Douglas Clement - Editor, The Region
Published September 27, 2011 | September 2011 issue
The collapse of the newspaper industry in recent decades is well-known, in part because of how extensively the media themselves have covered the trend. The consequences of this decline have been bemoaned by many, but rarely are they assessed with any degree of accuracy. Recent research by Minneapolis Fed Senior Economist Sam Schulhofer-Wohl and Miguel Garrido, an Edgeworth Economics consultant, provides a revealing, if preliminary, empirical demonstration that reduced news coverage due to fewer newspapers can indeed affect important civic activity and the democratic process itself.
In “Do Newspapers Matter? Short-Run and Long-Run Evidence from the Closure of The Cincinnati Post” (Minneapolis Fed Working Paper 686), the economists study elections in northern Kentucky communities near Cincinnati, Ohio, following the closure of a newspaper that had previously provided widespread coverage of the local political scene. They find that after the newspaper closed, fewer candidates ran for office, incumbents were more likely to be reelected, and voter turnout and campaign spending fell. Voter turnout remained depressed for nearly three years, though other effects diminished with time.
“The Cincinnati Post was a relatively small newspaper, with circulation of only 27,000 when it closed,” write the authors. “Nonetheless, its absence appears to have made local elections less competitive along several dimensions: incumbent advantage, voter turnout, campaign spending, and the number of candidates for office.”
The consequences of shrinking newspaper numbers—689 U.S. cities had competing newspapers a century ago; just 11 cities currently do—have been studied by a number of researchers, but few have managed to separate causation from correlation. A political trend following a newspaper’s closure may be due to the closure, but an unmeasured factor might have caused both, or the timing could be coincidental.
Schulhofer-Wohl and Garrido mitigate this problem by studying a “natural experiment” in which the timing of a newspaper’s closure was virtually determined 30 years in advance, so had little or nothing to do with current political trends. (They use several statistical checks to further strengthen their case.)
In 1977, the Post and a competing newspaper, The Cincinnati Enquirer, entered into a joint operating agreement (JOA), a legal arrangement established by Congress in 1970 under the Newspaper Preservation Act. With the goal of preserving a free press and diversity of editorial viewpoint, the act exempted competing newspapers in “economic distress” from antitrust laws and allowed them to charge monopoly prices for subscriptions and advertising. The Post-Enquirer JOA specified an explicit endpoint of Dec. 31, 2007.
Hit with an almost 90 percent decline in circulation over the next 30 years, the Post did in fact close its doors on that date, while the Enquirer continued to publish and, indeed, increased its news coverage of the Post’s former geographic strongholds in northern Kentucky. The predetermined closure date supports the economists’ argument of causality since the newspaper closed for reasons wholly unrelated to local politics.
The economists gathered data between 2003 and 2010 for 48 municipalities in the seven counties of the Post’s core or near-core areas of Kentucky circulation. The data include a number of stories about each municipality, results of local elections, candidate spending in each election and demographics in 2000. From the election information, the economists construct several measures of “political engagement and completion”: estimates of voter turnout, ratio of candidates to seats available, fraction of seats won by incumbents and total expenditure for local political office campaigns.
They then run regression equations to gauge statistically the impact of shifting levels of newspaper coverage on each of the dependent political variables, and the results are unequivocal: “On all four measures of political engagement and competition, we find indications that the Post’s closure made elections less competitive.”
Their estimates of political events in areas where the Post was more important than the Enquirer show that after the Post closed:
The economists provide numerous elaborations and qualifications to their findings: Controlling for race and age structure doesn’t affect qualitative results, for example, though the size of the effect does change somewhat; not all results are statistically significant at conventional levels; the short-run impact they’ve measured might diminish with time as other media replace the coverage previously provided by the Post. Nonetheless, the decline in political engagement occurred even though the Enquirer increased its coverage in Post strongholds after the Post closed.
Perhaps the largest qualification is that these results were obtained in just one geographic area. Would they be equally strong elsewhere? “Future research could investigate the consequences of closing of other news-papers,” write Schulhofer-Wohl and Garrido, “though a significant challenge is to find an exogenous or at least predetermined closing.”
Assuming their findings are more generally valid—that fewer news-papers suggests less competitive elections—the overarching question remains: How valuable are competitive elections? “If voter turnout, a broad choice of candidates, and accountability for incumbents are important to democracy,” the economists conclude, “we side with those who lament newspapers’ decline.”