The real damage is done when public policy responds, especially when the bulk of the population is wrong. Consider smoking: smokers and nonsmokers alike overestimate the risk of lung cancer from smoking (as estimated in the public health literature) by so much that Harvard Professor Kip Viscusi, an expert on risk analysis, calculates that if the public estimated this risk accurately, smoking rates would increase by 6.5 to 7.5 percent. More generally, a majority of the public has been supporting the Nanny State for many decades.  
  
          Why do many intelligent individuals so often mimic the opinions or behavior of others, however mistaken? Why do fads and fashions appear and disappear unexpectedly? Why do people seem to fall in imitation cascades? To answer these questions, economists have developed cascade theories over the last two decades.    
          Informational cascades, the main building block of cascade theory, have been analyzed by UCLA economists Sushil Bikhchandani, David Hirshleifer and Ivo Welch. Cascade theory reconciles herd behavior with the rational-choice approach in the social sciences: it is often rational for an individual to rely on information conveyed by others. The reason is that gathering information is costly, even if only in terms of time. Individuals will (formally or informally) buy information only up to the point where it yields no more net additional benefits than simply using the information conveyed by the behavior or opinions of others. Informational cascades occur when everybody relies on such "public" information.  
  
          It can be shown that wrong cascades — i.e., cascades that go against what most people would do or believe if they were left to their own private information — have a relatively high probability of developing. Another property of informational cascades is that they are fragile: since they are triggered by a small amount of original information and maintained by mere imitation, they can be reversed by a little bit of new information. This is why, like fads in general, public opinion often seems so volatile.  
  
          Cascade theory is relevant to many business phenomena, including management fads, adoption of new technologies and innovations, and the important issue of product reputation. But there is little evidence that cascade theory applies to "financial bubbles," because prices of financial instruments move to render herd behavior less and less attractive. When individuals have an incentive not to follow the crowd, they don't. In cascades, individual incentives remain biased towards herd behavior.  
  
          Besides information, another building block of cascades is personal reputation. Individuals often falsify their publicly-expressed preferences, and follow the crowd, in order to gain social acceptance. When this happens, cascades become more resilient, because individuals have incentives to disregard new information.  
  

« Cascade theory vindicates old classical liberal ideas about the importance of unpopular free speech, and of independent social elites capable of resisting public opinion and the state. »

  
          Consider political correctness. Individuals may use "politically correct" language even if they don't share the implied opinions. In Canada, anybody who wishes to obtain a broadcasting license from the Canadian Radio-television and Telecommunications Commission must present a proposal that uses all the Newspeak buzz words ("diversity," "community," "national"...). If one were foolish enough to plan a non-politically correct TV station, he (excuse me, "she") would have to defend the project with politically correct terms!  
  
          The idea of following the crowd in order to maintain one's social reputation is a major element of Timur Kuran and Cass Sunstein's theory of "availability cascades." (These cascades take their name from the fact that individuals often consider the greater availability of some information as an indication of its reliability.) In a major Stanford Law Review article, the two authors (respectively from the University of Southern California and the University of Chicago) explain why the regulation of risky products or environmental risks is so often driven by public opinion movements that run counter to scientific evidence. Once an availability cascade has started, nobody will want to risk his credibility by going against the trend.  
  
          The impact of special interest groups is increased by their ability to promote preference falsification among their members and supporters. This is a large part of what trade unions, corporate lobbyists, and subsidized public health organizations do. As run by pressure groups, writes Kuran, "politics consists of controlling society's choice" through manipulation of public opinion.    
          Cascades are efficient when they facilitate the diffusion of correct information (say, about a new technological innovation) at low cost. They are inefficient when they feed on false information. To reverse wrong cascades, credible individuals must obtain, and act on, correct private information. When reputational factors are involved, we also need free-spirited individuals who are relatively independent of social pressures and of government payouts. Cascade theory thus vindicates old classical liberal ideas about the importance of unpopular free speech, and of independent social elites capable of resisting public opinion and the state.  
  
          Lots of people can be wrong for a long time. Actual examples include global environmental scares, the jihad against smokers, the deemed usefulness of securities regulators, pidgin business ethics, politically correct corporate governance, and so forth. But there are good reasons to believe that wrong cascades, even supported by special interests, can be reversed by free speech, individual liberty, and the dispersion of power in society. Individuals and groups who defend such ideals are contributing to an efficient economy.