On the first day of early voting in Illinois, President Obama made a conspicuous effort to promote the initiative. “I’m so glad I can early vote here,” he said. Then shortly before leaving, he was sure to add, “Early vote, everybody!” Pundits and politicians have been quick to praise early voting for its democratic value. After all, it seems logical that early voting would make our elections more participatory. If we make voting easier, then more people will vote.

But studies performed by political scientists here at Yale and at the University of Wisconsin-Madison indicate otherwise. In “Election Laws, Mobilization, and Turnout: The Unanticipated Consequences of Election Reform,” researchers from UW-Madison write, “One of the most popular election reforms among state governments may inadvertently result in fewer voters at the polls.” They use empirical data gathered from analyses of voter turnout in the 2004 and 2008 elections to demonstrate this paradox of early voting.

By expanding the number of voting days, you naturally dilute the inherent potency of Election Day. A soft two-week period of early voting makes Election Day a subconscious perpetuity rather than a hard-and-fast deadline. With this new perspective, it’s easy to imagine how people may procrastinate, forget or outright neglect to vote. Your vote seems much less urgent, less impelling if Election Day can move according to your personal schedule. Ironically, the greater ease with which you can vote may actually discourage some people from voting in the first place.

In addition to reducing the civic significance of elections for individuals, early voting warps the behavior of crucial voting blocs and can hinder the mobilization efforts of political campaigns. In “Dynamic Voting in a Dynamic Campaign: Three Models of Early Voting,” a study out of Yale, the authors state that in the 2012 elections, high-participation voters took advantage of early voting while low-participation voters did not. Meaning, those who voted early were those who would have voted regardless of the existence of early voting laws. The authors go on to suggest that early voting tends to demobilize independents and young voters more than their partisan and older counterparts. If this is the case, early voting may compound the problem of polarization and partisanship.

The gains made by millions voting early are lost when millions more who would have showed up on Election Day fail to do so. The result, as both studies argue, is a net decline in voter turnout. Voters need an Election Day stimulus. Early voting instead depresses turnout by depriving Election Day of its stimulating effects and by its innate design, robbing Election Day of its special camaraderie. Early voting makes voting less and less of a singular jubilee process and more and more of a check-list item to postpone.

These studies don’t toss out the merits of early voting outright, but they do shed light on the complexity of voter behavior and the unintended consequences of election reforms. The debate over early voting is particularly relevant to New Haven because next week, Connecticut voters will have the chance to amend the State Constitution to permit early voting in the state. Question one on the ballot will ask, “Shall the Constitution of [Connecticut] be amended… to permit a person to vote without appearing at a polling place on the day of an election?”

As democrats (lower case “d”), we should support policies that encourage more meaningful voter participation. Because of this, evidence suggests Connecticut should leave its Constitution be. Early voting neither helps democratize our elections nor improves our politics. We should all seriously reconsider the conventional assumption that anything that makes voting easier will naturally increase voter turnout. As counterintuitive as it sounds, in most places, early voting has had the opposite effect. Like so many other policies, even the simplest and most benign changes can produce cascades of unintended consequences.

Doo Lee is a junior in Davenport College. Contact him at doo.lee@yale.edu.