We have just come off an election with surprisingly low voter turnout. Turnout fell more or less steadily from 1960 until 1996, when slightly less than half of the voting-age population went to the polls. But then it ticked up: 51 percent in 2000, 54 percent in 2004, and 58 percent in 2008. In 2012 these gains largely vanished. Turnout fell to 53 percent.

The decline in turnout from 2008 to 2012 is one of the most puzzling features of November’s election. It is unclear not only why turnout dropped, but also how Obama managed to win in spite of it. For several reasons, it seemed that a big decline in turnout would have harmed his reelection bid.

First, get-out-the-vote efforts, lauded as the backbone of Obama’s 2008 victory, were just as intense in 2012. The decade-long rise in turnout is often attributed to improvements in get-out-the-vote efforts, but last year’s decline ought to raise doubts about whether these activities can sustain elevated levels of participation.

Second, it was widely argued that Obama needed to maintain turnout close to 2008 levels in order to win. Indeed, in isolation, Obama’s 2012 numbers look discouraging. He received 69.5 million votes in 2008, but only 64.8 million votes this time around. Normally, losing almost five million votes from one election to the next would spell defeat. Mitt Romney, however, was unable to bring those votes to his side. He won 60.5 million votes in 2012—only 500,000 more votes than John McCain had in 2008. The result was good enough for a three-point Obama win and a big loss for conventional wisdom on turnout.

Third, much of the increase in turnout over the past decade came from newer voters, primarily minorities and young people. Newer voters are traditionally thought to be “marginal” voters, people who do not vote habitually and who are more likely to stay home in a low turnout year. It was reasonable to expect that if turnout dropped in 2012, younger cohorts and minorities would show the biggest declines in participation, hurting the Obama campaign. However, according to the exit polls, minority and young voters increased as a percentage of the electorate, compared with 2008, reflecting long-term trends.

So why were fewer ballots cast overall? The answer isn’t yet clear, but exit polls suggest that one demographic with a large decline in voting share was middle-income whites. Why they sat out is an even bigger question. Was it the campaigns? The candidates? The economy? It’s too early to say.