Approval Ratings Are Not Democracy — Rome Learned This the Hard Way

Every morning, somewhere in Washington, a political consultant opens a dashboard. It shows approval ratings broken down by demographic, issue-by-issue favorability, and sentiment trending across social platforms in near real-time. The consultant uses this information to advise their client — a senator, a governor, a president — on what positions to take, what language to use, and crucially, what decisions to avoid.

This is considered sophisticated political operation. It is also, from a historical perspective, one of the more reliable ways to slowly destroy a functioning democracy.

What Polls Actually Measure

Before getting into Rome, it's worth being precise about what polling captures and what it doesn't. A well-constructed poll tells you what a representative sample of people say they think about a question, at a specific moment, as the question was framed by the people writing the survey.

That's genuinely useful information. It is not, however, consent. It is not deliberation. It is not the kind of considered judgment that democratic theory — from Locke through Madison through every serious political philosopher who's thought carefully about self-governance — identifies as the actual foundation of legitimate democratic decision-making.

The confusion between "people say they currently like this" and "this has democratic legitimacy" is not a new mistake. It's an ancient one. And the historical record of where it leads is not encouraging.

The Roman Senate's Approval Problem

The late Roman Republic was, among other things, a sophisticated popularity-tracking operation. Roman politicians employed delatores — informers and observers — to gauge public sentiment in the Forum. They monitored crowd reactions at games and public events. They understood, with considerable precision, which positions played well with which segments of the Roman public.

They used this information to make decisions. Specifically, they used it to avoid making decisions that would be unpopular, even when those decisions were obviously necessary.

The grain supply is the clearest example. Rome's food security required constant, unglamorous administrative work: maintaining relationships with provincial suppliers, investing in storage infrastructure, managing the complex logistics of moving grain across the Mediterranean. None of that was popular. It didn't generate applause in the Forum. It didn't translate into favorable crowd reactions at the Colosseum.

What did generate applause was the annona — the subsidized grain distribution — and the spectacles that accompanied it. So Roman politicians invested heavily in the visible, popular parts of food policy and systematically underfunded the invisible, unpopular infrastructure that made the whole system work.

This continued until the infrastructure failed. At which point the popular politicians discovered that no amount of favorable crowd sentiment helps when the grain ships stop arriving.

The Tyranny of the Immediate

Human psychology hasn't changed in five thousand years, and one of its most consistent features is a preference for present comfort over future stability. This is not a character flaw — it's a deeply rational adaptation for most of human history, when the future was genuinely uncertain and the present was genuinely dangerous. Optimize for now. Deal with later when it arrives.

The problem is that democratic governance specifically requires overriding this tendency in structured ways. The entire apparatus of representative democracy — the insulation of appointed officials, the staggered terms, the deliberate friction of the legislative process, the independence of the judiciary — exists to create space between what people want right now and what gets implemented as policy.

Polling, in its modern form, collapses that space. When a senator can see in real-time that a vote on infrastructure funding is polling at 34% favorability, the institutional architecture designed to protect her from that number becomes very hard to use. The number is right there. The consultant is pointing at it. The reelection math is being explained.

So she votes against the infrastructure bill. The infrastructure continues to deteriorate. In fifteen years, a bridge fails or a power grid collapses or a water system contaminates a city, and everyone is shocked.

The Roman Senate would recognize this sequence immediately.

Athens Tried to Solve This — It Didn't Stick

Athenian democracy was actually quite sophisticated about the gap between popular sentiment and sound governance. The institution of ostracism — the practice of exiling a citizen by popular vote — was designed not to punish unpopular people but to remove individuals whose popularity had become dangerous to deliberative function. The Athenians understood that a politician who was too popular, too attuned to crowd sentiment, was a structural threat to the kind of slow, careful decision-making that kept city-states alive.

They also had the boule — a council of 500 citizens chosen by lottery — specifically to insulate certain decisions from the immediate preferences of the assembly. The lottery aspect was deliberate: it prevented the council from being populated by people who were skilled at being popular.

None of this solved the problem permanently. Athens still made catastrophic crowd-pleasing decisions, most famously the Sicilian Expedition of 415 BC — a wildly popular military adventure that destroyed the Athenian fleet and effectively ended Athenian dominance in the ancient world. The expedition polled great. It was a disaster.

What Real Democratic Consent Actually Looks Like

The historical record suggests that durable democratic stability doesn't come from leaders who track and follow public sentiment. It comes from leaders who track public sentiment, understand it, and then make the case for decisions that sentiment currently opposes.

Lincoln's approval ratings were, at various points during the Civil War, genuinely terrible. The Emancipation Proclamation was not popular. The draft was not popular. The decision to keep fighting through 1864, when the war was going badly and peace candidates were making real electoral headway, was not popular. None of that was consent-by-poll. It was leadership in the actual sense of the word: moving people toward decisions they hadn't yet made.

FDR's approval numbers during the early New Deal were volatile and sometimes alarming. The social insurance programs that now enjoy overwhelming public support were, at their introduction, deeply controversial.

The decisions that stabilized American democracy were not the ones that polled well at the moment of passage.

The Dashboard Problem

This is not an argument against polling. Understanding what people think is genuinely important, and the methodological sophistication of modern survey research is a real achievement. The problem isn't the data. The problem is what the data gets used for.

When approval ratings become the primary input to governance decisions — when the question shifts from "what does this situation require" to "what will this demographic in these swing districts support" — you haven't improved democracy. You've replaced it with something that looks like democracy and functions like paralysis.

Rome had bread and circuses because bread and circuses polled well. The infrastructure that actually kept Rome fed and safe didn't poll at all. You can see, from here, how that ended.

The dashboard on that consultant's screen is very sophisticated. The question it can't answer — the one that actually matters — is whether the thing that's popular today is the thing that will still be standing in twenty years. Five thousand years of data suggests that's the only question worth asking.