Biased? Who, Us?

There are two types of media bias: ideology, which reflects a news outlet's desire to steer reader opinions in a particular direction, and spin, which reflects the outlet's attempt to create a memorable story and, thus, gain or retain readership. But do these forms of bias have any measurable effects, in particular on consumer confidence?

That's what Dutch researchers Karel-Jan Alsem, Steven Brakman, Lex Hoogduin and Gerard Kuper set out to dilucidate in their CESifo Working Paper The Impact of Newspapers on Consumer Confidence: Does Spin Bias Exist?

Competition, point out the authors, plays a significant role in the prevalence of these forms of bias. On the one hand, it provides counterweights to a news outlet's ideological bias, as its competitors present their alternatively biased versions to the reader: the sum total of stories available shows truth to lie somewhere in the middle.

On the other hand, competition can exacerbate the temptation to use spin: the more you succeed in presenting your stories in an attractive manner, the larger your readership, and the more copies you sell.

The problem is that once these memorable stories stick to your mind, it is fair to presume that they can affect your opinion of things. And consumer confidence is one of those things that depend on your opinion.

In order to test whether a putative spin bias does affect readers' opinions, the authors had first to check whether such a bias can be identified in newspaper reports. For this purpose, and assuming that the fierce competition in the Dutch daily newspaper market stimulates newspapers to exaggerate their stories in order to make them more memorable to readers, they resorted to articles on the state of the economy published in Dutch newspapers from 1998 through 2003. They duly detected some evidence of spin bias in those articles. After reading such reports, readers may see their consumer confidence –which is in essence an assessment of the state of the economy– influenced by them.

If the authors' hypothesis is right and consumer sentiments are affected by what they read in the papers, then they will have a biased picture of the state of the economy. In such case, the business cycle perceived by consumers will show larger amplitude than the actual business cycle. Spin, then, can be interpreted as a magnifying force with respect to consumer sentiments about actual facts on the state of the economy.

The researchers make some caveats here, before taking the plunge onto their data: they are careful to point out that they have not looked into the problem of whether or not changes in consumer confidence exert an effect on the real economy, but only if consumer confidence itself is affected by the news. Additionally, the relationship between consumer confidence and the real economy is still not completely understood, and appears to be weak. All of this would point to producer and consumer confidence being influenced by different variables.

The authors put their spin hypothesis through two tests. First, they examined whether consumer confidence is more volatile than producer confidence. The latter is not influenced by sentiments, since it is measured by objective economic variables. Second, they developed a measure of how consumers perceive the state of the economy through what they read in newspapers, and then estimated whether this measure has a larger impact on consumer confidence than it does on producer confidence. If such turns out to be the case, they took it as evidence of the existence of spin bias.

Then, they estimated a vector autoregression (VAR) system with three variables: producer confidence, consumer confidence, and media. Impulse analyses were conducted on the basis of this system, using changes in media as impulses. And, bingo: changes in media have a significant, if short-lived, effect upon consumer confidence, but none on producer confidence. Adding a stock market index to the authors' VAR did not change the impact of media on producer and consumer confidence.

Still, this does not rule out other possible explanations for this effect, such as the mere fact of a story being written on a newspaper having a possible effect in readers' opinions. Furthermore, only newspapers were examined. Does this effect hold also for radio or TV reporting? Intuitively, one would surmise that it probably does. But these are scientists here, and they clearly detected a virgin field for further, in-depth research. Stay tuned.

Karel-Jan Alsem, Steven Brakman, Lex Hoogduin and Gerard Kuper: The Impact of Newspapers on Consumer Confidence: Does Spin Bias Exist?, CESifo Working Paper No.1328

• Other CESifo Working Papers dealing with news and the economy
• Other CESifo Working Papers by Steven Brakman
• CESifo Working Papers published in the past three months