If, say, you’re a woman who lives in Louisiana, you’re more likely to know that Mary Landrieu voted for it than you are that David Vitter voted against it.
That’s one insight from a new article (ungated) by political scientist Philip Edward Jones. He finds that women know more about how their female senators vote and are more likely to hold them accountable for those votes than when they’re represented by men. There’s also suggestive evidence that this might be true for male voters as well. Maybe electing women makes for better citizens.
Jones was curious about whether “descriptive representation” — women being represented by women — affected whether female voters paid attention to their senators’ activities. To find out, he used national survey data from the 2006, 2008, and 2010 Cooperative Congressional Election Study, which asked respondents if they knew how their senators voted on a variety of matters, including the Affordable Care Act, extending capital gains tax cuts, and Supreme Court confirmations.
Female voters were more likely to know the legislative records of their female senators than male senators. In 2006, for instance, Jones found that controlling for a host of other factors — partisanship, how long a senator had been in office, and so forth — women knew were able to correctly identify about 45 percent of the votes of their male senators. But when they were represented by women, they were able to say how they voted about 52 percemt of the time.
The votes of female senators were also more consequential for how their female constituents evaluated them. Jones looked at how often voters’ own policy positions matched the votes that their senators cast. He then examined the effect that had on senators’ approval ratings.
To see how this works, imagine two otherwise identical female voters, one whose positions were the same as her senator’s 25 percent of the time, and the other whose positions were the same as her senator’s 75 percent of the time.
The difference in the approval ratings those two voters gave their senator was about 40 points when represented by a man. When represented by a female senator, though, the gap was 54 points. In other words, the congruence between a female voter’s policy positions and her senator’s votes mattered more when the senator was a woman than when the senator was a man.
Jones doesn’t find the same statistically significant gaps when analyzing male voters, which he reasonably interprets as evidence that there’s something unique about women being represented by women.
My read of the data is slightly more expansive, however. Although the effects for women are clearer and stronger in a statistical sense, the overall patterns are quite similar for male voters. And in 2008, in fact, men were significantly more likely to correctly identify the votes of female senators than male senators.
Of course, the absence of statistical significance is meaningful, but I was struck by the consistency of the pattern for both women and men voters. Research examining knowledge of House members’ votes could shed additional light on whether these patterns are specific to women, or to voters in general.
Ultimately, Jones’ findings clearly suggest that being represented by a women makes women more attentive to the legislative records of their senators. On one hand, familiarity may be advantageous for women politicians, since they will be rewarded when they effectively represent their female constituents’ interests. But as Jones notes, greater knowledge could cut the other way if “female politicians face an electorate that is, overall, more attentive to any policy missteps that they make.”