strangers as such.

From an incentives standpoint, patronage starts emerging when economic specialization and long-distance trade create gains from cooperation that exceed what any kin group can capture internally. Individuals start defecting toward whoever can broker access to wider networks.

Why do the Afghanistans of the world remain kinship-based? Pastoral economies and mountainous or arid terrain reward mobile, defensible kin groups and punish centralization.

Step 3: Rule by Law

Why does patronage fail? Basically, scale and the principal-agent problem. A Tony Soprano who runs his patronage network through vassals can monitor them directly. But a king who governs a province through clients can’t monitor his clients’ clients. Each layer of delegation introduces slack, and personal loyalty doesn't transmit through the chain. On top of that, if you’re the emperor of the Roman Empire at its zenith, you won’t even know there’s a rebellion in the extremes of the Empire for a month.

You can muddle your way through this problem for a while if you have a superlative army. You just have to massacre anyone who rebels. But no military advantage is permanent, and eventually you can’t hold everyone in line just with fear.

In come rules! You establish a known set of rules that everyone plays by so you don’t have to directly monitor your clients as closely. Think law codes, written procedures, audits, examinations, standardized titles and tax rates.

The crucial thing here is that you still don’t have rule of law—you have rule by law. The king is still not subject to the laws himself. The laws are strictly a way for the king to maximize the efficiency of his empire.

Rule by law is sometimes stable, sometimes not. China has had rule by law for something like 2,000 years. The Soviet Union had rule by law for perhaps 40-ish years. Medieval England had it for several hundred, starting when William the Conqueror’s successors set up common law courts and ending certainly no later than the 1688 Glorious Revolution.

The upside of rule by law is increased ability to trust and trade with strangers throughout the empire. The downside is it can still be unpredictable—the king can always change the rules. That tends to chill investment and innovation.

If only there were a way to credibly bind rulers, it would spur greater investment and greater innovation! That system would then have an advantage over all competing systems, and slowly (or quickly) conquer or coopt them.

Step 4: Rule of Law

To answer the question from the beginning (why doesn’t everyone have rule of law): it is difficult for a ruler to bind themselves. Even if they voluntarily accept restrictions, every ruler dies. When they die, their successor may not be so noble—regression to the mean is a real pill.

So, how do you get rule of law? You need the ruler to bind himself, and a legal structure of some kind that can keep him bound. Depending on who you listen to, that may have only actually really happened once, in England in 1688. There, you had a fluke coincidence of: (1) a historically strong Parliament, (2) Catholic doctrine having weakened clanship networks and established laws that kings were supposed to honor, with real consequences if they did not, (3) a civil war that Parliament won, and (4) trusted legal institutions with deep roots.

But rule of law proved to be a huge advantage. The English system spread through revolution, conquest, and imitation—the United States, France, etc. The spread was easier in places that had almost but not quite achieved rule of law—much of western Europe, Japan, South Korea, etc. It hasn’t spread everywhere, but its success is undeniable.

Why Permanent Victory for Rule of Law is Impossible

Once the benefits of rule of law have been realized, the incentives of rulers change somewhat. The best of all worlds would be to get the investment and innovation gains from rule of law, but secretly (or not so secretly) subject yourself to a different set of rules.

The benefits of actual rule of law accrue to society writ-large, while the benefits of rule by law or patronage accrue to the rulers. This is a conflict of interest for rulers, and one of the ways it can be resolved is for a society to backslide. This tendency can get bad enough to knock a society back to patronage. One of the most difficult issues of constitutional government is how to avoid reverting back to the rule of patronage—re-patrimonializing, as Francis Fukuyama put it.

And in a way, this backsliding is to be expected. Rule of law is the most unnatural thing humans have ever built. Every instinct we inherited from evolution is to favor blood, to reward loyalty, to protect our own.

So, it isn’t just the usual rhetoric to say rule of law depends on everyone living by principles. We need to demand it of our rulers, or they might just stop giving it to us.

Subscribe now

I’ve written and said a lot about Arkansas’s Ten Commandments law, so I’m going to discuss this major new ruling this week.

First, I’m going to first briefly discuss what the Fifth Circuit ruling portends for Arkansas’s Ten Commandments statute. In a nutshell, I think there is serious reason to think when the Eighth Circuit (eventually) rehears the Arkansas Ten Commandments case en banc, it will uphold the permanent injunction against Arkansas’s statute.

Second, I’m going to complain about the Fifth Circuit majority opinion. There are enough problems with the opinion that I suspect the Supreme Court will have to blaze a new doctrinal trail if it wants to uphold these Ten Commandments statutes rather than simply go along with the Fifth Circuit’s argument.

So, basically, part 1 is going to be all about amoral reading of the tea leaves from a political perspective. Part 2 is going to be all about the actual substance of the ruling.

Subscribe now

Appointing Party vs. Appointing President

When I was a Hill staffer, I worked a lot on judicial confirmations and the partisan split of the various circuits. If we use the party of the appointing president as a proxy for ideology, the Fifth and Eighth Circuits are the two most conservative circuits, and the Eighth is the most conservative by a wide margin.

Counting only active circuit judges, the Eighth Circuit has 10 Republican-appointed judges and only 1 Democrat-appointee—91% Republican.

The Fifth Circuit has 12 Republican appointees, 5 Democrat—71% Republican.

The Fifth Circuit decision on the Ten Commandments was 9-8, with all Democrat appointees and three Republican appointees voting to strike down the Texas statute.

You might look at that math and think the Ten Commandments case is a slam dunk in the Eighth Circuit. After all, the split is 10-1 in favor of Republican appointees, and a 12-5 circuit just upheld!

But I looked at enough circuit judge nominations to know that Republican judges aren’t homogeneous. And you only need to look at the Fifth Circuit breakdown to see why. Every Trump appointee voted with the majority in the Fifth Circuit. All but one George W. Bush appointee voted with the dissent.

The breakdown of the ten Republican judges on the Eighth Circuit is: 1 judge appointed by George H.W. Bush, 5 judges appointed by George W. Bush, 4 judges appointed by Trump.

Toss in another wildcard: senior judges (i.e., non-active judges) can be appointed to the original three-judge panel in a circuit case. If they are, they typically get to participate in the rehearing en banc. There are enough Democrat-appointed senior judges that it could further swing the en banc court.

So, if the Eighth Circuit follows the Fifth Circuit by appointing president, the 10-1 Republican advantage actually turns into a 6 Bush (H.W. and W.) judges vs. 4 Trump and 1 Obama. That split could very plausibly yield a narrow decision going the other way—say, a 7-4 or 6-5 outcome in favor of striking down the statute.

Why do Republican Judges from Different Presidents Rule Differently?

I am generally sympathetic to the argument that the Bush-era judges were selected for different things than Trump-era judges, but I think there is a less arguable reason the Trump-era judges tend to rule in a more partisan fashion: The Trump-era judges are still of an age where it is entirely plausible that they could be appointed the Supreme Court. The Bush-era judges are not.

I hasten to add, the Bush-era judges are obviously young enough that they can physically do the job, but the obvious game theory of Supreme Court appointments is to put justices on the Court relatively young so they can be there for decades and decades.

Going back to 1975, Supreme Court appointees have averaged 52 years old when appointed. The youngest? Clarence Thomas, at 43 in 1991. The oldest? Ruth Bader Ginsburg, at 60 in 1993.

There were obviously substantive reasons for Clinton to appoint Ginsburg, but from an amoral strategic standpoint it was foolhardy. Baldly put, if Clinton appointed a 50-year old instead of a 60-year old, odds are quite good that Roe v. Wade would still be the law of the land today.

So, any president listening to his or her lawyers will from now on appoint judges in their low 50s at the absolute oldest. That means the Bush-era judges no longer need to worry about pleasing a conservative president who might bump them up to the Supreme Court. But Trump appointees absolutely do, which presumably explains why many do silly things to try to capture the attention of Trump. And the last thing they would do is rule against conservatives on a hot-button culture war issue like the Ten Commandments in public schools.

The Complaining Portion of Our Program

Enough politics. Let’s talk about the law of the Fifth Circuit’s ruling.

The majority ruled that the Texas Ten Commandments statute violated neither the Establishment Clause nor the Free Exercise Clause. I am going to spend more time on the Free Exercise Clause portion, but I’ll briefly address the Establishment Clause. If you want more background, see my previous post on the subject.

On the Establishment Clause, the majority’s new test is whether a statute resembles a founding-era establishment of religion—i.e., an official endorsement of a particular religion marked by legal control of church doctrine and personnel, compulsory attendance, religious taxes, etc. The majority said, in essence, that Ten Commandments posters in public schools are nothing like those establishments, so they must not violate the Establishment Clause.

There are two big arguments against that line of reasoning.

First, the Establishment Clause does not say “Congress shall make no law establishing a religion,” it says “Congress shall make no law respecting an establishment of religion[.]” So, right off the bat, you know that the Establishment Clause means more than just outlawing establishments. The majority opinion acknowledges this point, but never really addresses it in substance.

Second, the majority relies heavily on the idea that the clearest evidence of what “establishment” is can be inferred by what some states specifically got rid of in order to disestablish their churches. This approach is subject to the same limitation all modern religion clause cases face: Governments in general did so much less during the founding era that disestablishment meant a fairly limited list of things. It was not the theoretical limit of what an establishment could be, but rather reflected the specific things that states knew created establishment problems.

Free Exercise and Mahmoud v. Taylor

The free exercise side of the Ten Commandments cases is much more fluid and novel than the establishment side. The Supreme Court has narrowed Establishment Clause doctrine to the point that it would be mildly surprising if the Court does not ultimately overturn Stone v. Graham. (To be clear, I don’t think they should, but I think the conservative majority of the Court has a much more fixed view of what the Establishment Clause means.) But the Court has at the same time drastically expanded the Free Exercise Clause, so much so that it might effectively take over issues previously in the realm of the Establishment Clause.

The Court’s recent free exercise jurisprudence has opened the door to striking down the Ten Commandments displays, albeit almost certainly unintentionally. The key case here is from last year, Mahmoud v. Taylor. In Mahmoud, the Court ruled that a public school district violated the Free Exercise Clause by assigning pro-LGBT readings to students without an option for parents with religious objections to opt their children out of the assignments.

In my ideal world, I would say Mahmoud was wrongly decided, but the Ten Commandments displays still violate the Free Exercise Clause. But Mahmoud has set up a standard that is difficult for courts to distinguish from Ten Commandments displays. Thus, the really central free exercise issue in this recent Fifth Circuit case is how (and whether) Mahmoud does not control the outcome of this case.

The Fifth Circuit majority argued that Mahmoud involved coercion, and this case did not. In Mahmoud, the school district required students to read the material, and gave guidance to teachers on how to handle questions involving the lesson. Here, the Fifth Circuit held that the Ten Commandments are just a poster on the wall. The poster is not part of a lesson, and teachers are not directed to say anything about it. Hence, no coercion.

If this seems like a superficial reading of Mahmoud, well, it was. Under Mahmoud, mere “exposure” is not the end of the story. It is also relevant how old the children are, the setting, the message conveyed, and whether the state-imposed material creates pressure to conform.

The walls of an elementary school are a pedagogical space. They are used to list class rules, honor codes, patriotic symbols, mandatory lessons (think multiplication tables or the alphabet). I think the natural assumption of any student seeing something posted on the wall is some combination of: “This is true” and “My teacher wants me to know this.” But even putting that aside, the coercion problem is obvious the second you consider what happens when a student asks a question about the posters.

Leaving Coercion Up to the Teacher

Here are some possible answers and the problems with them:

• “They are the word of God transmitted through Moses, and the way we should live our lives.” Obviously, that would be the very religious instruction the Fifth Circuit says Texas does not authorize.

• “These are historically important rules that may or may not actually be from God.” The teacher has supplied a secularizing gloss that the statute itself does not include and that instill doubt as to the religious meaning that many families attach to the Ten Commandments. We are right back where the Fifth Circuit said Mahmoud put parents.

• “Ask your parents.” The state has pressured children to have a specific religious conversation with their parents. Could the state put up a poster that just says, “Ask your parents about Jesus Christ”?

• “Different people believe different things.” The teacher has turned the display into comparative-religion instruction without curriculum, training, standards, or parental notice.

• If the teacher refuses to answer, the refusal itself tells students that the state has placed a religious command in the room that the teacher cannot explain. The teacher also sounds bizarre in this scenario—why is there something the classroom they won’t talk about?

There is no good solution to this problem. That’s a consequence of the state trying to influence the religious views of children, but leaving it up to teachers to deliver the coercion in a way that is difficult for courts to review.

Misreading Mahmoud

One of the ways in which the Fifth Circuit tried to distinguish the Ten Commandments from Mahmoud was by misconstruing Mahmoud. The Fifth Circuit majority incorrectly cited to Mahmoud for the proposition that teachers in that district “were to inform students that their religious views on these topics were ‘hurtful, perhaps even hateful’ and that their parents’ views were wrong.”

If it sounded strange to you that a teacher would be directed to tell a child that they were “hateful,” don’t worry, the Fifth Circuit majority opinion hallucinated this. They cited to the majority opinion in Mahmoud, and here is the relevant cited portion:

The book [assigned by the school district] goes so far as to present a contrary view as something to be reprimanded. When the main character’s brother says “ ‘You can’t become a boy. You have to be born one,’ ” his mother corrects him by saying: “ ‘Not everything needs to make sense. This is about love.’ ” [Citation omitted]. The upshot is that it is hurtful, perhaps even hateful, to hold the view that gender is inextricably bound with biological sex.

So, the Fifth Circuit quoted the Supreme Court’s incendiary, defensive upshot of the book, then put that quote into the mouths of teachers (or, honoring the weird passive voice used by the court, the mouths of the administrators who supposedly told the teachers to call children hateful.)

Elsewhere in Mahmoud (not the part cited by the Fifth Circuit majority), the Court described where the “hurtful” language came from in the teacher guidance:

If a student claims that a character “can’t be a boy if he was born a girl,” teachers were encouraged to respond: “That comment is hurtful.”

Note that this guidance isn’t about declaring beliefs to be hurtful (let alone hateful). It is about a comment. Presumably, if a Christian schoolchild followed around a Jewish schoolchild and kept saying, “You won’t get eternal salvation if you don’t convert,” that would be a hurtful comment. It would not be an indictment of the Christian child’s beliefs to tell him that the comment is hurtful.

Next Steps

The Eighth Circuit has not scheduled oral arguments for the appeal of the permanent injunction against the Ten Commandments posters. It may well be some months before the case is heard and then, probably, reheard en banc. And then the Supreme Court will almost certainly weigh in regardless of how the Eighth Circuit rules. Stay tuned!

I wouldn’t worry about that. There wouldn’t be much point to a draft other than providing more ground soldiers. It’s entirely possible fighting could flare up again (as I write this there is a temporary ceasefire that seems to be teetering), but the war has involved virtually no soldiers on the ground. One big reason for that is the war is already unpopular, and ground soldiers would substantially increase U.S. casualties. And, of course, the politics of a draft are dicey to begin with.

But automatic draft registration might be more about aggregating and centralizing data for immigration enforcement purposes rather than a national security measure. Understanding why requires a look into the Selective Service System (SSS). Basically, the SSS might be more useful as a way for the federal government to aggregate data than for implementing a draft.

I actually did an interview for the local CBS affiliate Channel 5 last week on this, so I figured this week I would go into more detail on it. As a side bonus, I’ll also talk a little about why women will probably have to be included in the automatic registration soon.

Subscribe now

Draft Logistics

The case for automatic draft registration is pretty simple, and without any of the current political context, it would be all you’d need to know to support automatic draft registration. Under current law, men eligible for the draft (essentially all men who are U.S. citizens, permanent residents, or undocumented immigrants, aged 18-25) must register. Failing to do so is a crime. Coaxing men to register through advertising costs money, upwards of $30 million per year.

But if the federal government could just know where all the men in the country aged 18-25 were, the SSS could register those men themselves. No more advertising!

The tricky part I glided over just now was how the federal government could know where all the men are. That’s one of the big reasons to have voluntary registration—the men self-report their address and contact information. But if the federal government is going to do it on their own, they need the address data from somewhere.

The first big place SSS can get data from is the Social Security Administration (SSA). Anyone born in the country has a Social Security Number, and while the SSA doesn’t have their current address, it can pinpoint when they were born. Social Security data thus allows the SSS to figure out roughly who should be registered. The big exception here is undocumented immigrants—they do not have Social Security Numbers.

The best place for address data is state DMVs. Virtually all of these men aged 18-25 are going to want to drive, so they have to go to the DMV and report a current address. Most state DMVs are the way in which men register for the draft. When men come in to get a license, they fill out the form, and the DMV sends the information to SSS. The limitation here though, is that states usually only do that once for the initial registration. They don’t usually keep the address updated.

In theory, the federal government could use the SSS as a tool for the immigration enforcement raids that captured the country’s attention in 2025 and early 2026. Here’s the idea:

You have a list of (essentially) all the men in the country aged 18-25 with DMV data. You have a list of all the U.S. citizens of that age with the SSA data. You just find who is on the DMV list and not the SSA list, and that should logically be all the undocumented immigrants of that age cohort.

Two things make it difficult to do that: (1) if the information SSS has is not in an easily useable database, and (2) if SSS can’t share data easily with ICE and other immigration authorities.

One-Stop for Data + Easier Access for Other Agencies = More Enforcement

The automatic draft registration statute (part of the 2026 National Defense Authorization Act) doesn’t necessarily change the sources of data, but it allows SSS to smoothly integrate them. SSS has stated that it plans automated federal data feeds, matching algorithms, and elimination of “information silos.”

That’s the useable database part sorted. Automatic draft registration essentially requires SSS to become a giant, smooth database.

The other part was not really noticed by media outlets, but was noticed in data privacy circles. The SSS put a notice in the Federal Register one day before President Trump signed the NDAA updating its list of routine uses for the data in its possession. A “routine use” basically means the agency would not need to go through a cumbersome process to share the data. Those routine uses now include, across the various SSS databases, federal crime enforcement, disclosure for student aid eligibility, domestic and foreign law enforcement, and employment checks.1 All of those could be relevant for immigration enforcement purposes.

The potential for SSS to serve as a database to help hunt down undocumented immigrants has not gone unnoticed. Several contributors to Project 2025, including the current head of the “Mass Deportation Coalition”, have called for using SSS in this way.

How should we think about this situation? You can come to your own conclusions, but here are the factors I think are most relevant:

• Hijacking a national security institution for domestic purposes is pretty generally a bad idea.

• Creating a giant federal database carries risks. Foreign governments could hack or alter the data. Political groups could use the data for campaigning or to track disfavored people. (This was the standard conservative argument against federal aggregation of U.S. citizen data before the Trump era.)

• Given that the Trump Administration has already been doing seemingly everything in its power to target political opponents, I especially would not want such a database created now.

• I oppose chaotic, violent ICE raids to round up and deport non-violent undocumented immigrants. To the extent transforming the SSS into a database facilitates increasing those raids, I don’t support it.

• Automatic draft registration probably makes sense in the abstract, but it could be implemented in a way that completely avoids the dangers discussed above: simply don’t let the data be used for immigration purposes (i.e., remove the relevant routine uses).

Bonus: Why Women Will Likely Soon Enter the Draft

The Supreme Court ruled in 1981 in Rostker v. Goldberg that excluding women from the draft was constitutional not because of traditional gender roles, but because the Court found that combat restrictions on women meant that for purposes of the draft, men and women were not similarly situated.

Since then, the factual predicate of that holding has fallen apart. Since 2015, the Department of Defense has permitted women to serve in any combat role open to men. So far, the Trump Administration has not changed that. Defense Secretary Hegseth has been proposing changing fitness standards to de facto exclude more women, but that would not have legal relevance over whether physically able women are not registered.2

If DOD does not consider men and women differently situated for roles in combat, then Rostker seems to indicate that the men-only draft is doomed.

The Supreme Court has itself suggested as much. In 2021, the Supreme Court denied certiorari on a case challenging the constitutionality excluding women. In the denial of certiorari, three justices (including Justice Kavanaugh) wrote to emphasize that the Court was giving Congress time to change the law on its own. But that effort seems moribund, so the next time a case comes before the Court, it is unlikely the Court will deny certiorari.