The Decline of the Mainstream Media and the Theory of Informational Cascades

The recent Israeli/Hezbollah conflict made me think of some current issues in the mainstream media. While I do think that my game theoretic analysis of this recent conflict was good -- and may end up being right on yet -- it certainly did not answer all my questions. As that conflict went on, additional questions and issues emerged. One that has been written about in the blogosphere a fair amount concerns the way that Hezbollah seemed to control the battlefield of the media. While Israel may have gained some strategic advantage from the conflict (granted that the final outcome is still unclear) I don't think anyone would argue with the view that Hezbollah won the communications war.

So let's put this into a broader context.

And I do want to give the Economist credit for its cover story this past week, "Who Killed the Newspapers?" There was not too much to follow inside the cover, but a few points were made.

Since the onset of the TV (geez, I remember our first black and white!) the newspapers have fought for market share of readers and of advertisers. The recent Internet-based economic changes have further affected the market position of not only newspapers, but the whole "mainstream media," TV and news magazines as well. Richard Posner had a great review article in the Sunday New York Times a while back, laying out very well how the new technology destroys the old economies-of-scale model of the old media business (see here).

The point is that the old business model of the TV news service, newspaper, or news magazine is seriously suffering. They are now seriously losing readers and advertisers to the Internet.

Whether this is the correct response or not, I don't think there can be argument over the proposition that the mainstream media are cutting their investigative journalism budgets. In the face of declining revenues and profits, a natural reaction is to cut the number of reporters.

So we have a situation where the amount of true investigative reporting in the industry, worldwide, has declined. This now sets us up for situations where information cascades and "rational herding" will occur more frequently. Once I explain these, you ask yourself if you do not agree that we are seeing more herding and cascades in the media.

A good link to the theory of informational cascades and herding is by Ivo Welch at Brown, here.

Let me briefly describe the theory. I think informational cascades are the easiest and most illuminating. Suppose you are uncertain about something, say whether a news story is true or not. You have some private information on whether the story is true or not, and so does everybody else. You don't observe other people's information directly, but you do observe how other people behave -- say, whether they choose to pass the story on to other people. If other people only pass on stories that they believe are true, then when you see someone passing on a story, you should rationally infer that their information supported the truth of the story. With fairly reasonable underlying assumptions on the structure of the information, once you observe even two people pass the story on, you will have to rationally assume from that point on that the story is true. You will therefore pass the story on, and now the next person has seen three people act as if the story is true. They will have even more reason to disregard their own information and act just like everyone before them...hence the term informational cascade.

A key point, of course, is that those two people who made the initial decisions to pass the story on could have had bad information (this is all in a world of uncertainty). No matter, once they decide to pass the story on, everyone thereafter will behave in the same way. So we will get a lot of false stories passed on as truthful, and everyone will believe them to be true!

And not to get ahead of ourselves too much, but suppose you know that this is how the world is operating, and you (Hezbollah) decide to be those two people who make the first decisions that everyone else is going to use to infer what is actually going on...

This is the theory of informational cascades. The theory of rational herding is very similar; the term herding refers to the tendency of people to make similar decisions under conditions of uncertainty. Herding can be reinforced by other incentive issues. For example, in the investments world, managers of mutual funds might "herd" not only because they are watching one another, but because their compensation is based on relative performance. If I know what other fund managers are doing in terms of stock picks, then if I mimic that, I cannot go too far astray in my performance.

I think the applications to the media industry are now apparent. With the decline in resources devoted to true investigative reporting, the tendency towards informational cascades and herding are stronger. Nobody really knows what is going on, so when we see someone with some information, we will rationally believe it. I cannot yet present data to support my claim, but my casual observation is that "herding" is more rampant in the media than before. It seems that one story or one fact has much longer and stronger "legs" than ever before. Deaths in Iraq are one example; someone puts out the data on how many were killed, and everyone reports that story. The craziness over Mr. Karr who confessed to killing Jon Benet Ramsey is another good example. Does anyone want to write a paper with me that would create some measure of "herding" for the media and show that it is negatively correlated with the resources devoted to investigative reporting by the mainstream media?

Two further observations. The first is really important, I think. In a world with less investigative reporting by the Fourth Estate, the ability for governments or other organizations and individuals (read Hezbollah, OBL, etc.) to influence people's beliefs is enhanced. I hinted above how someone like Hezbollah could start an informational cascade. Does that story describe pretty well what happened in the recent Lebanese conflict? I think so. Any country that enters into a conflict without a grand strategy of controlling the media is in for a real battle. Israel certainly lost the battle for world opinion, and increasingly, it appears to have done so on the basis of just a couple stories that were exaggerated.

Amazingly, in this new world of technology, we are getting less information being produced, yet more (false) consensus in the world on what is truth, and therefore worse decisions being made.

The second observation related to an earlier post of mine on the tendency for bloggers to simply link to other sites (see here). Bloggers are not yet fully replacing the investigative reporting role of the mainstream media. They (and I) are serving at best an analytic role, trying to opine on the facts that we assume are being collected by others. This is a classic free-riding situation! Who is going to start collecting the information that the old-style reporters used to collect? Discovering the economic model that will support bloggers actually doing more primary data collection will be a challenge, but the potential economic rewards could be huge.