Jesse M. Shapiro is professor of economics and Robert King Steel Faculty Fellow at the University of Chicago Booth School of Business.Download Working Paper
Jesse M. Shapiro
Professor of Economics
Voter turnout increases when a newspaper enters a market
Hit by a depressed economy, weighed down by debt, and hurt by advertisers and readers moving to digital media, US newspapers have fallen on hard times in recent years. For instance, in 2008, media giant Tribune, owner of the Los Angeles Times and Chicago Tribune, declared bankruptcy. In the same year, the McClatchy Company, which owns 30 daily newspapers across the country, cut 10 percent of its workforce; another 15 percent lost their jobs the following year. While many newspapers have folded, others shrink as they continue to lay off workers, including reporters covering essential beats.
Local news, in particular, has suffered, according to a June 2011 Federal Communications Commission report on the changing media landscape. The report says that many communities now face "a shortage of local, professional, and accountability reporting," which is "likely to lead to more government waste, more local corruption, worse schools, a less-informed electorate, and other serious problems in communities." The number of statehouse reporters, for instance, dropped by one-third from 2003 to 2008, while state government spending rose substantially. Fewer reporters mean that newspapers have less time to spend on in-depth stories.
Of grave concern is the effect of newspaper closings on electoral politics. People may be less inclined to vote if their local newspaper, which keeps them current on political issues and events, suddenly folds. Without newspapers, it may be easier for public officials to get away with corruption and challengers lose an important platform that would help them run against incumbents. Further closures due to the internet may also result in an increased polarization of views, if people instead flock to websites that confirm their prejudices.
To understand how newspapers can affect electoral outcomes, University of Chicago Booth School of Business professors Matthew Gentzkow and Jesse M. Shapiro joined with Michael Sinkinson of Harvard University to analyze a massive dataset of every daily newspaper published in the United States from 1869 to 2004. Their goal was to find out if the entry and exit of newspapers in this period had a significant influence on voter turnout, party vote shares, and the advantage of incumbent politicians. "This is useful evidence from the past that we can use to think about some of the issues still important today," says Shapiro.
The results of the study, which are discussed in a new paper titled "The Effect of Newspaper Entry and Exit on Electoral Politics," are consistent with the idea that newspapers affect the political process by providing valuable information. The entry of a newspaper significantly increased voter turnout, and the impact on congressional elections remained strong even after the introduction of radio and television.
Finding the Impact of a Newspaper
The authors collected data on US newspapers never before digitized on such a large scale. For every year, they matched each newspaper to the county where it was published. Many markets had two or more competing newspapers during the sample period, but the number of counties with more than one paper steadily declined throughout the twentieth century, and most counties in recent years had only one paper.
The study analyzed the impact of the entry and exit of newspapers by comparing the electoral results in counties that experienced a change in newspapers with those counties in the same state and year that did not. Their strategy exploited the fact that exits and entries of newspapers caused large changes in newspaper readership. Changes in readership before or after such events were small, relative to the effect of the event itself.
To be sure that any effect on electoral outcomes is indeed due to newspaper entries or exits, the authors controlled for other factors that may confound the results. For instance, the number of newspapers in a market is primarily determined by population and income growth, both of which also help predict the number of people who will vote.
If people are more likely to vote when they are better informed, then newspapers can play an important role in increasing turnout at elections. Newspapers usually devote many pages to discuss the issues at stake and the candidates' characteristics and platforms. Newspapers also can encourage people to care more about the outcome of an election, or they may simply remind people of the fact that an election is taking place.
The effect of newspapers on voter turnout may vary with the extent of competition and the availability of alternative news sources. More newspapers can lead to lower prices and better reporting, which could expand the market and newspapers’ impact on elections. On the other hand, the first newspaper to enter a market can have a much bigger impact than later entrants, whose readers may already be subscribers of the first newspaper. The effect on voter turnout, particularly for presidential elections, also may be much stronger before the introduction of radio and television.
The study finds that after controlling for population growth, an additional newspaper increased voter turnout at presidential and congressional elections by about 0.3 percentage points. The effect of the first newspaper on turnout was one percentage point, while the effect of the second and third newspapers was much weaker. Thus, newspaper competition is not a key driver of voter turnout.
How big is this effect? If the only newspaper in a county closes, the impact of its absence on voter turnout is about eight times larger than rain on an election day. Newspapers seem to be more persuasive in getting people to vote than direct mail solicitations, but somewhat less effective than talking directly to people by going door to door or calling them on the phone.
Newspapers in the radio and television era seem to matter more for congressional than presidential races. The effect of newspapers on voter turnout at presidential elections diminished in the period after radio and television were introduced, but the impact on congressional elections remained significant. "One possible reason is that radio and television provide a lot more national news relative to local news," says Shapiro. "Having a good local newspaper remains important for congressional elections."
At the moment, the data do not allow for a convincing analysis of the internet’s impact on voter turnout. But just like radio and television, newspapers may lose some of their influence if the internet becomes a good substitute for providing information about local elections. Whether people will be more or less informed of local policies if the internet drives newspapers out of business remains to be seen.
Party Vote Shares and the Advantage of Incumbents
The introduction of a partisan newspaper may persuade people to shift their votes to its preferred candidates by slanting coverage or publishing biased information to sway voters. While it may seem straightforward that the introduction of a Republican newspaper would increase conservative vote shares, newspapers may choose a market that is already conservative for commercial rather than journalistic reasons.
The study finds no evidence that partisan newspapers affect party vote shares. In fact, the persuasive effect of partisan newspapers is much lower than estimates from other research of the effect of Fox News on the Republican presidential vote share or the Washington Post on the Republican gubernatorial vote share in Virginia. That newspapers choose their affiliations to match existing beliefs of consumers in a market may be one reason for this weak effect. Also, partisan newspapers have a known bias so readers may not take them very seriously when deciding whom to vote for. "There is not much information in their recommendation," says Shapiro.
Newspapers can affect the power of incumbents by making it easier for challengers to become known, especially if they are shut out of other communication channels. Newspapers also inform voters about the performance of incumbents, which can hurt corrupt officials but benefit those who are honest and competent. Ultimately, if what newspapers do is expose who the better candidate is, then it is unclear whether newspapers can affect an incumbent’s chances of winning an election. Indeed, the study finds no evidence that newspapers systematically hurt or help incumbents.