Turnout Is Not the Variable Everyone Thinks It Is

The turnout theory of American elections is simple and was, for a long time, roughly correct. Occasional voters leaned one way, habitual voters leaned the other, and anything that raised participation moved the result in a predictable direction.

That relationship has weakened to the point of unreliability. The educational and geographic realignment of the past decade rearranged which coalitions contain the habitual voters, and it did so unevenly across regions, which means the national effect of higher turnout now depends on where the additional voters live.

Why the theory persists

It persists because it is operationally convenient. A campaign built on mobilization has a clear task: identify supporters, contact them, get them to the polls. A campaign built on persuasion has a harder one, involving people who disagree with it.

Mobilization is also measurable in ways that satisfy donors. Contacts made, doors knocked, ballots chased. Persuasion produces numbers that are noisier and arrive later, which makes it a harder budget line to defend in a meeting.

A theory that makes the work legible will outlive the evidence that supported it.

What the data actually shows

Studies of recent high turnout cycles find effects that are small, inconsistent in direction, and swamped by composition. The identity of the marginal voter matters far more than the count of them, and the marginal voter is not the same person in a suburb of a growing metro as in a rural county two hours away.

This has a practical implication that campaigns are slow to absorb. Universal mobilization in a district where the opposition has more low propensity supporters is a way to lose by turning out the other side.

The uncomfortable version

The honest conclusion is that persuasion is back, and nobody in either party's professional infrastructure has been trained for it. An entire industry grew up around list building, targeting, and contact rates during a period when the mobilization theory was true enough to work.

Rebuilding a persuasion capability means hiring different people, measuring different things, and accepting that a substantial share of the effort will not be attributable to any specific outcome. That is a hard sell in an industry that sells attribution.

The theory was not wrong when it was formed, which is what makes it hard to dislodge. For a long stretch, occasional voters did lean one way and habitual voters another, and mobilisation genuinely moved outcomes.

What changed is the composition of the coalitions rather than anything about turnout itself. When the educational and geographic alignment shifted, the correlation between voting frequency and partisan lean shifted with it.

The voting supplement is the standard source for who actually votes, and it is published after every federal election with enough detail to test the assumption rather than assert it.

The assumption survives anyway, and the reason is organisational rather than evidentiary. A campaign built on mobilisation has a task list, and a campaign built on persuasion has an argument to construct.

Campaign finance reporting shows where the money goes, and the categories that dominate are the ones that produce countable outputs. Contacts made and doors knocked are auditable in a way that a changed mind is not.

That is the mechanism by which a theory outlives its evidence. It is not that anyone is fooled. It is that the alternative cannot be reported to a donor in the same format.

The operational risk is specific and increasingly plausible. Universal mobilisation in a district where the other side holds more low propensity voters delivers turnout to the opponent, at the mobilising campaign's expense.

That failure mode is invisible in the metrics a mobilisation programme collects, because every one of its numbers went up.

The administrative layer matters here too and is usually treated as neutral. The rules that govern access determine who finds voting easy, and changing them alters composition rather than volume.

Election administration carries most of that in practice, through decisions about hours, locations and equipment that are made by officials nobody campaigns against.

The uncomfortable conclusion for the professional infrastructure is that persuasion is the binding skill again and almost nobody currently working has been trained in it.

Rebuilding that capability means hiring different people, measuring different things, and accepting that a large share of the spending will not be attributable to anything. It is a harder sell than a dashboard, and it is the work.