Affiliation:
1. Boğaziçi University, Turkey
Abstract
This chapter discusses the challenges and opportunities of incorporating computational social science (CSS) into political science. Using an original dataset of quantitative methods courses offered at the top-40 schools, this study shows that CSS courses are currently underrepresented, but their number is likely to augment because of the competition among schools. Yet, this trend is likely to stay confined to North American and Western European schools. Also, collaboration in CSS requires reconciling research priorities in engineering and physics (prediction) and in political science (explanation). Regarding research, big data does not mitigate concerns about causation and representativeness. But simulations allow studying rare phenomena and phenomena for which experiments are not appropriate. By running thousands of experiments at various combinations of explanatory variables, simulations explore counterfactuals, assess the explanatory power of competing theories, and make forecasts.