Abstract
This article proposes stochastic, spatial model of roll-call voting as a benchmark for evaluating more complex models of congressional behavior. On the basis of several evaluative criteria, this model outperforms Weisberg's two-party and three-party models and Hammond and Fraser's random-voting models. Particularly, the spatial model successfully recovers the occurrences of hurrah, party unity, conservative coalition, and Civil War votes as well as the marginal distribution of yea votes. At the same time, the spatial model is parsimonious, requiring that more parameters be estimated than the twoparty model does but fewer than the three-party model.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call