A New Explanation for the Gender Gap By Scott Blinder and Meredith Rolfe

From the “angry white male” of the Reagan era to the “year of the woman” in national electoral politics in 1992, from “soccer moms” and “NASCAR dads” to “security moms” in more recent elections, gendered differences in party identification and voting behavior have become a staple of popular and scholarly commentary on American electoral politics. Scholars have documented the existence of a gender gap in voting behavior, partisan preferences, and issue positions—more women than men vote Democratic, identify themselves as Democrats, and hold more liberal or Democratic positions on many domestic policy issues, particularly on “social welfare” and military issues.

With the occasional notable exception, two classes of explanation of the gender gap predominate. First, the “compassion” theory holds that women are more compassionate than men, presumably because of different patterns of childhood socialization or experiences in adulthood. This compassion gap leads more women to favor the Democratic Party because of its more compassionate policies on social welfare spending and equality for racial minorities. Second, the “state dependency” theory notes that women are more often economically needy and rely on state provision of welfare. This leads to more liberal issue positions and, in turn, to more frequent Democratic voting and party identification.

In what follows, we move toward an alternative explanation of the gender gap, but, equally importantly, we attempt to dispel these prevailing explanations. In our view, a key flaw in the gender gap literature is that it selectively rejects data that do not exhibit the gender gap phenomenon, notably including opinions of African Americans and other racial minorities.

We attempt to address this deficiency by examining the gender gap in partisanship and issue attitudes among white and black Americans using data from several large national surveys. Two are quite commonly used: the American National Election Study (ANES) and the NORC General Social Survey (GSS). A third, the National Black Politics Study (NBPS) of 2000, is widely used among researchers on black public opinion, as it contains a large representative sample of African Americans. Finally, we use a national survey conducted by Knowledge Networks in 2003 and 2004 that contains a large oversample of African Americans.

According to the compassion theory, heightened compassion leads women to support government spending on so-called “social welfare” programs more often than men do. This difference causes gender gaps in partisanship and vote choice, as more women than men choose the party with more liberal or generous issue positions on these social welfare issues. How can we test these assertions?

In our view, if women’s compassion drives the gender gap, the data should show the following:

We test the first two implications of the compassion theory by comparing African American opinion to white American opinion. As shown in Figure 1, the partisan gender gap is much weaker among African Americans than among white Americans—it appears later and less consistently over time, and is smaller when it does appear. Black men self-identify as Democrats almost as often as black women and much more so than white women.

Of course, African Americans’ overwhelming Democratic partisanship is well known, but the issue positions of black men may be more surprising, and more damaging to the compassion theory. As indicated by the selected items in Table 1, a thorough review of survey questions over all available years on spending on numerous government programs indicates that gender gaps on social welfare issue positions were rare among African-American survey respondents. With the exception of child care spending, African-American men were every bit as liberal as African-American women across a range of social welfare issues. Moreover, they were also much more liberal than white women on all “compassion” issues. Even on the issue of child care, which one might think would be essential to many more women than men, black men were more supportive of spending than white women, and nearly as supportive as black women.

Hierarchy and Identity: A New Explanation for the Gender Gap