Almost two and a half months have passed since the Iowa caucuses, and we seem (at least on the side of the Democrats) to be at least another few months away from choosing a presidential nominee. We hear the constant debates over February momentum and superdelegates, but does this painfully drawn out process to select a party’s best candidate actually lead to different results than one simultaneous nation-wide election?

Well, as we know from the study of information cascades, early decisions can have a profound impact, for better or for worse, on later choices. But how do cascades affect sequential elections? Do later voters inefficiently follow the herd and pick a candidate that they don’t like, or do the inferred signals of previous voters lead to a more informed decision. In “Information Asymmetries and Simultaneous Versus Sequential Voting”, Rebecca B. Morton and Kenneth C. Williams asses the effect of different voting processes on electoral results, making use of many information cascade concepts to analyze the qualities of each system.

The results of experiments show quite strikingly that sequential elections do lead to different voter choices than simultaneous elections. Furthermore, the study finds that the information cascade of sequential voting, by allowing later voters to infer the signals of earlier voters, actually lead to voters choosing candidates that better reflect their preferences, a “better” electoral outcome. While simultaneous elections generally favor better-known candidates, sequential elections allow lesser-known candidates to gain momentum. This sequential process often leads to a more moderate nominee, a choice that makes more voters content.

So, as we sit through the remaining months of the nominating struggle, we can console ourselves with the fact that chances are we’ll eventually make the right decision.

Here’s the original article:

“Information Asymmetries and Simultaneous Versus Sequential Voting”

Rebecca B. Morton and Kenneth C. Williams

http://www.jstor.org/view/00030554/di011604/01p0355r/1?frame=noframe&userID=80540094@cornell.edu/01c0a848740050ba75b&dpi=3&config=jstor