The Gender Gap Mystique

E ver since the Greek poet Aristophanes in 411 B.C. fired men's imaginations with the play Lysistrata, the possibility that women might band together for their own purposes has evoked strong feelings, ranging from ridicule to apprehension. Until the twentieth century, the prospect of such female solidarity remained largely fictional.

In this century, the anticipation that American women, armed with the vote, would create a unified political force, evoked similar reactions. For example, in 1921 when the newly formed League of Women Voters moblilized to oppose New York Sentor James Wadsworth, his ally, Governor Nathan Miller, lambasted the women for introducing "sex antagonism," which he described as a "well-known socialist tactic" used by the Bolsheviks. These women, he said, were "a menace to our free institutions and to representative government." The governor singled out the evils of bloc voting. There was no place in American politics, he said, for "a league of women voters," just as there was no room for "a league of men voters."

The governor need not have worried. Women remained much less inclined than men to vote at all, let alone to use the vote for their own purposes. From 1920 until 1968, when women began to vote in numbers comparable with men, little attention was paid to women voters by candidates, political parties, or even political scientists. A remark made by a woman suffragist in the 1920s remained apt for nearly half a century: "I know of no politician who is afraid of the woman vote on any question under the sun."

The recent introduction of the term "gender gap" into the American political lexicon suggests a renewed possibility of women's political solidarity. The National Organization for Women coined the term in response to exit poll data from the 1980 presidential race that showed women's support for Jimmy Carter exceeded men's support by a statistically significant margin of nearly 8 percent. (Women gave 45 percent of their votes to Carter, while men gave him 37; women gave 46 percent to Reagan, while men gave him 54 percent.) Most analysts in 1980 had focused on the 1 percent difference in how women voted, not on the wider gap between the way men and women voted.

Initially, the gender gap was assumed to be feminist because it was identified during a presidential campaign in which the Republican Party renounced its 40-year support for the Equal Rights Amendment and its nominee, Ronald Reagan, inveighed against abortion rights. It was presumed to favor women candidates. Some predicted that the Democrats were poised to reap the benefits since they were arguably better than Republicans on feminist issues. Still others found in the gender gap political support for the proposition that women speak "in a different voice."

Over the last decade, gender has correlated with partisan preference or election outcomes in enough high-profile elections to give credence to the belief that women are becoming a voting bloc. In congressional races, more women than men have preferred Democrats by a consistent three to five point margin. In 1990, for instance, the difference was 52-48. Women's votes put Democrats Barbara Roberts of Oregon and Ann Richards of Texas into governorships in 1990. In 1992 the four women elected to the U.S. Senate received a majority of women's votes.

Yet the gender gap turns out to be far more complex than first presumed. As a construct, it is unable to capture many dimensions of women's new influence on elections. Indeed, the gender gap's dominant characteristic is its elusiveness, materializing in some races but not others, sometimes waxing and sometimes waning during an individual race. When it does emerge, it is almost never based simply on a candidate's own gender or feminist views. Women voters do not automatically favor women candidates. Nor is the women's vote a monolithic voting bloc.

Nonetheless, women's votes are more important than ever, and over time they have had the effect of electing more women to office, increasing the electoral salience of issues that interest women voters, and reshaping the policy agenda.

H ow can both things be true? At the most basic level, since women now comprise 54 percent of the electorate, all candidates, male or female, must compete for women's votes and take care not to alienate them. A gender gap may not be stable or robust, but the fact that a gender gap can emerge at all gives women voters new clout. As one political consultant put it, "Any candidate down ten points with women has to worry."

×

Gender and Party. Despite evidence that women voters have a slight preference for the Democratic Party, Republicans have won more of women's votes in three of the last four presidential elections (Reagan twice and Bush once). These feats were not accidental but the product of strategic targeting and mobilization, induced by the need to win women voters and hence to take them seriously. Republican strategists, not Democratic ones, have proven astute at disaggregating the supposed women's vote into as many as 64 subcategories. In national races, Democrats have been more concerned with keeping blue-collar and white male ethnics in the party than with understanding women voters.

Gender and Issues. While the predominant concerns of women voters, like male voters, are economic, the new competition for women's votes has raised the political salience of broadly defined women's issues. These include any issue on which polls show women "care" more than men (health, welfare, child care); women's rights (reproductive rights, equal opportunity, anti-discrimination laws, and so on); and issues that affect mostly women (breast cancer funding, enforcement of child support obligations, and so on). The categories themselves indicate the complexity of defining a women's issue. The latter two, despite their close identification with women, may or may not reveal a gender gap since both male and female voters sometimes support women's issues. In any campaign, opposing male and female candidates may battle each other over who has the better record on women's issues.

Gender and Candidates. From 1920 to 1968, politicians had few incentives to listen to women voters or to mobilize women as a constituency. In a 1990 interview, Margaret Chase Smith, long the Senate's sole woman, discussed women voters as she looked back over her long congressional career, which included the House and a primary race for the presidency in 1964: "Had I been a woman candidate, I wouldn't have been elected." Comparing women with other groups whose members did not make up a numerical majority, such as blacks or labor, she said bluntly, "No minority could be elected."

With Smith's disclaimer that she "never was a woman politician," she more than just rejects feminism. She is making a strategic judgment--widely shared--that women voters, as a numerical minority, couldn't deliver the electoral goods.

For years, three cardinal rules have governed women's campaigns. Don't call attention to yourself as female (dress as "like-male" as possible). Don't single out women's issues. Don't try to mobilize women voters as a separate or special constituency. With the exception of a handful of self-identified feminists, the ubiquitous disclaimer by women candidates that they were not "running as a woman" captured the fear of violating these rules.

By 1992 women candidates running for top-level offices were largely freed from these constraints. Women candidates can now put women's issues at the forefront of campaigns and remain viable. Victorious Senate candidates Dianne Feinstein, Barbara Boxer, Carol Moseley-Braun, and Patty Murray, along with scores of female congressional candidates, unabashedly mobilized women voters and made the absence of women in politics a central campaign theme. While this new freedom might be generally attributed to the modern women's movement, feminism is not the primary basis for their success. The relationships between women candidates, women's issues, and women voters are much more complex.

I n the early 1970s, the idea of female solidarity infused the rhetoric of feminism's early leaders. When the National Women's Political Caucus was founded in 1971, Betty Friedan was among those feminists who predicted that a powerful new consciousness would create a new majority vote, composed of "women who are Democrats and Republicans, women disgusted with both parties and young women turned on to neither. . . . [who] would upset all the old political rules . . . [and] that bosses won't be able to contain nor polls predict."

Throughout the decade, however, these claims repeatedly failed to translate into votes: the Equal Rights Amendment languished in state legislatures, and the women's vote failed to punish offending legislators. As late as the 1980 presidential primaries, Joanne Howes, on Senator Edward Kennedy's campaign staff, searched for evidence of a women's vote in the hope that it would elevate the significance of female campaign operatives as well as provide a strategy to peel votes away from President Carter. She found little evidence.

ENTER THE GENDER GAP

The prospect of an identifiable women's vote was to await Eleanor Smeal's analysis of the Carter-Reagan election exit polls. In addition to highlighting the eight-point gap between women's and men's preferences in 1980, NOW demonstrated from other polls significant gender differences in attitudes toward war and government spending on social programs, as well as assessments of the Reagan presidency.

Sure that these differences were the product of the modern women's movement, Smeal prophesied what earlier men had feared: "Reagan can be defeated on the women's vote alone [in 1984]." Opinion polls showed that only 38 percent of female voters thought Reagan was doing a good job. A 10 percent difference between men and women voters was all that was needed to defeat Reagan.

This analysis helped persuade Walter Mondale to select Democratic Congresswoman Geraldine Ferraro as his running mate. Paul Kirk, chairman of the Democratic Party, and Lane Kirkland, head of the AFL-CIO, blessed the Ferraro candidacy on the grounds that she would intensify the gender gap. But after a women's vote failed to materialize in 1984, interest in the putative clout of women voters waned again. While exit polls indicated Ferraro neither helped nor hurt the ticket, disgruntled Democrats labeled women just another special interest and held them partly responsible for Mondale's defeat.

Yet periodic evidence of a gender gap continued to appear. In 1988, for instance, George Bush appeared to be in trouble with women voters. Six months before the November election, women preferred Governor Michael Dukakis over Bush by a substantial 16 to 18 percentage points, similar to their earlier disaffection with Reagan. By November, however, Bush had won not only the election, but like Reagan before him, the majority of women's votes.

REPUBLICANS "GET IT" FIRST

The key to understanding the Republicans' success with women in 1984 and 1988 lies in their careful targeting and mobilization of them. In a sense, Republican strategists paid women voters the respect of taking their influence more seriously than Democrats did--if only to frustrate feminist aspirations and issues.

A special unit, headed by Michael Deaver, was set up in the White House. Party officials briefed operatives and candidates on the changing demographics of American voters and especially of American families. Secretary of Labor Elizabeth Dole was charged with publicizing Reagan's record on women's issues. According the Dole, "This was the first time we'd had a lot of assistants to the president working on women's issues--child care, dependent care, issues like enforcement of child-support laws."

Richard Wirthlin, pollster to Reagan and Bush, divided women into 64 subcategories. Ads emphasizing economic issues were targeted to the three groups of women identified by Wirthlin as persuadable--single working women, married working women, and elderly or widowed women. Reagan also campaigned at hospitals and other workplaces that were predominantly female to reinforce the campaign message that Republicans were the best bet for a healthy economy and continued employment. Reagan remained comfortably antiabortion and still won a majority of women's votes.

By contrast, the Democrats, after selecting Geraldine Ferraro to strengthen the women's vote, failed to mobilize subgroups of women, believing Ferraro's presence on the ticket would sufficiently motivate women to vote Democratic. The Mondale camp discouraged Ferraro from campaigning on women's issues and made just one TV spot aimed at women. (It was made so late in the campaign that it ran mostly as a news item rather than as a strategic pitch.)

A replay of sorts occurred in the 1988 campaign. Pollster Linda DiVall, who helped Bush overcome his deficit with women voters, said women initially disliked Bush because they saw him as "somewhat removed from what they were going through on a day-to-day basis." First, the Republicans used the national convention to showcase the Bush family. "Successfully," says DiVall, "since the numbers that moved the most after that speech were with women." Then in speeches aimed at women, Bush developed the theme of economic empowerment and argued that private-sector job creation was the key to economic security. He also focused on personal security and crime. Women initially thought Dukakis was tougher on crime, but by the campaign's end, polls showed them convinced that Bush was tougher.

The Bush campaign targeted young women, single working women, middle-aged, married working women, and elderly or widowed women as those who would be most receptive to his campaign messages. As with Reagan, Bush felt no need to articulate "feminist" issues and remained strongly against abortion. Meanwhile, the Dukakis campaign, like Mondale's, had no strategy to mobilize women voters. The campaign organization itself reflected a sensitivity to the charges of special interests that had plagued Mondale (desks were organized by geography, not constituency).

Most campaign operatives would argue that segmenting women voters, as Wirthlin and others have done, is not particularly unique--segmenting male voters is common practice. And that may be precisely the point: for the first time in history, women voters are being targeted and mobilized because, as a majority, their votes cannot be taken for granted any more than men's can.

As women have entered the work force in greater numbers, their relationship to the political system has changed: their political interests frequently are independent of their marital status. It was this truth that the Reagan (and Bush in 1988) campaign captured. For all the Republican rhetoric about family values, their targeting strategy was based on a realistic picture of the working household, not a nostalgic fantasy.

GENDER IN 1992

There was no substantial gender gap in the 1992 race. Although women preferred Clinton over Bush (45 percent to 37 percent), so did men (41 percent to 38 percent). Nor were Perot voters significantly divided by gender. Perot received 17 percent of women's votes and 21 percent of men's. According to analysis of exit polling data by Joe Goode of the Greenberg Research Inc., Clinton's margin with women was due entirely to black women. White women split their votes evenly between Clinton and Bush (41 percent to each candidate). Clinton won working women by 10 points and younger women by 15 points, two targets of the campaign, and lost homemakers to Bush by 9 points. Democratic strategist Celinda Lake said, "Women were key to the stability and the base of Clinton's votes throughout the campaign as men ebbed and flowed with Perot's fortunes."

Indeed, the dramatic gender news in the Clinton campaign concerned men, especially white men who had been defecting to the GOP. According to Goode, "Clinton gains, the Bush collapse, and Perot's emergence allowed for very big swings to the Democrats among male groups who normally gave Republicans pluralities--namely, white men (Bush's margin dropped from 27 points to just 4 points), male college graduates (Bush's margin dropped from 27 points to just 2 points), and independent men (Bush dropped from an 18 point win to a 3 point loss)."

Other recent campaigns demonstrate the importance of women's votes and women's issues, combined with the absence of a simple or stable gender gap. In the 1990 California gubernatorial race, early polls showed Dianne Feinstein had a gender advantage from women who said they were going to vote for her over her opponent, Attorney General John Van de Kamp. By the close of the primary, however, this advantage had disappeared. Two-thirds of undecided women voters eventually voted for Van de Kamp. Men who decided late did just the opposite--they gave two-thirds of their votes to Feinstein. While Feinstein won the primary, the two candidates divided men's and women's votes almost equally.

In this race, feminist issues, which generally are more salient in a Democratic primary where activists wield more clout, were used against Feinstein. Van de Kamp was strongly prochoice and had championed other women's issues. Feinstein had vetoed several items considered important to the feminist community, including a reproductive rights commemorative day and comparable worth legislation. Former Congresswoman, now Senator, Barbara Boxer, who two years later would campaign with Feinstein as the "Thelma and Louise" of California politics, supported Van de Kamp in 1990, calling him "the best feminist in the race."

In the general election for governor against former Senator Pete Wilson, early opinion polls showed that Republican women were considering crossing party lines to vote for Democrat Feinstein. By the campaign's end, however, Wilson had won them back, narrowly defeating Feinstein. The predominant issues were the economy and crime, reflecting the vulnerability women feel in the face of high crime rates.

But the issue that "stopped Republican women cold," as pollster Mervin Field put, was quotas. Feinstein proposed that women and minorities be appointed to public positions in rough proportion to their numbers. Wilson quickly depicted Feinstein as advocating quotas. A Wilson ad asked, "Can we afford a governor who puts quotas over qualifications and promises before performances. . . .it is unfair, it's extreme, and it's wrong." Then in a hypocritical twist, the ad pointed out that "in nine years as mayor of San Francisco, the number of women appointed by Feinstein increased by only 1 percent."

Field, who does the California Poll and is a long-time observer of California politics, thinks this issue's strong effect on Republican women made it "easier [for them] to drift back into the fold." Thus an issue that could be considered feminist--parity between men and women in public appointments--was used to win women's votes for a male candidate. In November Feinstein lost to Wilson by 4 percent.

By contrast, a significant gender preference did develop for Ann Richards when she ran for governor of Texas in 1990. Six months before the election, State Treasurer Richards was more than 20 points behind Clayton Williams, a wealthy businessman who had never held a political office but who had spent over $4 million in the Republican primary. Just before Labor Day, when Richards had narrowed Williams's advantage to 10 points, polls showed she had roughly 44 percent of potential women voters, compared with 40 percent who favored Williams, not a significant preference. Male voters were giving a clear preference, 57 percent, to Williams.

Richards invested heavily in strategies that targeted women, notably by mobilizing the prochoice community to bring out Democrats and independents and to win the support of crossover Republican women. Richards also flipped on its head the traditional advice to women candidates not to call attention to themselves as women. Richards had a photo taken of herself with her hairdresser that emphasized her high-maintenance hairdo, a sweep of teased white hair, often referred to as a "power do." She then sent the picture on a postcard to every beautician in Texas, managing in one mailing to hit two vast constituencies--beauticians and their clients. In her book Storming the Statehouse, Celia Morris argues that the mobilization of a highly developed network of women's organizations was critical to Richard's victory.

O bservers also agree, however, that the action that turned the tide had nothing to do with feminist networks or issues. Clayton Williams publicly snubbed Ann Richards, refusing to shake her hand. Texas voters were appalled. The belief that "You just don't do that to a lady" signaled a substantial shift of support from Williams to Richards during the last week of the campaign. (Had Richards been a male opponent, the snub might have gone unnoticed--just a case of good ol' boys duking it out.) Williams lost even more support when it was revealed that he had tax problems.

According to Candace Windel Beaver, managing director of the Texas Poll, an unheard of number of voters changed their minds in the last few days of the election. Both men and women jumped to Richards, with women doing so in greater percentages. In exit polls, 10 percent to 12 percent of the men said they shifted to Richards on election day, while nearly 22 percent of women voters said they did. Richards beat Williams (52 percent to 48 percent), gaining 57 percent of women voters, clearly the margin of victory.

During the same 1990 election cycle, Republicans fielded a number of women candidates for the U.S. Senate, hoping they could use the gender gap to bump off veteran Democrats. For instance, Congresswoman Lynn Martin challenged Paul Simon in Illinois, and Claudine Schneider took on Claiborne Pell in Rhode Island. Though challengers, both women were considered viable candidates with some chance of winning. Neither did.

Pollster DiVall, who had worked for both Schneider and Martin, admitted, "The mistake we made was thinking the gender gap would favor women candidates." She concluded that "women are still more inclined to vote for Democrats, even male Democratic incumbents, over female Republican challengers."

In 1992 the Senate primary victories of Barbara Boxer and Dianne Feinstein, following Lynn Yeakel's upset victory in the Pennsylvania primary and Carol Moseley-Braun's in Illinois, were assumed to be part of a sweep by women voters angry over Anita Hill's treatment before the Senate Judiciary Committee.

Yet if not a single woman had voted in the California primary, Feinstein and Boxer would have won their nominations on men's votes alone. (And both were announced candidates before the Anita Hill episode.) Feinstein and Boxer won more men's votes than their opponents, a fact that does not diminish the extraordinary and historic achievement of either candidate. In a two-way race, men gave Feinstein 53 percent and her opponent, Gray Davis, 36 percent. In a three-way race, Boxer won a plurality of men's votes over her two opponents, 39 percent to 35 for Leo McCarthy and 23 for Mel Levine. Women supported Feinstein two to one over her opponent, while Boxer won twice as many women's votes as each of her opponents.

Indeed, the ever-elusive gender gap did show up in both Boxer's and Feinstein's general election victories. Both won a majority of women voters, and in Boxer's case, women gave her a clear margin of victory (61 percent to 46 percent). Both men and women gave Feinstein her winning majority (66 percent to 52 percent). The Thelma and Louise act by Feinstein and Boxer in the last days of the campaign stands to date as perhaps the most aggressive strategy ever successfully used by women in top-level campaigns to mobilize women voters--and without losing too many men's votes.

Both candidates did particularly well with working women (Feinstein, 67 percent; Boxer, 64 percent), with young women (Feinstein, 64 percent; Boxer, 60 percent), and even with older women voters, frequently a tough sell for women candidates (Feinstein, 66 percent; Boxer, 57 percent).

WOMEN'S POLITICAL ECONOMY

Bill Carrick, a strategist for Feinstein, remembers the exact moment he realized there was a new relationship between working women and the Democratic Party. "I remember sitting in a coordinated campaign meeting with the Clinton folks looking at tracking numbers and the usual get-out-the-vote stuff. . . . Our polling data was showing we had 70 percent working women [for Feinstein]. Clinton and Boxer's numbers were a little less but in the same neighborhood. And I said, 'Let me tell you something, guys. We ought to start thinking about working women as a Democratic Party base constituency group.' "

Feinstein and Boxer didn't do as well with homemakers (Feinstein, 45 percent; Boxer, 35 percent), following a national pattern in which homemakers tend to prefer Republican candidates. Among the newly elected female Democratic senators, Patty Murray of Washington--who ran on the slogan "a mom in tennis shoes"--won the most homemaker votes (50 percent).

Carol Moseley-Braun's election reveals a little noticed gender gap between black men and women, when compared with the Feinstein and Boxer contests. Black women gave Feinstein, Boxer, and Moseley-Braun more than 90 percent of their votes, consistent with historic black voting behavior that favors the Democrats. However, black men gave such a large percentage only to Moseley-Braun, who is black. An astonishing number of black men (22 percent) preferred Republican Bruce Herschensohn, one of the most conservative candidates in the country in 1992, over Barbara Boxer (78 percent). Similarly, 18 percent of black men preferred moderate Republican John Seymour over Feinstein, who received 82 percent, suggesting that black men may be reluctant to support white women candidates, even liberals with whom they otherwise might agree.

The slight edge that working women seem to be giving Democrats, however, was not strong enough to elect Democratic women who ran against Republican members of the Senate Judiciary Committee. Easily defeated were Gloria O'Dell, who challenged Robert Dole in Kansas, and Jean Lloyd-Jones from Iowa, who challenged Charles Grassley. Lynn Yeakel, who took on Arlen Specter in Pennsylvania, lost narrowly. In that race, Yeakel did win a majority of women's votes (54 percent), but lost men's votes (44 percent). Yeakel had majority support among working women and women of all age groups except older women, who preferred Specter.

In all these races, respondents to exit polls ranked the economy, jobs, and the U.S. deficit or the economy, jobs, and health care as the most important issues determining their vote, although women were more inclined to place health care over the deficit. Voters who said abortion--the only feminist issue listed on these exit polls--was among the top reasons for their vote ranged between 6 percent and 18 percent.

The data are quite clear that explicit feminism--in an ideological sense--does not underlie the development of candidate preferences by gender. Scholars have found little connection between feminist issues and women's weaker support for Reagan or Bush at the national level, even though a link between a gender gap and feminism seemed intuitively correct. Studying the same data analyzed by Smeal, most scholars insist that the composition of the gender gap in 1980 was based on economic and foreign policy issues, not on women's rights. In fact, political scientist Jane Mansbridge has pointed out that the 1980 data showed both men and women shifting their votes to Carter if women's rights were important to them. In general, numerous polls show that similarly situated men and women do not differ significantly in their support for women's rights.

A t the same time, feminist issues can become cutting-edge issues in campaigns. Abortion rights has a strong constituency among both men and women. Male and female candidates alike may win on this issue, as New York Congresswoman Nita Lowey and Virginia Governor Douglas Wilder illustrated.

Lowey won her first election to Congress in 1988 by upsetting a two-term incumbent through an intensive mobilization of Democratic, independent, and Republican pro-choice men and women. At the twentieth anniversary of the National Women's Political Caucus, Lowey described the race as "very targeted," a coalition of men and women where the issue of abortion rights "clearly made the difference," particularly among 18-to-34-year-olds. This lively race featured accusations of radical feminism reminiscent of past elections. "You're just a Bella [Abzug] without a hat," charged her opponent, Joseph DioGuardi. "And you're just a Dan Quayle without the good looks," quipped Lowey in return.

In 1989 reproductive rights and sex education became a central issue in Wilder's campaign for governor of Virginia, occurring as it did just four months after the Supreme Court's Webster decision, which began to restrict abortion rights. Large percentages of men and women cited abortion as important to their vote (men, 33 percent; women, 44 percent). Frank Greer, who prepared Wilder's media, said there was a "sea change" among young voters after the Webster decision. In the last week of the campaign, Democrat Wilder overcame a 15-point deficit and surged ahead by 8 points.

Campaigns in which abortion rights are front and center, however, are atypical. More common are campaigns in which candidates compete for ownership of the more broadly defined "women's issues." Iowa Attorney General Bonnie Campbell, elected in 1990, has made child support enforcement a cornerstone of her term. She has run "most wanted" ads for deadbeat dads and placed graphic advertisements in virtually every Iowa newspaper, including small towns, showing in detail what child support pays for: a lunch pail, a winter parka, school supplies.

Democrat Campbell, at this writing, is seriously exploring a run for governor. Her most likely opponent, incumbent Governor Terry Branstad, has tried to match her aggressiveness. Branstad is heralding a child support enforcement plan, which he claims will "have a bigger impact than just putting pictures on 'wanted' posters."

In California, Lieutenant Governor Leo McCarthy used his post to seek support among women voters by declaring breast cancer a "state emergency" when he ran against Barbara Boxer in the Senate primary. It is this way, and not via a simple gender gap, that women's issues and the female electorate are making serious headway in American political life. Nancy Kopp, a Maryland state legislator, noted recently with some irony that men are moving into legislative areas that have been strongly associated with women--education, for example--to improve their legitimacy with women voters. Conversely, female candidates who try to assert their superiority on women's issues often encounter stiff opposition from their male opponents, who argue that "all issues are women's issues," as Pete Wilson retorted to Feinstein in the 1990 campaign.

Ethel Klein, a political scientist and strategist who has studied the gender gap extensively, emphasizes that "it is an issue vote and can be triggered only by policy discussions that incorporate women's perspectives." Klein notes that women are "markedly more willing to express policy preferences" today than they were before the women's movement. Her comments point to a link between women's increased political involvement and the women's movement but do not confuse the two.

Klein highlights the multiplicity of interests women have developed, not what women have in common, a view consistent with the material change in women's economic roles, particularly their work force participation. By no means are all political interests economic. But these new roles have enabled women to develop an interest in government actions similar to men's. Yet Klein also emphasizes that "different groups of women base their value preferences on different sets of issues" and that all the traditional variables that affect actual votes, such as socioeconomic class, age, or even region of the country, come into play. Precisely how preferences will translate into votes, says Klein, "all depends on who is doing the organizing."

F or Democrats who wish to do the organizing, the experience of the last decade suggests that party strategists need to acquire an understanding of women's votes that goes far beyond simplistic notions of a gender gap. If the pitfall in the past was to assume that women's interests were identical to men's, it is equally misleading today to equate the gender gap with an emergent female voting bloc, let alone a monolithic one. Moreover, to argue that women voters should be taken more seriously is not to suggest that men's defection from the party be treated lightly. Democrats therefore need to think much more creatively about working families--in all their variety--in order to identify interests and to construct issues that might unite men and women voters.

Women voters may indeed have interests that are distinctly different from men's--a classic gender gap. But the translation of those interests into political issues--even "women's issues"--may or may not elicit gaps in male and female voting preferences. The potential of a majority coalition of progressive women and men will be realized only if strategists appreciate the complexity of women's voting preferences. If they target women voters based on stereotypes and oversimplifications, they will only reinforce the defection of white men without capturing offsetting women's votes.

The gender gap is elusive, but to view it as merely fickle misses the point. Like the electorate it reflects, the influence of gender on voting is substantial, subtle, and complex. And those who court the gender gap will be rewarded only if they treat it with appropriate respect.