Bayes’ Theorem in Politics
This article uses Bayes’ Theorem to evaluate Emerging Democratic Majority (EDM) theory. EDM theory states that recent changes in society are causing a push towards support of the Democratic party over the Republican party. The idea of realignment in general suggests that roughly every three decades, one party becomes dominant over the other. EDM theory applies this to the modern day, arguing that we are currently going through a Democratic realignment.
Bayes’ Theorem states that, given two events A and B, P(A|B)=(P(A)*P(B|A))/P(B). This article specifically uses Bayes’ Theorem to evaluate how the new data provided by the current election should be interpreted as it relates to EDM theory. In other words, if the results of the election show support for the Democratic party, how likely is it that we are in fact undergoing a Democratic realignment, and vice versa. The article uses P(H) to represent the prior probability of realignment and P(E|~H) to represent the probability of a false positive (the data predicts EDM theory to be accurate, when it reality the data was caused by some other explanation.
The author uses an example with weather predictions to show just how surprising the implications of Bayes’ Theorem can be. Consider an area where it rains only 5 days per year. If a meteorologist with a 90% accuracy rate predicts rain the next day, there is actually only an 11% chance it will rain that day. While 90% accuracy seems high, the fact that rain is highly unlikely means that a prediction for rain is probably a mistake by the meteorologist.
The trouble in using Bayes’ Theorem to evaluate EDM theory lies in coming up with the probabilities to plug in. For example, an EDM theorist might plug in values that result in the argument that evidence of Democratic support corresponds to a 97% chance of an Emerging Democratic Majority. A skeptic, however, might plug in a higher probability of a false positive, yielding a significantly smaller chance of supportive evidence corresponding to an actual Emerging Democratic Majority.
The insight this article provides into Bayes’ Theorem relate to our discussion in class of the theorem and how even doctors often misinterpret probabilities since the theorem yields such unintuitive results.