Political science isn’t science

last updated 8 January 2015

Jacqueline Stevens

The New York Times

Is political science really science? asked Jacqueline Stevens. Congress recently decided it wasn’t, and voted in May to eliminate $11 million in government grants for political science research. My fellow political scientists are outraged, but Congress is right. In the past, federal grants have gone mainly to researchers who analyze “probability studies and statistical significance” to make political predictions, often for think tanks and government agencies. And in that forecasting work, “my colleagues have failed spectacularly.” No political scientist, for example, foresaw the fall of the Soviet Union. Nor did they predict the rise of al Qaida, or the Arab Spring. Domestically, well-known political scientist Morris P. Fiorina predicted in 1992 a “persistent pattern” of Americans electing Republican presidents and Democratic Congresses. Voters quickly proved him wrong. Psychologist Philip E. Tetlock famously found in the 1980s that political scientists were no better at prophesying the future than “chimps randomly throwing darts” at a range of outcomes. So let’s stop pretending that anyone can somehow divine the future of a complex world using “contrived data sets and models.” Chimps who throw darts don’t deserve taxpayer-funded grants.