While the big story for Republicans continues to be the rise of Donald Trump, perhaps the biggest surprise on the Democratic side is tepid turnout, which (rightly or wrongly) has raised questions about whether Bernie Sanders’s political revolution is fading. As Jeff Stein reports, “Participation is down this year for Democrats by about 23 percent, or about 150,000 votes, through the first three states.” And on Saturday turnout was down in South Carolina by almost one-third.

Lower turnout potentially matters for both Hillary Clinton and Sanders. But arguably it matters more for Sanders, as the key to his political revolution all along is the promise of expanding the electorate by bringing more first-time, young, and lower-income voters into politics. The hope is that, after they become engaged electorally, they’ll remain engaged in the policy debates long after the election.

So far, his campaign has generated a lot of enthusiasm in surveys, but this isn’t always translating into voting. Why not?

First, it’s important to consider another possibility: Despite lower overall turnout, perhaps there has been an increase among the kinds of people most likely to support Sanders (and a decrease among those most likely to support Clinton). This is a claim that many Sanders supporters have recently offered.

Below are poll data from Iowa, New Hampshire, Nevada and South Carolina for 2008 and 2016, showing the percentage of the electorate coming from the kinds of groups most likely to represent Sanders’s expanded electorate. In almost all cases, these groups don’t make up more of Democratic primary voters in 2016, compared with 2008. Given lower overall turnout, this also means that there are fewer of them showing up this year.

The one exception was young voters in Nevada, whose percentage increased from 13 to 18 percent. Yet given that Nevada’s turnout reportedly dropped by one-third, it’s still the case that fewer young voters turned out this year than in 2008.

In sum, overall turnout is down, and there’s no evidence that Sanders has mobilized a large number of new, young, and/or lower-income voters.

So how do we interpret this non-revolution? One explanation making the rounds is that Democrats have an “enthusiasm problem” this year.

Maybe so. But I want to offer a different interpretation. Converting support into action is really difficult among people who do not habitually go to the polls. One of Sanders’s main ways of trying to motivate these folks is to talk about their everyday financial concerns — making the personal political.

The problem is that, when political rhetoric reminds voters of their personal financial constraints, it demobilizes them. Voters are busy, and their financial situation is increasingly precarious. This is especially the case for the kinds of voters that Sanders is trying to appeal to. Moreover, when it comes to participating in primaries and especially caucuses, voters are more likely to perceive that politics demands time.

Thus, although it may sound like a good idea to mention issues that remind voters of their financial concerns, such reminders can backfire. Arguably this is why Sanders’s (very successful) fundraising rhetoric emphasizes different considerations.

The upshot is that we shouldn’t attribute lower Democratic turnout to an enthusiasm gap that will continue to matter in the general election. Instead, we should attribute it to primary election rhetoric reminding potential new voters of their many other pressing concerns. During the general election, Democratic voters will largely have other cues to rely on, not least of which is a Republican alternative to oppose.

One way to avoid unintentionally demobilizing voters is to downplay any self-undermining quality of economic insecurity. That is exactly what happened along the Las Vegas Strip during the Nevada caucus. There, casino workers were given paid time off to caucus, which allowed them not to be preoccupied with financial concerns and instead spend time caucusing. In addition, there was probably social influence generated among co-workers — people got excited about caucusing as their co-workers were talking about it and going. Overall, there is reason to believe that these factors contributed to a high turnout in these areas.

Adam Seth Levine is an assistant professor in the Department of Government at Cornell University.