American Presidential Primaries and Information Cascades
An Information Cascade occurs when an individual bases his decision on his observation of the actions of other people, rather than his own information signals. Social observers have long recognized that this effect rises from human beings’ deep-rooted proclivity to imitate. Countless models, spanning many fields ranging from politics to finance to medicine, exhibit the effects of such cascades. One such model is the American Presidential Primaries process.
The notion of momentum in elections has always received much attention. It has often been argued persuasively that early victories make a huge difference in the prospects for any candidate. This faith is reflected not only in the disproportionate amount
of campaign funds that are used to convince early voters, but also in the data gathered regarding voting behavior and polls over the years. In a two party electoral system, the voter set can be divided into three groups : supporters, neutral and opposition-supporters. Assuming the first and the last group are convinced about their choices, it is just the neutral voters who get swayed by the results of prior elections as they feel that others are more informed than them and hence take their views into account while making their decision. The neutral voters know that the candidate will always receive votes from his supporters, no matter what, and hence they only take into account how many neutral voters swayed in favor of a particular candidate. If a candidate wins in a state where he is anticipated to have little support as compared to his rival, it is a clear indication that the neutral voters voted for him based on their own opinion (cascade has not started yet as it is early stage of elections). Thus such a surprise victory improves the candidates standing with remaining neutrals. By the same logic, a win in a state where he is anticipated to have lots of support suggests nothing about neutral voters and thus does not initiate a cascade. Moreover, ‘surprise’ victories are given more media coverage and attention, and consequently increases their impact on remaining neutral voters.
The impact of ‘surprise’ victories is exemplified in Jimmy Carter’s victory in the 1976 Democratic primaries. Carter, a candidate from the South, was believed to have little support in the state of New Hampshire, yet he won the elections, thus demonstrating to the voters, the financiers and the media that he could win without his partisan support in the South and that he was the right candidate.
History also indicates that victories in a candidate’s home state or where partisan support is highly likely have less of an impact on the behavior of remaining neutral voters. In the 1984 Primaries, Walter Mondale of Minnesota was expected to have much support in the neighboring state of Iowa. Thus despite being victorious in the caucuses, he was largely overshadowed by Gary Hart. The fact that Hart was able to garner 16.5% of the votes in Iowa generated a burst of interest from voters and the media. By performing better than anticipated, Gary Hart was suddenly perceived as a better candidate.
Links
1. http://www.dictionaryofeconomics.com/article?id=pde2008_I000103
2.http://levine.sscnet.ucla.edu/archive/refs4782713000000000062.pdf?origin=publication_detail