A Closer Look at the Gender Gap

IN the month since the election, I've had a chance to analyze the large election-day survey (sometimes known as an "exit poll") taken by Voter Research and Surveys (VRS), the four-network consortium. One of the most interesting findings to emerge is the extent of gender differences in the balloting.

So far as we can tell, in the first century after Wyoming led the way in extending the franchise to women - the territory did so in 1869 - women and men almost never differed significantly in their patterns of party and candidate support. In the 1930s, for example, the Democrats had large and identical leads over the Republicans among women and men alike.

Beginning in the late 1970s, however, partisan differences by gender began to appear - typically not huge, but persistent. According to a series of Gallup surveys taken in late 1991 and early 1992, the Democrats had an edge in party identification of 4 percentage points among women, while they trailed the Republicans among men by 7 points - a gender gap of 11 points. (Throughout this column, I will compute the "gap" by subtracting a candidate's or party's margin of support among men from that among women .) Similar differences have appeared in congressional and presidential voting, as was evident again this year.

According to the VRS survey, women gave Bill Clinton a 9-point margin, while men favored him over George Bush by just 3 points. This gender gap of 6 points was considerably smaller than it had been in each of the three preceding presidential contests. But the gap this year was large and striking in 10 of the 11 Senate races where a woman faced a man. Women voters are generally less supportive of Republican candidates than are men, and the gender gap refers to this partisan difference. In 1992, however, w henever a female candidate faced a male candidate for Senate, she did better among women voters than among men, whatever her party - although it must be noted that there was only one instance in which a Republican woman ran against a Democratic man.

A number of observers, among them political scientist and CNN analyst William Schneider, have suggested that the gender gap is concentrated among young men and women, and among those with high levels of education. The 1992 VRS findings generally support this interpretation. For example, among elderly voters (60 years and older) on Nov. 3, there was no difference between men and women in party identification. But among the youngest cohort (18-29) the gender gap was large: Republican identifiers outnumbere d Democrats by 41 to 31 percent among young men; Democratic identifiers led 40 to 33 percent among young women.

The VRS survey found no gender gap in party identification among voters with a high school education or less. But for those with more than four years of college, the gap was a very wide 18 percentage points. The same thing can be seen in the presidential vote. There wasn't any gender split in Bush-Clinton preference between men and women of high school training. In the ranks of those with graduate training, though, Mr. Clinton led Mr. Bush by a whopping 25 points, 55 to 30 percent, among women, while he led the president by just 7 points, 47 to 40 percent, among men.

The presence of large gender differences among those with graduate training carried over into House of Representatives voting. Women postgraduates went heavily for House Democrats, 62 to 38 percent. Men with graduate training, however, gave an edge to Republican candidates by a margin of 51 to 49 percent.

It is important not to overstate the gender gap. As noted, it was a modest 6 points in the Bush-Clinton contest. Still, the growth of the women's movement and its increased political activity over the past 15 years seem to have contributed to persisting gender differences in voting (and, as well, on various policy questions). Women's greater entry into the labor force and the sharp rise in single-parent, female-headed households have probably added to the gender gap as well.

This being so, it is not surprising that young women, lacking experience in earlier times when gender differences were not evident politically, show the impact of the new pressures somewhat more than their elders. Similarly, college training, especially advanced graduate education, is a carrier of the newer values.