In class, we discussed the fact that if certain restaurants get mentioned by many people, you discover that some of these people had mentioned the majority of these highly-recommended restaurants when you asked them about new restaurants. Since you now know that these people have valuable information in terms of restaurant selection, you trust their judgment more and their opinions are given more value. In fact, one will now venture out to try the obscure restaurants these people have sugested because now their judgment can be trusted.

This phenomenon of observing what others have done before you and following their example, regardless of your own private information signals, is known as an information cascade. You do not know from personal experience that these obscure restaurants are good. Most likely, your own private information signals were steering you away from these restaurants before these “restaurant experts” suggested the restaurant to you. However, since you are listening to the advice of certain “restaurant experts”, you decide to eat at the restaurant. Since it is usually sensible to do what other people are doing, your behavior is governed by rational choice.

These cascades of information can be broken easily because new information will be able to overturn long-standing decisions.For example, a particular “restaurant expert” might be able to persuade a group of people to go to a particular restaurant everyone else in the group views as obscure if and only if this “restaurant expert” has suggested good restaurants in the past countless times.

The phenomenon of information cascades has been seen in voting trends as something called the bandwagon effect. For example, people vote for the particular candidates who are likely to succeed, based on what the media is telling people, because they want to be on the winning team. Since voting booths in California close three hours after voting booths in New York close, voters in California have a chance of knowing who won an influential state such as New York before going to the voting booths.

Therefore, voters in California will be influenced by the decision of the New York voters.

This phenomenon has a more heavily felt influence in primary elections because in these, certain states vote a month ahead of other states. These states that vote early in the primaries, such as Iowa, have more influence in the overall process than states who vote later.

A study in The Journal of Politics (http://www.jstor.org/journals/00223816.html?cookieSet=1)

explored the voting patterns of 180 college undergraduates at the University of Kentucky. These students were assigned randomly to nine different groups and were questioned about the exact same set of election scenarios. Almost 70% of participants received information about the expected winner. Independents, who had no prior political party affiliation, were strongly influenced towards voting for the candidate who was expected to win.

If a particular candidate is initially favored by a small margin, poll results showing this candidate as the leader of the race at that particular point in time will cause the margin to increase. If these poll results continue to be repeated, the effect of the information cascade will tend to accrue and become very beneficial to leading candidates.

If a candidate has a very large majority of the votes of the rest of the country, the news media will not wait for Californians and other people on the West coast to vote. For example, in 1992, NBC News proclaimed Ronald Reagan as the winner of the presidency several hours before the voting booths even closed in the west, based on the exit polls of the rest of the country. In fact, all of the states on the West coast voted for Reagan, but their decision was most probably influenced by the NBC news station’s proclamation. Since Californians continued to see poll results with Reagan as the winner, they realized that if most everybody else in the country had already voted for Reagan, Reagan was the winning candidate regardless of who they voted for.