A Formal Analysis of Legislative Turnover: Women Candidates and Legislative Representation

A regular Markov process relating a group's candidates to that group's proportion in a legislature is specified with a set of assumptions and definitions. The model is relaxed to allow for bias relating to incumbents and new candidates. The model is then applied to women candidates. Analysis shows that in the absence of bias the proportion of women in the legislature eventually will equal the proportion of new women candidates. The rate at which the proportion of women in the legislature approaches the proportion of women among new candidates is dependent upon the incumbency factor: the larger the incumbency factor the slower the rate of approach. Bias relating to new women candidates will serve to increase or decrease the stabilized proportion of women in the legislature. Bias relating to women incumbents has more complex effects, but generally negative bias has less impact than positive bias on female legislative membership. Increasing the return rate of female incumbents can affect the level of female representation more than increasing the number of women candidates.