The Founders Built a Machine for Humans They Never Actually Met

Let's be clear about something before we start: this isn't a takedown. The men who designed the American constitutional system were operating at a genuinely extraordinary intellectual level, synthesizing Enlightenment philosophy, classical history, and hard-won political experience into something that had never existed before. The framework they built has lasted nearly two and a half centuries, which is not nothing.

But they were also working without behavioral science. Without polling. Without any systematic way to study how actual human beings — not the idealized rational actors of Enlightenment theory — behave when you put them inside a political system and watch what happens.

They had a model of human nature. It was sophisticated for its time. And on several of the most important questions, it was wrong in ways that became apparent almost immediately. The gap between their predictions and actual outcomes is one of the most instructive psychological datasets we have.

Madison's Great Fear, and Why It Didn't Materialize the Way He Expected

James Madison was obsessed with factions. Federalist No. 10 — arguably the most important single document in American political theory — is essentially a long argument about why the Constitution's design would prevent any single faction from permanently seizing power. Madison believed that a large republic with many competing interests would produce a kind of natural equilibrium. Groups would check each other. No one coalition would grow large enough to dominate.

He was specifically worried about a faction of the poor majority using democratic power to redistribute wealth from the propertied minority. This is the faction he most wanted the system to neutralize.

Within a decade of ratification, something happened that Madison had not predicted: the factions didn't multiply and balance each other. They consolidated. By 1800, American politics had organized itself into two broad opposing parties — the Federalists and the Democratic-Republicans — with stable coalitions, partisan newspapers, coordinated electoral strategies, and something that looked remarkably like the permanent factional dominance Madison had designed the whole system to prevent.

Madison himself ended up leading one of these parties. He helped build the very thing he'd theorized against.

From a behavioral science perspective, this makes complete sense. Humans are intensely tribal. Coalition-building is one of our deepest social instincts. The idea that competing interests would remain fragmented and mutually checking assumes a level of sustained independence from social pressure that simply doesn't describe how people actually behave in groups. We sort. We affiliate. We build teams. A system designed around the assumption that we wouldn't do this was always going to be surprised.

Hamilton's Bet on Elite Disinterest

Alexander Hamilton believed — sincerely, not cynically — that if you filled government with educated, propertied men of standing, you would get disinterested public service. The assumption wasn't that elites were morally superior. It was that men who already had wealth and status had less to gain from corruption and more to lose from a damaged reputation.

This was a reasonable hypothesis. It was also wrong almost immediately.

The speculation scandals of the 1790s, in which Treasury insiders and their associates profited handsomely from policies Hamilton himself designed, demonstrated something behavioral economists would later formalize: people with access to resources and influence don't become less motivated to accumulate more. They become more so. The psychological research on this is now robust — wealth and power tend to reduce empathy and increase risk tolerance in ways that make elite corruption more likely, not less.

Hamilton wasn't naive. But he was working from a model of human motivation that treated self-interest as something that could be satisfied and set aside. Modern psychology suggests it doesn't really work that way. Self-interest is generative. It expands to fill available space.

Jefferson's Vision of the Rational Yeoman Voter

Thomas Jefferson's political philosophy rested on a specific human type: the independent farmer. Economically self-sufficient, educated enough to reason from first principles, with no employer to pressure him and no debt to make him desperate. This citizen would evaluate candidates and policies on their merits and vote accordingly.

Jefferson's faith in this figure was so strong that he opposed institutions — banks, cities, manufacturing — that he believed would create dependency and corrupt the independence of the citizen body.

What actually happened, within Jefferson's own lifetime, was the rapid expansion of exactly the conditions he feared. Urbanization. Wage labor. Financial dependency. And with those conditions came political behavior that looked nothing like the rational deliberation Jefferson had imagined. Voters responded to charisma, to tribal loyalty, to economic anxiety, to rumor and slander and theatrical spectacle. The 1828 presidential campaign between John Quincy Adams and Andrew Jackson — which Jefferson lived just long enough to see approaching — was arguably the first modern negative-campaign election, full of personal attacks, emotional manipulation, and organized get-out-the-vote machinery.

The rational actor model of political behavior that Jefferson assumed would characterize American democracy had a lifespan of roughly forty years before it was visibly obsolete.

This isn't surprising if you've read the behavioral economics literature. Daniel Kahneman's distinction between System 1 (fast, emotional, associative) and System 2 (slow, deliberate, rational) thinking maps almost perfectly onto the gap between Jefferson's expected voter and the actual voter. People vote with System 1 the vast majority of the time. Jefferson designed for System 2.

What They Got Right (And Why That's Also Instructive)

To be fair — and the historical record demands fairness — the Founders got several things profoundly right, and those successes are also psychologically instructive.

Madison's instinct that you need structural friction in government to slow down impulsive majorities was correct, even if his mechanism (faction multiplication) didn't work as planned. The separation of powers, whatever its frustrations, has repeatedly slowed the conversion of temporary popular passion into permanent authoritarian change. That's the system working as intended — just through different mechanisms than Madison drew up.

And Hamilton's insistence on a strong federal government with real financial credibility turned out to be correct in ways that mattered enormously for long-run stability. His model of human economic behavior was flawed. His institutional instincts were sound.

The Experiment We're Still Running

What makes the Founders' predictions so valuable as a psychological dataset is their specificity. These weren't vague aspirations. Madison, Hamilton, and Jefferson wrote down, in extraordinary detail, what they expected people to do. And then three million people went ahead and did something else.

No IRB-approved lab study gives us that. No survey of undergraduates runs for two hundred and thirty-five years. The American founding is a documented, large-scale natural experiment in the gap between theoretical human nature and actual human nature — and the results are sitting in the Library of Congress, available to anyone who wants to look.

The Founders weren't failed prophets. They were brilliant theorists who built the best model they could without the tools that would have made it better. The grace note here is that they knew this was possible. Madison wrote in Federalist No. 51 that the system had to account for the fact that men are not angels. He just underestimated, in a few key places, exactly how un-angelic the specifics would turn out to be.

We're still living inside their experiment. It still tells us something new every election cycle.