MAY 2005

• • • • • • • • • • • • • • • • • • • • • • • • • •

Spin Cycle

Truth and Lies in the Market for News

Research by Matthew Gentzkow and Jesse M. Shapiro

Editors and reporters have many different methods by which to shape the facts and vary the impression conveyed by their news stories. Different stories may be based on the same set of facts, but each outlet may convey a radically different version of the same story via selective omission, choice of words, and varying the credibility ascribed to sources. Such choices are commonly referred to as "spin" or, depending on one's perspective, "bias."

Recent examples of media bias include the different slants in television coverage of the war in Iraq between U.S. networks and Arabic-language news channels such as Al Jazeera.

Concern about media bias is now prominent in many public policy debates, from public diplomacy in the Middle East to media ownership regulation by the Federal Communications Commission. In addition, survey evidence revealing rising polarization and falling trust in the news media has prompted concerns about the media's ability to deliver credible information to the public.

In the recent study "Media Bias and Reputation," University of Chicago Graduate School of Business professor Matthew Gentzkow and Jesse M. Shapiro, the Initiative on Chicago Price Theory's inaugural post-doctoral Becker Fellow, present a new way to understand media bias. They find that patterns of bias need not arise from consumers' desire for information that confirms their existing beliefs, reporters' interest in promoting their own views, or direct political influence on media firms. Instead, the bias observed is often a natural consequence of media firms' desire to build reputations as high-quality sources for accurate news.

A New View of Media Bias

Gentzkow and Shapiro's analysis begins from the premise that media firms try to build a reputation for truthful, accurate reporting.

The high costs that media firms are willing to incur to gather information provides strong evidence of the reputation-building incentive, as does the response of media firms whose reports are revealed to be inaccurate. For example, Jayson Blair's fraudulent reporting at The New York Times prompted the resignation of top-ranking editors Howell Raines and Gerald M. Boyd.

Gentzkow and Shapiro point out that when consumers are uncertain about the accuracy of a news source, consumers will try to discern its quality by looking at how it has reported the news in the past. Reports that prove to be inaccurate naturally lead them to downgrade their estimate of quality. By contrast, surprising reports that turn out to be true increase consumers' confidence in the source.

Of course many stories are never shown conclusively to be either true or false. In this case, consumers' prior beliefs will play a key role in their judgment of quality.

If a news source reports something consumers believe to be highly unlikely-for example, Elvis being spotted in Central Park-consumers will rationally infer that the source had poor information, exercised poor judgment, or had some objective other than reporting the truth. By the same logic, a reader who believes the American military takes extraordinary measures to avoid harming civilians in Iraq will be skeptical of a news source that suggests civilians are being intentionally targeted.

Gentzkow and Shapiro argue that this mechanism can explain a large body of evidence documenting a strong connection between consumers' prior views and their assessments of information sources.

For example, in a recent survey, nearly 30 percent of people who identified themselves as "conservative" indicated that they thought they could believe all or most of the information provided by the Fox Cable News Network. Less than 15 percent of self-described "liberals" subscribed to this view. The opposite pattern occurred for National Public Radio. More than 35 percent of liberals believe reports by National Public Radio to be accurate, while only 20 percent of conservatives hold such a view.

What Determines Media Bias?

Building on this intuition, the authors show that bias can arise even when consumers care only about the truth, media firms care only about maximizing profits, and eliminating bias could make all agents in the economy better off. Their study identifies three factors that play key roles in determining the direction and strength of media bias: 1) consumers' prior beliefs; 2) ex-post feedback; and 3) competition.

A media firm concerned about its reputation for accuracy will be reluctant to report evidence at odds with its consumers' attitudes, even if it believes the evidence to be true. The more that consumers' prior beliefs favor a given position, the less likely a media firm becomes to print or broadcast a story contradicting that position.

"Your prior beliefs about what is likely to be true affect how you draw inferences about the quality of a news report," says Gentzkow. "Consumers' prior beliefs play a large role in what media firms want to say in order to appear to be the most accurate source."

To illustrate the connection between consumer beliefs and media slant, the authors point out that in the 2000 election, the vast majority of newspapers in traditionally Republican states endorsed Bush, whereas this pattern was often reversed in more Democratic states.

The second determinant of media bias is ex-post feedback, which refers to information that confirms or refutes a set of facts after a story is published. The more likely the truth is to come out after a report, the stronger is the firm's incentive to tell the truth.

Gentzkow and Shapiro's model of media bias therefore predicts that there will be less media bias where facts are concrete and outcomes are immediately observable, as is the case with weather and sports. Bias is much more likely in the coverage of topics such as foreign wars, discussion of alternative tax policies, and summaries of information about global warming. In these cases, the truth is hard to see and often not known until long after reports are made.

The third determinant of media bias is competition. The authors find that greater competition in news markets leads to lower bias, because with more news outlets reporting, there is less chance of getting away with an error.

For example, on September 8, 2004, former CBS News anchor Dan Rather reported the emergence of new evidence indicating that President George W. Bush's family had pulled strings in order to get him into the Texas Air National Guard to avoid serving in Vietnam. When later information indicated that the documents on which the report was based may have been fabricated, both Rather and CBS News President Andrew Heyward issued apologies emphasizing the importance of a reputation for truth-telling in journalism. The fallout from the incident eventually led to the resignation of four CBS employees, including senior vice president Betsy West, executive producer Josh Howard, and senior broadcast producer Mary Murphy.

"Having many independent voices in the market means the truth will come out eventually," says Shapiro. "If everything was owned by CBS, we might not have seen this response. The Rather example illustrates the role that competing media outlets can play in exposing flaws in journalism."

Regulation and Public Diplomacy

In the current debate over FCC regulation, the main argument in favor of limits on consolidation has been the importance of independent voices in news markets. Gentzkow and Shapiro offer a new way to understand the potential costs of consolidation. Independently owned media outlets can provide a check on each others' coverage and thereby limit media bias, an effect that may be absent if outlets are jointly owned.

A larger issue that concerns the authors is the public diplomacy crisis that the United States currently faces, especially in the Muslim world.

The U.S. government is currently engaged in a debate about the most effective way to counter what it sees as anti-American bias in the Arab media, especially Al Jazeera. Policy approaches have included the condemnation of Al Jazeera by top U.S. officials, appeals to the Emir of Qatar (who sponsors the network) to change the tone of Al Jazeera's coverage, and the closing of Al Jazeera's Baghdad office by the U.S.-backed Alawi government in Iraq in August 2004.

Gentzkow and Shapiro suggest an alternative approach. They recommend that the United States support the growth of the Middle Eastern media market, and in particular, increase the availability of alternative news sources in local languages. Aside from the direct effect on beliefs of viewers in the region, introducing more news outlets could have the effect of disciplining existing stations and reducing the amount of media bias in the region.

Relatively low-cost solutions include subtitling reports by CNN in Arabic. More generally, policymakers may consider introducing other credible Western news sources into the Middle East media market.

"If we provide Arabic CNN, and it's a reliable source, then anyone who wants to contradict that source would be taking a risk," says Gentzkow. "The United States can get a lot of bang for its buck if we can achieve journalistic credibility in the Middle East."

Gentzkow adds, "A clear implication of this paper is that in public diplomacy, we should focus on trying to improve the quality and quantity of information available. Unlike current approaches, this method would be consistent with the values the United States says it is promoting: competitive, open free markets, and freedom of the press."

• • • • • • • • • • • • • • • • • • • • • • • • • •

Matthew Gentzkow is assistant professor of economics at the University of Chicago Graduate School of Business. Jesse M. Shapiro is the inaugural post-doctoral Becker Fellow for the Initiative on Chicago Price Theory at the University of Chicago Graduate School of Business.

>> View the study: Media Bias and Reputation