Jeffrey Lazarus and Amy Steigerwalt argue that “gendered vulnerability” accounts for a variety of legislative behaviors that distinguish female members of the U.S. Congress from their male counterparts. Gendered vulnerability refers to female candidates’ perceptions that they are at risk of losing their bids for office. Women congressional candidates expect to be subjected to skepticism about whether women can do the job and other gender-related stereotypes, in part because women are socialized to doubt their own abilities. They face obstacles other than their perceptions, including frivolous news coverage of their appearance rather than the substance of the campaign, more primary challengers, and more serious general-election challengers than male candidates, simply because they are women. These doubts and obstacles are experienced by female incumbents as well, even when they have won previous elections by comfortable margins. The authors offer an extended argument for the existence of gendered vulnerability based on evidence from political science, psychology, and other fields on women in the workplace, interviews with former members of Congress and congressional staff, and vignettes about specific women in elections. Establishing that women in Congress are subject to gendered vulnerability leads to predictions about how their legislative behaviors should be affected.

The authors ask whether women are more likely than men to engage in activities that would limit their electoral vulnerability, specifically those that David Mayhew (The Electoral Connection, 1974) identified as fostering reelection: credit claiming, position taking, and advertising. The authors also ask whether women members are more responsive to constituency interests. These activities are covered in four empirical chapters, in most instances using data from the 101st through the 110th Congress (1993 through 2009). The start date coincides with the gains in women’s representation in Congress in the 1992 election. Quotations from the authors’ interviews illustrate the analysis.

The findings of multivariate analyses indicate that women House members send more franked mail than do male representatives, although this is not true for senators. No gender differences are found with respect to spending on travel to the district for House members (a factor not examined for senators). Women Senate members (but not House members) employ more district staff than men, allowing for more casework. Analyses of earmarks in 2008 and the Obama Stimulus Package in 2009 show that women in both chambers procured more and higher dollar earmarks for their constituencies and that women in the House (but not the Senate) sent more stimulus dollars home. Moreover, women representatives were attuned to constituency need, with more stimulus aid going to districts with higher poverty rates.

Analyses of bill and resolution introductions and cosponsorships show women to be more active than men in both chambers. Evidence for whether women utilize more “messaging” bills than men (messaging is inferred, rather than coded directly)—and thus have more opportunities for position taking that they can advertise to constituents—is strong only for House members. Committee membership analysis shows women to be more likely than men to sit on committees with jurisdiction of direct relevance to constituencies, with Senate results that are far more robust than those in the House. In contrast, findings on the policy content of bills introduced are strong for the House (women more likely to introduce bills in areas of direct relevance to the district), but null for the Senate. DW-NOMINATE scores are used to create a measure of the proximity of roll-call voting to constituent preference, and both bivariate and multivariate analyses indicate that women members deviate less from constituency than do men, with stronger results in the Senate.

In general, women invest more time in communicating with constituents, secure more pork for the district, take more positions to send messages to constituents, and vote more in line with constituency preferences. Women are working to engage in activities with an electoral payoff, and benefiting the constituency in the process. Whether they are substantively better representatives than men depends on which side of the trustee-delegate divide one favors, but women pay greater attention to constituent demand as measured here. On the basis of their interviews, the authors speculate that women in Congress are more pragmatic and likely to engage in bipartisanship than men, and so an increase in their presence should dampen the effects of polarization that leads to gridlock.

The strong point of this book is the idea of gendered vulnerability and the argument that it is pervasive and subjective. The phenomenon is well documented in the wide variety of literature that the authors cite and is likely to resonate with any female academic or student reading the book. She has been there. Predicting women’s behavior based on this argument moves the study of women legislators in a direction beyond the notion that the utility of having women in legislatures is that they will provide better substantive representation for women—a group that is by no means homogenous in preferences. The use of predictions of multiple types of behaviors where women should differ from men is a strength, as is the use of multiple measures for several of those behaviors.

The book is not without flaws. Much of the analysis depends on data pooled across Congresses. This is an efficient means of presentation, and possibly necessary to overcome the relatively small number of women in a single Congress. But the results of pooled analysis can mislead by artificially inflating significance tests. I would like to know if the results on gender held up when Congresses were analyzed individually. The authors offer no discussion of the validity of many measures, and some are questionable. For example, sponsored bills that do not go far in the legislative process are deemed to be messaging bills (analyzed on pp. 127–35). But are they? It would be impractical to code each of these thousands of bills, but a sample could be selected and coded to establish the validity of this contention. Or, why use the poverty rate in a district rather than the unemployment rate as a measure of need for the economic stimulus bill (Table 4.4)? Tables 6.2 and 6.3 are particularly hard to follow because the same symbol is used to denote disconfirming results and the absence of observations. More seriously, there is a tendency to overstate the results—mixed results between chambers are sometimes downplayed in the individual chapter conclusions. If the mixed results were always in the same direction (e.g., confirming results for the House but not the Senate), I would be less bothered by the fact that sometimes the results are confirming for the Senate but not the House, or vice versa. There is also a tendency to offer post hoc explanations for results that do not support the predictions. And some will find an irony to the conclusion that women are better legislators based on the Mayhewian activities examined - Mayhew saw the emphasis on pursuit of reelection (or the cultivation of the personal vote) as anathema to good lawmaking.

Despite these criticisms, Lazarus and Steigerwalt have made a useful contribution to the literature on women legislators. They offer a new thesis, ask new questions, and use a variety of data to address them. Gendered Vulnerability opens new avenues for research on women in politics generally.