TT Epaper
The Telegraph
Graphiti
 
CIMA Gallary

THE MARKET FOR FACTS

- An economist must make sense of the statistics India offers

Although I am an economist, I have never been able to make much sense of the media industry. I have a number of young fans; some of them dream of becoming columnists like me, and ask me how to do it. I have no idea. As far as I can see, a horde of potential writers chases a small number of newspapers and magazines. Their owners have plenty of choice; they pick lucky numbers, and turn them into writers. My number has been quite lucky; ever since I started writing in a Delhi magazine in 1964, I have had a ready market. I have written steadily except in periods when I was out of the market. One needs only one outlet at a time. I have always had one when I needed one, sometimes two; whenever one outlet threatened to make itself unavailable, another has always turned up. Now I need to be lucky for only a few years more; even if I lose my little space, there are plenty of online opportunities.

Matthew Gentzkow has specialized in the economics of the (American) press and come up with some innovative analytical techniques and reached surprising conclusions. Some eight years ago, he started with three simple propositions. Readers would prefer a newspaper which shares their preconceptions. An editor who knows this will design a newspaper that matches the preconceptions of the largest possible number of readers. Hence the press will be inexorably biased towards the prejudices of the majority of readers. Gentzkow and Shapiro (Journal of Political Economy, April 2006) illustrated the bias by citing the report of the same bit of news by Fox News, Al Jazeera and The New York Times.

But the game changes if the public can verify the veracity of news. In that case, the seller of news is torn between giving his readers the real news and pandering to their prejudice; he is likely to lean towards the truth. The more the newspapers, the greater the competition, and the more will they lean towards truth. Their bias will be least where the reader can experience the truth for himself. For instance, most of us are hopelessly biased in favour of our national cricket teams. But news about cricket matches is never biased, because anyone can find out the scores. Ditto with the weather; if I see it raining outside, newspapers tomorrow will never say it was sunny. It is different if the newsmaker can manipulate the news. Those who invest in shares like their companies to do well. So managers often maximize short-term profits at the expense of their companies’ long-term future. That is why bonuses based on current performance are a bad idea.

Gentzkow then turned to the newspaper market in Washington (American Economic Review, June 2007). The advent of online news has been a matter of great concern for American newspapers. They have started their own websites; but then they have to decide whether to charge for online access or not. The Washington DC market had two newspapers, The Washington Times and The Washington Post; then the Post started an online edition. With sophisticated econometrics, Gentzkow showed that while the revenue of Post.com was $33.2 million, it had reduced the profits of the Post by $5.5 million and those of the Times by $2.7 million. More important, however, the online edition had increased consumers’ surplus by $45 million.

Gentzkow and Shapiro make a distinction between product markets and information markets (Journal of Economic Perspectives, Spring 2008). Product markets have a number of producers competing in it. News is different. It is produced in one place at one time, and the media carry reports of it. It may be carried by only a handful of media, but there is no limit to who will carry it; and with the advent of the television and internet, the number of ways it can travel from any of those news sources to a consumer is enormous. Because the same news may be picked up by and reach a consumer through a number of channels, it is impossible to manipulate the news; the consumer will compare sources and find out. When he does, whoever disseminates false information will pay a high price: he will lose reputation, which is difficult to regain. So Gentzkow and Shapiro told the US government to stop worrying about competition in media in general. If there was a problem, it was in small cities; some of them had few newspapers, and their newspapers were generally too poor to spend much on news gathering.

People have political views and loyalties. So do newspaper owners. There is always the question: how far can the press influence or mould public opinion? If it had a strong influence, it could jeopardize democracy. And there is no Chinese wall between news and views. News can be slanted, and opinion pieces can be objective. Gentzkow and Shapiro invented a slant index (Econometrica 2010 January): from Congressional Record they identified phrases that were closely identified with one of the two parties. They then measured the frequency with which newspapers used these phrases to measure their bias. The bias of readers could be estimated more directly from the proportion of Republicans and Democrats in different zip code areas. Profit maximization would imply that newspapers would try and match their readers’ slant. Not surprisingly, Gentzkow found that they did. But that still left the question open: did readers bias the press, or did the press sell bias to readers? Gentzkow found an index of bias that the press could not influence, namely the readers’ religion, and showed that newspapers’ bias accounted for only a small part of readers’ bias. Then he asked whether the newspapers reflected the bias of their owners, which he measured on the basis of their political donations and the bias of other newspapers they owned. He found that they did not. In other words, owners gave their newspapers a slant to maximize their sales; they did not try to sell their own prejudices if that went against profit maximization.

In their most path-breaking study, Gentzkow, Shapiro and Sinkinson (American Economic Review, December 2011) fitted equations to US data from 1869 to 2004. They showed that the publication of the first newspaper in a place had a huge effect on voting percentage, and that more newspapers also led to a rise in voter participation, though smaller. This effect was just of the appearance of a newspaper, and not of competition in the media. They also showed that newspapers were pretty ineffective in persuading voters to change their party preferences, and that they had no influence on the chances of incumbents.

If we want to know how our media perform and how far they affect our politics, we should invite Gentzkow to spend some time in India and try his hand on our statistics. We have masses of statistics in this country, but they are useless without the magic wand of a good economist.