What political science needs is a good dose of mathematics, according to SFI Omidyar Fellow Nathan Collins.
How do people choose whether to vote? Why do they prefer one candidate over another? How do they form their beliefs about political candidates? Mathematical models, Nathan says, can help answer all these questions.
Take voting behavior, for example. Like a bowling ball, voters are subject to inertia: They tend to vote if they voted last time, and they tend not to vote if they didn’t. Unless, that is, they’re dissatisfied – then they switch strategies. When expressed mathematically, this principle alone can explain long-term trends in voter turnout remarkably well, Nathan found.
He created a super-simple model in which all voters are dedicated Republicans or dedicated Democrats, and they decide only whether or not to vote. Voters whose party consistently wins may decide voting isn’t worth the bother. Conversely, non-voters whose party loses may get prodded into voting.
Of course big external events like a war or an economic collapse can impact voter satisfaction too, and Nathan’s model can’t predict those. During stable periods, though, he argues that his model captures much of the dynamics driving long-term trends.
To test the model he examined voting data from six countries during the relatively stable period between the end of World War II and the beginning the oil embargo. “Our model was a closer t than any other reasonable model,” Nathan says.
The spike in voter turnout in the U.S. in the last election, he says, may indicate that recent events have changed basic satisfaction rates, leading to a longer-term change in voter turnout.
Nathan earned a master’s degree in physics at MIT before tackling political science for his PhD at Stanford.