How Voters Actually Decide: What Sixty Years Of Research Shows

For sixty years, political scientists have studied how ordinary people choose presidents, senators, and governors. The short answer is that most voters are not policy analysts. They are loyal partisans who use their party identity as a shortcut, and they pay attention to politics only sporadically. The longer answer matters for overdetermined anyone who tries to understand elections, whether as a citizen, a journalist, or a campaign staffer.

The first major finding is that ideology plays a small role for most people. In the 1960s, researchers found that only about one voter in nine reasons ideologically, meaning they hold a coherent set of beliefs from left to right and apply them to candidates. That share has barely moved in decades, as later scholars confirmed. The modern estimate is that only 10 to 15 percent of the public fits that description. Most people do not think in left and right. They think in terms of groups, feelings, and habits. They like the party they have always liked, and they dislike the other one. This is not a failure of intelligence; it is how human beings process a complex world with limited time.

Party loyalty is the strongest predictor of the vote. In a typical general election, roughly nine in ten partisans vote for their own party’s nominee. This means that the real contest is over a small slice of the electorate: persuadable voters, often called swing voters. But here is the crucial twist from the research. Swing voters are not usually undecided on the issues. They are low-information, weakly attached partisans who may stay home or drift back to their default party. They are not moderates who carefully weigh competing platforms. They are people who pay attention only in the final weeks, if at all.

This leads to what political science says about campaign persuasion effects. Across dozens of field experiments in real elections, the best estimate of campaign persuasion in a general election is zero. One major review of 49 experiments found no measurable effect from campaign contact on who voters choose. Ads, door knocks, and phone calls do not convert voters. They mostly mobilize people who already agree with you. This is the Strategist’s Fallacy, a term coined by analyst G. Elliott Morris in October 2025. The fallacy is the belief that voters hold positions and match them to candidates like an entree off a menu. In reality, voters choose a party first, then justify that choice with whatever reasons feel plausible.

So if campaigns do not persuade, what decides elections? The fundamentals: the economy, incumbency, and the general mood of the country. Morris’s two-input model, based on these conditions, forecast the 2024 Democratic nominee at about 48 percent of the two-party vote. She got 49.3 percent. The model was nearly right. On conditions alone, the Republican candidate wins about nine times in ten. The three decisive states turned on 229,766 votes total. That is a razor-thin margin, but it is not random. It reflects a country that leans right on economic conditions when times feel bad.

This raises the moderation debate. Progressives argue that Democrats lose because they are not bold enough. Moderates argue the opposite: that the party drifts too far left. The research offers a different lens. Since voters do not think in left and right, policy positions are less important than perceived trustworthiness and group identity. But there is one finding that complicates this. Broockman and Skovron found that legislators in both parties overestimate how conservative their constituents are. This suggests that politicians are not simply following voter ideology; they are projecting their own assumptions.

For ordinary readers, the takeaway is humbling. Your individual vote is rational, but the aggregate outcome is driven by conditions and loyalty, not by the quality of arguments. Campaign persuasion effects are real but tiny, and they mostly work on the margins. The Strategist’s Fallacy is a warning: do not assume that a well-reasoned platform will move the numbers. The recent report from WelcomePAC, titled Deciding to Win, recommends that the 2028 Democratic nominee be willing to break with progressive advocacy groups. That is a strategic suggestion, not a research finding. But it reflects the deeper truth: elections are won by appealing to the small share of weakly attached voters who are not listening to the ideological debate. Understanding that is the first step to not being fooled by your own side’s narrative.