Thursday, September 14, 2006

Condorcet Jury Theorem

Here's something interesting I came across while reading for my Public Choice class tonight. It's a theory that explains some of the positive effects of using a simple majority rule over unanimity in decision making:

Let us assume that after hearing all of the evidence in a case that the probability of a judge reaching the correct verdict regarding the accused’s innocence is 0.6. Obviously in trials presided over by one judge, the correct verdict will be reached 60 percent of the time.

A tribunal that employed the unanimity rule would make the correct decision only 21.6 percent of the time. On the other occasions it would either fail to reach a unanimous verdict or would unanimously reach the wrong verdict.

If, however, the tribunal used the simple majority rule, it would always reach a verdict, and would reach the correct verdict 64.8 percent of the time. Moreover, the probability that a panel of judges reaches the correct verdict grows continuously as its size increases – provided that it employs the simple majority rule.

-- Mueller, p. 128-129

Of course, there are other problems that can arise in using majority rule as a decision making procedure, but I thought this aspect of it was non-obvious and fascinating.

One of the biggest detriments of using majority rule for political decision making is tyranny of the majority. A good political institution should use a variety of rules with more concensus being required for weightier decisions.

I am thoroughly enjoying this Public Choice class so far and it is getting me to look at political institutions and decision making in an entirely new light.

It has also been a tremendous opportunity to listen to lectures by two of the chief architects of this school of thought while here at GMU -- Gordon Tullock and Nobel Laureate James Buchanan. In addition to being here at GMU, something else I share in common with these two great scholars is that we've all three done our time at the magnificent Virigina Tech.

Read more on Condorcet's jury theorem here.

No comments: