- UID
- 279019
- 在线时间
- 小时
- 注册时间
- 2007-9-27
- 最后登录
- 1970-1-1
- 主题
- 帖子
- 性别
- 保密
|
Professor Cao Huining: Does Conversation Help Us Make Better Choices? How do people make choices to invest in a certain company, or buy a particular product? Do they make rational choices or do they just follow the crowd? Is following the crowd the most rational thing they can do? And do their choices get better if they witness the outcome of decisions others have made?
--------------------------------------------------------------------------------
CKGSB Professor of Finance Huining Cao BEIJING – 15 July 2011 – In a forthcoming article in the Journal of Economic Theory, CKGSB Professor of Finance Huining Cao, Bing Han, and David Hirshleifer show that people mostly follow others even when they are wrong. Moreover, witnessing and discussing the outcome of other peoples’ decisions doesn’t necessarily stop us making bad choices. Jumping on the Bandwagon The herd mentality - jumping on the bandwagon, following fads and fashions – is a human characteristic that behavioral scientists are keen to get to the bottom of. In finance for example, it is a key reason for stock and asset bubbles and crashes. Although many such actions seem irrational, it is still possible to gain useful insight into why they happen. Behavioral scientists do this by bringing together theories from psychology and economics, and statistically analyzing their findings. For Cao, Han, and Hirshleifer, this involved defining the ideas to be examined, making assumptions and propositions, and working out the probability of particular outcomes. The Information Cascade Rather than making individual choices, people are often drawn towards taking certain actions by those around them, by observing their behavior and its outcomes, and engaging in conversation. This is the social learning part of decision-making. It is seen as sensible to do what other people are doing, as their actions are presumably based on some favorite information. In turn, those who follow us believe our actions are informative, too. When this happens to a chain of individuals time and time again, it's called an information cascade. The problem with information cascades is that they often result in uninformative outcomes as information cascade does not aggregate the information of all participants in the cascade. Just observing the actions of others and making the same choices based on those observations alone can lead to arbitrary or wrong decisions. Often, a long time has passed since someone actually examined the logic behind following a particular path. The lack of information from alternative choices means that further down the line, people have even less on which to base a potentially dissenting decision. "I don't know why I did it, it's always been like that," sums up this line of thought. Applied to China today, the popularity of Louis Vuitton bags is a good example. Why, despite the fact that other cheaper bags offer the same quality, does LV sell so well? Rationalizing the Sheep Mentality Maintaining the view that people are rational, scholars have come up with a number of reasons why information cascades continue despite repeated sub-optimal outcomes. Often, they are taken not as the result of individual weighing up of costs and benefits and actively selecting the path to greatest personal benefits, but simply by following "received opinion." When people are stuck on the wrong choice, new information from payoffs, which help explain the real value of any project, sometimes jostles subsequent players into experimentation.
Many have suggested that something blocks our access to the independent information we need to base decisions on. The something appears to be the behavior of the people who made decisions earlier on. The information cascade theory has been analyzed with models permitting different degrees of access to information surrounding a decision. In earlier analysis, "society is trapped in mistaken choices until an external shock arrives" as nothing but the previous decision is known. In their study to be published in the Journal of Economic Theory, Cao, Han, and Hirshleifer feed additional information into the model. They allow for observation of and conversation about payoffs. Given this extra information, they find that the probability of an information cascade lasting forever is lower. When people are stuck on the wrong choice, new information from payoffs, which help explain the real value of any project, sometimes jostles subsequent players into experimentation. Information Blockage However, society can settle into making continued wrong choices that informational shocks are not strong enough to dislodge. Information blockage needs only limited restrictions to observation or conversation in order to occur. Information is aggregated inefficiently because of information externalities. The first is that individuals fail to see the benefit to later decision-makers of adjusting their actions. The second is that the quality of payoff information affects the quality of later decision-makers' choices. Positive developments in society are harmed by a lack of experimentation on "the road less travelled by." Take the example of scientific research grants. Although individual researchers may benefit from having a small number of broad funding channels, society will gain from having more diversity of funding organizations. This will enable more research into less frequently investigated areas. A Little Less Conversation Observation of payoffs does not necessarily help reduce bad decision cascades. The earlier payoff information becomes known, the sooner a re-enforcing cascade will start up. Conversely, if there is less conversation at the start, more information will be available for aggregation by later players before "cascades clog the flow of information." Moreover, conversation about payoffs doesn't seem to help. Conversation is often "noisy," meaning that it lacks factual substance. Engaging in conversation makes it hard to pick out the real payoff information. The rise of the mass media and of interactive communication technology has made it easier to observe the payoff outcomes of others. Although this seems to offer the decision-maker more information on which to make a choice, this is true only at the beginning of a chain of decision-making. For later actors, decisions further up the chain are seen as opaque. For these people, it makes sense for them to follow prior decisions. The void of information about alternatives means that they might be stepping off a cliff if they don't. Abstract:
Taking the Road Less Travelled: Does Conversation Eradicate Pernicious Cascades?
We offer a model in which sequences of individuals often converge upon poor decisions and are prone to fads, despite communication of the payoff outcomes from past choices. This reflects both direct and indirect action-based information externalities. In contrast with previous cascades literature, cascades here are spontaneously dislodged and in general have a probability less than one of lasting forever. Furthermore, the ability of individuals to communicate can reduce average decision accuracy and welfare. |
|