In a compelling, multimethod study of the relationship between political party organizations and women's representation in state legislatures, Kira Sanbonmatsu provides significant new insight into the puzzle of women's underrepresentation. She offers a powerful critique of the conventional wisdom that strengthening party organizations will lead to an increase in women's representation. She also breaks new ground in the study of parties, calling into question the candidate-centered model of U.S. elections.

Sanbonmatsu begins with the observation that the percentage of women serving in state legislatures has plateaued (22.5% in 2005). As demonstrated by multiple scholars, when women run, women win at the same rate as their male counterparts. Why then, do not more women run for office? The effects of party activities have been largely missing from scholars' lists of possible explanations. Analyzing the role of party organizations in the nominating process and the implications for women's candidacies, the author demonstrates that fewer women run for office in states where legislative parties actively seek to influence nomination contests.

There is no consensus about the role or strength of state party organizations with regard to the candidate nomination process in the contemporary primary system. Sanbonmatsu's first challenge, then, is to assess specific party activities in the nomination process for state legislative office. She disentangles two different mechanisms of party organizational strength—recruitment (seeking out and encouraging candidates to run) and gatekeeping (discouraging candidates from running or taking sides in a primary). She draws upon evidence from 227 personal interviews with state legislators, state and legislative party officials, and other actors, such as women's groups in Alabama, Colorado, Iowa, Massachusetts, North Carolina, and Ohio, in addition to mail surveys of party leaders from all 50 states. She finds convincing evidence that parties actively—and increasingly—recruit candidates for the state legislature. Gatekeeping is also common, though less widespread, as interview data revealed that actually dissuading candidates conflicted with some leaders' commitment to an open electoral process or brought about political risks.

Sanbonmatsu uncovers more party activity in the nominating process than conventional wisdom suggests. Readers, however, should not overlook the interesting and sometimes large differences among states, parties, and types of races that emerge from the interview data. Parties do not play a central role in the nomination process in Massachusetts or Alabama. In four of the six states, party leaders work with interest groups to some degree, but groups only have real influence in Alabama. Not surprisingly, parties usually target the most competitive seats. State and legislative party leaders in five of the states may work to prevent a primary, but only in Ohio do they actually become involved in primary races. Legislative caucuses and campaign committees play a greater role in party recruitment than do state parties.

After demonstrating that parties typically recruit candidates to run for the legislature, Sanbonmatsu investigates how gender factors into their strategies with survey data from legislative and party leaders from all 50 states. While almost half of those interviewed reported no advantage for men or women, the remaining respondents were closely divided in their opinions about which sex possesses an electoral advantage. More telling, over half indicated that there were at least a few state legislative districts where it might be hard for a woman to win an election. Leaders stressed the importance of finding the best candidate for the district, regardless of gender. But in a survey of Ohio legislative candidates, more than half of the female candidates indicated that the party is more likely to encourage men than women to run, while only 5% of the male candidates concurred. The rich interview data underscore that gender dynamics in recruitment practices vary by leader and by race, and, taken together, offer both promising and discouraging signs regarding increasing women's representation in the state legislature. Leaders in Ohio and North Carolina were more likely to suggest that party recruitment practices are a barrier for women, citing examples of the “old boys' network” and concerns about women's electability. At the same time, female party leaders and women's groups work to mitigate these negative effects.

In separate multivariate analyses of the effects of party gatekeeping and recruitment activities on the percentage of women party nominees in a state, Sanbonmatsu shows that the effect of legislative (but not state) party gatekeeping is negative and statistically significant. When the party is concerned enough about a race to become involved, the combination of informal networks that advantage men and doubts about women's electability will increase party support for male candidates. Contrary to her expectation that active party recruitment would increase the number of female candidates, she finds no systematic effect.

The distinction between party recruitment and gatekeeping activities may give some readers pause. While the author finds that parties engage in these activities to different degrees and their effects on women's representation differ, she acknowledges that “recruitment may imply gatekeeping” (2006, 30). Indeed, the evidence she presents reveals that some party organizations pursue a combined strategy with respect to targeted, competitive races—the races most likely to result in new members of the legislature. Including the results of an analysis of women's representation in states where parties employ a combined strategy might have strengthened the analysis. Additionally, a comparison of the success of women's candidacies in the approximately one-third of states that use preprimary nominating conventions with women's success in states without preprimary conventions would have provided additional insight into party activities and the role of party activists in particular.

Where Women Run makes significant contributions to the literatures on women and politics, political parties, and state politics. While readers should not overlook the positive developments that Sanbonmatsu uncovers in some states, such as women party leaders' efforts to recruit women candidates, they should also be aware that she convincingly adds party gatekeeping to scholars' collective list of explanations for women's underrepresentation. Her analysis not only helps to explain the paucity of female state legislative candidates but also sheds light on the lower level of women's representation at the federal level, as the state legislature is the best pipeline to the U.S. House of Representatives.