The study of economics has always had a strong preference for theory rather than experimentation. New theories that gain wide acceptable are internally consistent with other theories, even if empirical evidence shows otherwise. Contrast this with the study of psychology, where experiments under laboratory conditions are widespread. Economics, on the other hand, often deal with social issues of such scale that experimentation is impossible — it is often not feasible to change one element of a highly complex system and observe the results.
Every once in a while a natural experiment presents itself, however. Sam Schulhofer-Wohl and Miguel Garrido of Princeton University study the impact of the closing of a newspaper on the level of civil and political engagement among readers in a working paper titled, “Do Newspapers Matter?”. Often times, a newspaper closes due to declines in readership, but in this case, the closing of the newspaper was contractually agreed upon thirty years in the past.This allows the researchers to examine the impact of the closing of the newspaper without the impact of other factors that normally accompany the closing of a newspaper. In other words, it’s a natural experiment — something that normally would only be possible in a laboratory setting happens to manifest itself in real life.
Schulhofer-Wohl and Garrido state that “a century ago, 689 cities in the United States had competing daily newspapers; at the start of this year, only about 15 did.” One might say that these newspapers have been replaced by something better (the Internet), but Schulhofer-Wohl and Garrido make a compelling argument that the closing of newspapers have a significant negative effect on the level of civic engagement and political awareness of readers:
The logo of the E.W. Scripps Co., printed on the front page of all its newspapers, is a lighthouse. This paper describes what happened when one of Scripps’ lights went out. The Cincinnati Post was a relatively small newspaper, with circulation of only 27,000 when it closed. 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. We caution that although our preferred point estimates tell a compelling story, the results are statistically imprecise and sometimes sensitive to the treatment of very small municipalities. Further, our results cover only the Kentucky suburbs, because Ohio has not held regular municipal elections since the Post closed, and represent only the short-run consequences of the paper’s closing. Future research could investigate whether political engagement and competition return to their pre-closure level in the long run.
Several other well-known newspapers have closed since the Post (the largest being Scripps’ Rocky Mountain News , circulation 210,000, just this February) and more are in danger. Observers are energetically debating whether these closings matter: Do newspapers play a valuable, irreplaceable role in American democracy? Starr (2009) argues that the newspaper industry’s decline raises practical questions for anyone concerned about the future of American democracy.” On the other hand, after the Rocky closed, U.S. Rep. Jared Polis, Democrat of Colorado, said the paper’s demise was mostly for the better” (Crummy, 2009). Whether our results support Starr’s view or Polis’ depends on how one values competitive elections. But if voter turnout, a broad choice of candidates and accountability for incumbents are important to democracy, we side with those who lament newspapers’ decline.
Read the full paper here.