The old consensus was not really. Voters, it was argued, basically reflect the views of non-voters. In recent years, some political scientists are not so sure. Two empirical facts drive the new skepticism.

First, despite significant gains in high turnout elections like 2020, the electorate that turns out remains unrepresentative on axes of class, race, and age. Apart from seniors, the general rule is that individuals at the top of the socioeconomic ladder are much more likely to turn out on Election Day than those at the bottom. Income and education, however, are not the only axes of distortion. Americans who are over 65 remains significantly more likely to vote than those who are under 35.

The demographic biases of the electorate are most acute during primaries, midterms, and off-cycle elections—including the age bias.

The old view was that none of this mattered because the electorate’s partisan preferences roughly match that of the eligible electorate. That may still be true, but the conclusion that non-voting doesn’t matter may still be wrong. The strongest data pertain to the gap between young and old voters. A series of PEW studies show that younger Americans have distinctly more progressive views on a range of views, regardless of party affiliation from gay marriage and racial inequality to global warming and the role of government. Data also show that young voters are significantly more likely to prefer the Democratic Party. (My personal view is that we underestimate how the age bias of the electorate explain mid-term backlash against the party in power—the Nate Cohn graph on below 60% turnout would seem to confirm that hunch.)

Beyond age, the class bias of the electorate may also matter. Nonvoters generally have different life experiences, and this seems to translate into different policy preferences—if not partisan preferences. Kay Schlozman’s 2012 study, The Unheavenly Chorus, found inactive voters are not only much more likely to report struggling to pay bills, obtain healthcare, and find decent housing, but also much more likely to have utilized public benefit programs. Data also suggest that economic position affects economic policy preferences from welfare spending to taxes—though not a range of other hot-button policy issues. For example, Pew found, in 2014, that “Opinions about the factors that result in wealth and poverty differ across income categories,” and further that this appears to translate into different views of the value of government assistance, reporting that “People with lower incomes express more positive views of government programs to aid the poor than do those with higher incomes.” Academic studies on the New-Gilded Age and the distinct views of the super-rich, including the work of Benjamin Page & Cari Lynn Hennessy, are consistent with the broader observation that views on economic policy tend to track class position.

Apart from the above, a recent Knight Foundation study emphasizes that perhaps the greatest difference between voters and non-voters is their lack of faith in the electoral system.