Local newspapers provide an invaluable service to their communities, covering regional and community items that often go unnoticed on a national scale. And Americans’ trust in these outlets remains unusually high: a study released last year by journalism nonprofit Poynter Institute discovered that Americans’ trust in media remain highest at the local level, with nearly three-quarters of Americans — 73 percent — citing trust in their local paper, compared to 59 percent who said they trust national newspapers.

However, the mass disappearance of news outlets around the country has resulted in a waning interest in local political coverage among Americans in exchange for an increasingly nationalized media diet. And outlets with a national reach find themselves thriving in a hyperpartisan political climate, with conflict in Washington driving revenues and subscriptions at outlets like The New York Times, The Washington Post and The Wall Street Journal, all of which recently surpassed earnings expectations and saw increased subscribers, particularly for their digital editions.

According to a recent study, Americans’ shift in media consumption habits could be driving political polarization throughout the country, which, in turn, may also be having an effect on how we vote.

The study, which was co-authored by communication and political science scholars at Louisiana State University, Texas A&M University and Colorado State University, measured political polarization across the country by looking at split-ticket voting habits, the practice wherein voters choose candidates from different political parties in elections where multiple offices appear on the ballot.

Writing for the Nieman Journalism Lab, the study’s authors drew attention to the notion that split-ticket voting is a practice that has dipped markedly in recent years. In 1992, according to the study’s authors, “voters in more than one-third of the states holding Senate elections elected a Senator from a different party than they voted for in the presidential election.” During the 2016 election cycle, however, “there were no states in which voters did that, and more voters cast straight party-line ballots than at any point in the past century.”

The study examined the number of the daily/weekly newspapers that had shuttered or merged with other outlets between 2009 and 2012 (according to the study, 110 closures were tallied, the majority of which involved weeklies). It then compared split-ticket voting habits during the 2012 election in areas that had lost a local paper with areas where local outlets remained.

The study discovered that in communities that lost a newspaper, voters were nearly two percent more likely — 1.9 percent — to vote same-party in Presidential and Senatorial elections than communities where a newspaper didn’t shutter.

A decline of local newspapers, in other words, could be contributing to increasingly partisan voting habits. As Americans begin turning to nationalized news content in greater numbers, our political identities become a greater driving force in informing our decisions at the ballots.

“We considered two explanations for our finding: Either people switched to national news once their local newspaper closed or they lost the information they needed to vote in more local races,” the study’s authors wrote. “We assessed these options by looking at ‘ballot roll-off,’ when voters leave their ballots blank for lower-level offices. If roll-off increased, that would show that people did not feel like they had enough information to vote in state elections. We didn’t find higher roll-off after newspapers closed. The partisan voting we found is probably due to switching to national news, not feeling less informed about local politics.”

To account for the possibility that others factors—a weaker local economy, for instance—could be the cause for an uptick in same-party votes instead of losing a local paper, the study also analyzed split-ticket voting numbers during the 2012 election in places that had a local paper but lost it shortly thereafter. According to the study’s authors, no correlation was found.

The study, “Newspaper Closures Polarize Voting Behavior,” appears in The Journal of Communication, a bimonthly peer-reviewed academic journal published by Oxford University Press on behalf of the International Communication Association.