The Dynamics of the News Framing Process:
From Reagan's Gender Gap to Clinton's Soccer Moms
Pippa Norris and Susan J. Carroll
Harvard University and Rutgers University


Abstract

The "gender gap"--a term referring to the differences between women and men in their political preferences and voting behavior--has been widely covered in the press since the early 1980s. Stories about women candidates and elected officials are also common. The central argument of this paper is that structural biases in practices of news reporting have frequently led to journalistic frames which provide oversimplified, and sometimes misleading, coverage of gender differences among voters and candidates. Working within the context of framing theory, we develop a benchmark of empirical evidence about trends in the number of women candidates and the electorate's political preferences compared with content analysis of how gender-related stories have been covered by the media in successive American elections. We conclude by considering the consequences of this framing process for journalism, for women's power at the ballot box, and for our understanding of American elections.

Presented at the 1997 Annual Meeting of the Southern Political Science Association, Norfolk Virginia, November 5-8.


The Dynamics of the News Framing Process:
From Reagan's Gender Gap to Clinton's Soccer Moms

The widely accepted principle for the American media, set out by the Hutchins Commission on Freedom in 1947 and reflected in professional codes and guidelines, is that journalists should meet high standards of informativeness, truth, accuracy, objectivity, pluralism, and balance (McQuail 1992, 117; Fink 1995). Many American journalists give high priority to these values (Weaver and Wilhoit 1996, 136). In evaluating how far the American press falls short of these standards, critics have often focused on the alleged tendency for journalists to slant the news towards the "left," due to liberal values (Lichter et al. 1986), or towards the "right," due to establishment pressures and over-reliance on official sources (Lee and Solomon 1992; Hertsgaard 1989; Hermann and Chomsky 1988; Parenti 1986; Naureckas and Jackson 1996). This perennial debate over press bias, generating more heat than light, remains unresolved since it is difficult to establish any agreed upon benchmark to measure ideological "bias" (Lacy, Fico and Simon 1991; Fico and Soffin 1995; Weaver 1972).

Yet, routine journalistic practices may contribute more to systematic biases in news reporting than any liberal or conservative leanings within the press. Decisions in the news-gathering process--determining what and how stories are covered--may prove critical here. Out of the myriad ways of describing events in the world, reporters, editors, and producers rely upon frames to convey dominant cultural meanings, to make sense of the facts, to focus the headline, and to structure the story-line. Although the specific details of a day's occurrences might be unique to that day--a plane crash, a presidential speech, a local murder--the way that journalists observe and report these occurrences has a lot to do with how similar events have been framed in the past (Bird and Dardenne 1997).

Although framing is widely discussed in political communication, social psychology, and public opinion research, the central theory remains somewhat underdeveloped. To clarify how the central concept is used in this study, following Gitlin (1980, 1994), we understand media frames to be persistent patterns of selection, emphasis, and exclusion which furnish an interpretation of events. News frames provide contextual cues, which give meaning and order to complex problems, actions, and events, and slot the new into familiar categories (Norris 1997a; Gamson 1991; Entman 1991, 1993). Frames represent consistent, predictable narrative stories which are embedded in the social construction of reality. Without such devices, journalists--and readers--would lack coherent news narratives which link disparate facts.

Like art, frames are often difficult to define, yet easy to recognize. Familiar examples include the "terrorism" frame to describe acts of political violence and attribute blame (Cohen and Wolfsfeld 1993), the "Cold War" frame to understand friends and enemies in foreign affairs (Norris 1997b), and the "natural disaster" frame commonly found in coverage of the developing world. Campaign news commonly emphasizes the poll-driven, horse-race frame (who's ahead, who's behind) as well as "game" over "substance" (Patterson 1994). The women's movement and feminism have also been subjected to media framing (Creedon 1993; Costain, Braunstein, and Berggren 1997).

Moreover, studies suggest the framing process has certain important consequences. Iyengar (1991) demonstrated that the selection of an episodic or a thematic frame in the news influences how viewers attribute responsibility for policy problems. For Zaller (1992) the process has a significant role in democracies, since frames determine our interpretation of public opinion. And according to Entman (1993), journalists may attempt to be "objective" in their reporting, and yet serve to reinforce rather than challenge the dominant framing in the news, a process which prevents their audience from making a balanced assessment of events.

Yet, although framing is commonly regarded as theoretically important to political communications, few studies have systematically compared an independent benchmark of "reality" (the actual events in the news) against the dominant frames journalists use to explain them. Moreover, we have little evidence to understand the dynamic process about how frames evolve over time. Do frames lag, mirror, or perhaps even lead, changes in the "real" world?

This paper aims to provide insights into these questions by examining one category of newspaper stories-- those covering gender-related developments in American election campaigns from 1980 to 1996--to see how accurately news frames reflect the realities of gender politics during these years. The first sections lay out the hypotheses and research design. We go on to establish baseline trends about women voters and candidates from 1952-96 based on aggregate and survey data. Subsequent sections compare the framing process in the media using three approaches: a Lexis-Nexis keyword search for framing terms in all major U.S. papers; content analysis of a sample of gender-related stories in The New York Times and The Washington Post; and finally a critical qualitative reading of selected stories. Like adjusting the focus on a microscope, this provides progressively narrower and more detailed perspectives on the content of news frames. Finally, the conclusion considers the implications of our findings.


Hypotheses

Based on previous studies of reporting routines and news narratives, we hypothesize that stories about gender and elections will be framed in certain consistent ways. First, other scholars have observed a tendency towards "pack journalism" (Sabato 1988) and preferences among journalists for narrative "pegs" around which to organize their stories (Bird and Dardenne 1997). As a result of shared news-room cultures, journalistic routines, and organizational structures, the selection of news stories by different journalists has been found to be fairly predictable:

Patterns of presidential election coverage are remarkably uniform, regardless of a newspaper's partisan orientation. Media personnel at highly regarded papers everywhere select the same kind of stories and emphasize the same kinds of facts, despite the wealth of diverse stories available to them (Graber 1997, 246).

As our initial hypothesis, we therefore expect to be able to identify certain predominant frames used by the press to structure different stories about gender politics in election campaigns. Some journalists might challenge the predominant frame, but most could be expected to reinforce it.

Second, we hypothesize that the framing process is dynamic so that the predominant frames change over time. For example, we expect to find that from 1982 to 1992 the gender gap was conventionally framed as a political problem for the Republicans, with stories focusing on women becoming more Democratic. In contrast, when journalists had to explain the election of Republican majorities in both the House and Senate in 1994, we anticipate that the "Year of the Angry White Male" emerged as a dominant frame in news coverage, emphasizing change among men. Similarly, we expect to find a return to a focus on women when explaining Clinton's victory in the 1996 election with the "Soccer Mom" as a predominant frame used for writing about gender differences. We also expect that the major national 'papers of record' (The New York Times and The Washington Post) may be particularly influential in setting the predominant frame and that other regional papers will tend to follow their lead.

Lastly, at least one previous study has found much hyperbole and exaggeration in media coverage of the gender gap (Ladd 1997). Given the perceived need for news to focus on drama and novelty and to emphasize change rather than continuity (Gans 1979; Graber 1997, 118-9), we also hypothesize that the frames used for writing about gender politics will commonly overstate and magnify changes across election cycles in the actual number of women candidates and the behavior of voters. We expect to find that the patterns evident in empirical data will show much greater continuity across elections than media coverage would suggest.


Research Design, Data and Methods

To explore these hypotheses, we used four strategies. First, to establish a suitable independent baseline for comparison with media coverage, we examined empirical evidence about changes among women voters and women candidates over time. We analyzed trends in the gender gap from 1952 through 1996, examining exit poll data (rather than data from the National Election Studies or some other source) because these are the data journalists had available to them on election night and in the period immediately following the election. We also examined data on the numbers of women candidates and officeholders provided by the Center for the American Woman and Politics (CAWP). CAWP has tracked the number of women candidates in every recent election and has made the information on candidates available to the press.

Second, to examine trends in the news media coverage, we conducted a keyword search in the Lexis-Nexis database of major papers (MajPap), counting the number of stories (N=2228) which included the key terms gender gap and election, Year of the Woman and election, angry white males/men and election, and soccer mom and election. We focused on the period during and immediately after the most intense portion of the general election campaign (from September 1 to November 30) in the 1980, 1982, 1984, 1986, 1988, 1990, 1992, 1994, and 1996 elections. We did not attempt to distinguish between lengthy treatments of these themes and passing references to these terms. Neither did we examine the contents of these stories. We use the results of this Lexis-Nexis search to establish trends over time in a wide range of newspapers.

Third, to look at the framing process in more depth, we conducted a content analysis of stories in two major newspapers, The New York Times and The Washington Post, which are widely regarded as "papers of record" and which influence coverage in the regional press. We used keyword searches in the Lexis-Nexis data base, looking for all stories published in September, October, and November of 1982, 1984, 1986, 1988, 1990, 1992, 1994, and 1996, which referred to elections and women voters, women candidates, or the gender gap. We identified a total of 321 stories across all elections. These were then coded by the date, source, gender of reporter, length, headline priority/running order, the primary and secondary subject themes, sources interviewed, and the major and minor frame.

Lastly, we also gave a close textual reading to selected lengthier stories to understand the framing process from a more qualitative perspective. As Entman (1993) notes, the major task of determining textual meaning cannot be accomplished by simply adding up messages without understanding the clusters of salient messages which constitute the broader frames. For example, the phrase "angry white man" was used only once or twice in some articles, but often the whole story was written to describe and support this interpretation of the outcome. We looked at the first election stories we could identify from our searches which referred to terms like the 'gender gap" or "soccer mom," and then examined subsequent extended treatments of these stories in more typical coverage.



Baseline Evidence on Women as Voters and Candidates 1952-96

To establish an independent baseline to compare against media coverage, we examined trends in electoral behavior and women's representation in elected office using available empirical data. We summarize those data here.

In order to understand the power of the modern gender gap frame in political science as well as the media, we need to remember the conventional wisdom which predated it. The first empirical study of gender differences in voting behavior was Maurice Duverger's classic work, The Political Role of Women (1955). Duverger found women voters to be marginally more right-wing than men in Norway, France, and Germany, and similar patterns were evident in the United States. As Campbell et al.(1960) noted, in the 1952 and 1956 elections, women were slightly more likely than men to vote for the Republican presidential candidate, in the region of 3 to 5 percentage points, although this "gender gap" in voting could be attributed to gender differences in age, regional turnout, and education. Gallup polls also showed stronger female support for Republican presidential candidates during the 1950s (see Figure 1).

[Figure 1 about here]

During the 1960s and 1970s, the voting differences between American women and men as reported in surveys were usually statistically insignificant (see Figure 1). The modern gender gap became evident in 1980; since then American women have consistently given stronger support than men to Democratic presidential candidates, while men lent stronger support than women to the Republicans, generating a flourishing literature seeking to explain this phenomenon (see, for example, Mueller 1991; Shapiro and Majajan 1986; Wirls 1986; Bandana and Lake 1994; Clark and Clark 1996; Cook and Wilcox 1991; Conover 1988).

The gender gap in presidential voting, defined as the difference in the proportion of women and men supporting the winning candidate, has ranged from a low of 4 percentage points in the 1992 elections to a high of 11 percentage points in 1996. Throughout the 1980s, the gender gap in presidential voting, as reflected in exit polls conducted by major media, was fairly consistent in magnitude, ranging from 6 to 9 percentage points (CAWP 1997a). Despite the fluctuation in the size of the gender gap between the 1992 and 1996 elections, the gender gap in presidential voting has been a consistent and persistent feature of the electoral landscape since 1980.

[Table 1 about here]

The gender gap in voting in non-presidential races also has shown consistent patterns with a few notable fluctuations over time. As Table 1 indicates, the gender gap in congressional races was 3 to 4 percentage points in every election between 1982 and 1992. In 1994, the gender gap in congressional voting increased to 11 percentage points, due largely to the fact that men voted more Republican and less Democratic than they had in 1990 or 1992 while women voted much the same as they had in both previous elections.

In statewide races (for U.S. Senator and governor), there have been gender gaps in a majority of the races where exit polls have been conducted throughout the 1980s and 1990s. For example, the Center for the American Woman and Politics (CAWP) reports that there were gender gaps of at least four percentage points in 78 percent of the races where Voter News Service conducted exit polls in 1996, in 81 percent of the races in 1994, in 67 percent of the races in 1992, and in 61 percent of the races in 1990 (CAWP 1997a). In addition, in every election since 1982, it has been possible to identify several specific statewide races where not only was there a gender gap, but also women's votes provided the margin of victory (CAWP 1997a). There is no trend evident in the number of such races over time; for example, women's votes provided the margin of victory in 3 races in 1992, 10 races in 1986, 4 races in 1992, and 8 races in 1996 (CAWP 1997a).

[Table 2 about here]

Just as the data on the gender gap show considerable consistency over time with a few notable fluctuations, so too do the data on the number of women running for and elected to public office. Table 2 presents data on the number of women running for the offices of U.S. Senator, U.S. Representative, governor, lieutenant governor, and state legislator in elections from 1982 to 1996. Few clear patterns over time are evident in these data. The numbers of women running for the U.S. House and Senate fluctuated somewhat, while remaining consistently low, throughout the 1980s. The number of women candidates for Congress increased considerably in 1992, and in 1994 and 1996 the numbers of women running for the U.S. House continued to increase. Thus, in both 1994 and 1996, as well as in 1992, record numbers of women ran for the U.S. House.

The data on gubernatorial candidates indicate that a record number of women ran for this office in 1986; that record was broken by a new record high number of candidates in 1994. 1986, 1990, and 1994 (especially the latter) were years when record numbers of women ran for lieutenant governor. While record numbers of women were running for Congress in 1992, the numbers of women candidates seeking the office of governor and lieutenant governor in 1992 were small relative to other years. At the state legislative level, the number of women candidates increased with each successive election through 1992; in 1994 and 1996 the number of women state legislative candidates declined somewhat.

[Table 3 about here]

Data showing trends over time in the number of women serving as public officials are presented in Table 3. Numbers of women state legislators have increased in each successive election. This same pattern is not evident for the U.S. Senate and U.S. House. In both Congress and the state legislatures, the largest increase in women members came in the aftermath of the 1992 elections. However, all three of the women who ran for governor in 1992 lost, adding no new numbers to the three women governors serving at that time. Only 15 women have ever served as governors, and the largest number ever to serve simultaneously was four in 1994 and for part of 1991 (CAWP 1997b).

Keyword Search of Framing Terms in Major Papers

In recent years, certain gender-related phrases --the gender gap, Year of the Woman, angry white male, and soccer mom-- have come to be associated with election coverage of women in politics. As a first step in identifying possible news frames, we examined the frequency with which these key phrases appeared in articles in major newspapers, focusing on the elections in the years between 1980 and 1996. Table 4 presents the results.

[Table 4 about here]

Because the number of papers added to the Lexis-Nexis major papers data base has increased with each successive election, the data in this table cannot be used to examine possible changes over time in the number of gender-related stories. However, for each individual year, the data do provide a snapshot of where the media were focusing their attention.

Articles mentioning the term "gender gap" were common across all post-1982 elections. None of the articles published in 1982 mentioned any of the other gender-related terms.

In 1984, when Geraldine Ferraro became the vice-presidential nominee on the Democratic ticket, a small number of articles made reference to the "Year of the Woman." This phrase continued to appear occasionally in newspaper coverage in subsequent elections. However, the phrase became particularly prevalent in coverage of the 1992 elections when record numbers of women ran for and were elected to Congress. In fact, "Year of the Woman" was mentioned four times more often in election coverage that year than was the phrase "gender gap." In coverage of the 1994 elections, references to the "Year of the Woman" continued to appear frequently as reporters undoubtedly drew contrasts between 1992 as the "Year of the Woman" and 1994 as the "Year of the Woman--Not." However, 1992 and 1994 were the only two elections where references to the "Year of the Woman" were more frequent than were references to the "gender gap" (Table 4).

The other key phrases--"angry white male" and "soccer mom"--are of relatively recent vintage. "Angry white male," first appears in the aftermath of the 1994 elections when reporters were attempting to explain the outcome of an election where Republicans gained control of both houses of Congress for the first time in four decades. Some references to "angry white males" are also found in coverage of the 1996 elections, again likely in reference to the previous election. However, much more common in 1996 are references to the media's newest discovery, the "soccer mom." Although references to the "soccer mom" are less common in 1996 than references to the "gender gap," both occur with considerable frequency (Table 4).



Content Analysis of Gender-Related Frames in News Coverage

The above examination of the use of key phrases in media coverage could be considered a relatively weak test because we have no way of establishing a suitable 'universe' of stories to say how far these frames are representative of 'typical' coverage. We find what we are looking for. Moreover some frames -- such as women as outsiders --may not be so clearly embodied in single key phrases. Our content analysis of coverage in The New York Times and The Washington Post provides a stronger test for gender-related frames, providing more detail but limiting the focus in terms of the number of sources and news stories. Rather than simply tabulating the number of stories which contain an explicit reference to frames which we anticipated might exist, in this part of the analysis we look in greater depth at a sample of stories which could be regarded as typical election coverage of gender politics to see which gender-related frames, if any, emerge.

First, however, to put our analysis of possible news frames in perspective, we attempted to assess the importance of news about gender politics by gauging the number, length, location, and headline treatment of gender-related election stories. Overall, the coverage in these two papers suggests that in most recent elections gender-related phenomena have been viewed as moderately important by the news media.

The number of stories we identified in these sources varied somewhat over time with the largest numbers of stories, as one might expect, appearing in three presidential election years--65 in 1996, 53 in 1992, and 51 in 1984. Generally, fewer gender-related articles were published in non-presidential election years with only 25 articles appearing in 1994 and 26 in 1990. Nevertheless, fewer articles, 28, were published during the 1988 presidential election than in the off-year elections of 1982 and 1986, when 39 and 34 articles were published, respectively, suggesting that gender was not viewed by the media as a particularly interesting or important factor in the 1988 elections.

We coded the approximate proportion of each article which referred to gender politics and found that about one-third of all stories contained only a passing reference to gender, and another third devoted less than half the article to gender-related content. However, the remaining third provided extensive treatment of women as voters and/or candidates. This pattern remained fairly stable over time.

The location of stories in the newspaper serves as a third indicator of the priority given by the media to gender politics. We found that a fifth of all stories were front-page news, and almost half were on other news pages; the remainder were found on editorial and op-ed pages or in other sections. Notably, less than a dozen stories were in style or women's sections.

Lastly, we examined whether the headlines for these stories highlighted gender politics. Of the 206 articles which made more than a passing reference to gender, the headlines of half (50.0%) made no reference to gender. In the articles with headlines which focused on gender, women candidates and women voters received almost equal attention, with 21.4 percent of the headlines focusing on women voters or the gender gap and 18.4 percent focusing on women candidates.

In testing for the presence of gender-related news frames in coverage in The New York Times and The Washington Post, for each story we determined whether there was a dominant gender-related frame around which the story was structured. From this content analysis, we identified eight possible gender-related frames, or themes, that were evident in these election-related stories (Table 5). Of the stories that made more than a passing reference to gender, the discussion of gender in 60.5 percent (N=200) was structured around one of these themes. Three of the themes--the gender gap, the angry while male, and the soccer mom--focused on voters. The other five--the Year of the Woman, record numbers of women candidates, the first woman, a bad year for women, and women as outsiders/agents of change--focused on women candidates.

Because of the prominence of the gender gap theme across elections, we conclude that it clearly deserves to be considered as a media news frame (Table 4). This frame is apparent in every election; it dominates coverage in 1988 and is the most common gender-related theme in 1982, 1984, and 1996. The angry white male and the soccer mom are perhaps most appropriately considered as variations of the gender gap frame that were employed by journalists to adapt the gender gap frame to fit the peculiarities of a particular election--the angry white male in 1994 and the soccer mom in 1996.

The candidate-related themes present a more complicated picture. None of the candidate-related themes is predominant across years in the way that the gender gap frame is. Yet, several seem closely interrelated. For example, in some years, such as 1986 and 1990, a substantial proportion of articles were structured around the theme of "record numbers of women running for office." In at least one other year, 1984, the predominant candidate-related theme seemed also to focus on a record but of a different kind--"the first woman" running for a given office.

If the record number of women running is great enough, as it was in 1992, the theme may become the "Year of the Woman." Just as the "Year of the Woman" and the "record numbers of women running for office" themes seem highly related, at least conceptually, so too are the "Year of the Woman" and "women as outsiders/agents of change" themes. Moreover, these latter two themes seem to be prominent in the same years--1990, 1992, and to a lesser extent, 1994. The themes of "Year of the Woman" and "women as outsiders/agents of change" are not important in other years.

The final candidate-related theme, one we did not initially anticipate, shows up in both the early 1980s and the mid-1990s; this is the theme of "a bad year for women candidates." This frame seems almost to be the flip-side of "record numbers of women running for office." When it is not a record year for women candidates or when women candidates do not fare as well as expected, the frame of "a bad year for women candidates" seems to be the one that journalists adopt for writing about women candidates. From these data on candidate-related themes in media coverage, it appears that journalists have used a variety of interrelated frames over the years to structure their stories on women candidates (Table 5).

Table 5 also highlights the shifts in gender-related frames across elections. In 1982 the gender gap clearly was the frame which dominated media coverage. In 1984 three frames were employed almost equally--the gender gap, the first woman candidate, and a bad year for women candidates. Two frames dominated the 1986 election coverage--the gender gap and a record number of women candidates. The only gender-related frame apparent in coverage of the 1988 elections was the gender gap. In 1990 candidate-related frames were much more evident than voter-related frames, but no single frame dominated coverage. In 1992, the interrelated frames of the Year of the Woman and women as outsiders/agents of change were most evident. In 1994, coverage was focused on the angry white male and a bad year for women candidates. In 1996 the gender gap frame was again dominant with some focus on the soccer mom. In short, while common frames are evident in media coverage over time, considerable shifting among these frames is evident from one election cycle to another.



A Critical Analysis of Gender Frames

Yet content analysis provides only limited insights into the framing process, mainly because the overall frame may only emerge from the narrative construction of the story, rather than from its particular components. Like the role of theories during periods of normal science, frames may provide the underlying structure and rationale for a news story, and yet be relatively implicit rather than overt. In this final section we therefore take a more in-depth, qualitative look at the frames that have emerged from our content analysis of news coverage. We also assess the validity of these news frames in the light of the empirical data on women voters and candidates presented earlier, as well as some additional data relevant to specific frames. We first examine the "gender gap" news frame and its "angry white male" and "soccer mom" variations. We then turn to candidate-related news frames.

The Gender Gap

According to Kathy Bonk, the National Organization for Women (NOW) played a major role in attracting media attention to the "gender gap." NOW originally developed the phrase as a dramatic and simple way to publicize the women's vote (Bonk 1988). The story first broke with Judy Mann in an article entitled, "Women are Emerging as Political Force," in The Washington Post (October 16, 1981). The theme was picked up again in a lengthy, front-page article in The New York Times in June 1982 by Adam Clymer, who throughout focused on the change among women rather than men:

HEADLINE: WOMEN'S POLITICAL HABITS SHOW SHARP CHANGE

The political habits of women appear to be undergoing deep changes that worry the Republicans and raise long-range hopes of the Democrats. A variety of newly available statistics show that women, who in the past have voted at a lower rate than men, are now voting at roughly the same level. These statistics also show that women, whose political attitudes used to be barely distinguishable from those of men, are beginning to take positions on issues that sharply differ from those taken by men.

By October 1982, the gender gap had become established as one of the major stories in the campaign and was an issue discussed by candidates, campaign managers, and pollsters. The gender gap frame could be used as a peg to make sense of diverse races, campaign issues, and candidate strategies. Throughout the 1982 campaign, the focus remained on how women's voting behavior had shifted, speculation about the possible reasons for this shift, and reaction by party spokespersons. The story was increasingly picked up in analysis of the mid-term results in 1982, where the gender gap was treated as one of the most significant developments in the election, indeed the basis for a possible realignment. In Adam Clymer's words:

But another kind of realignment, in which more women are tying themselves to the Democratic Party while men vote more heavily for Republicans, could have a more profound impact on American politics than a shift in partisan dominance. This so-called gender gap....may influence American life in the 1980s as much as the civil rights revolution did in the 1960s (The New York Times, October 31 1982).

Coverage of the gender gap in 1983 and 1984 helped to keep women's political concerns on the political agenda and strongly influenced the choice of Geraldine Ferraro as Vice-presidential candidate on the Democratic ticket (Borquez, Goldenberg, and Kahn 1988). In the 1980, 1982, and 1984 elections, the gender gap could be dismissed as a short-term reaction to President Reagan's leadership and the mobilization of the New Right. But subsequent trends, with American women consistently leaning more heavily towards the Democrats while men favored the Republicans, convinced even the most skeptical journalist that this was not merely a temporary blip. Once politicians awoke to the electoral salience of gender, they tried different strategies to mobilize women's support at the ballot box, including recruiting more women to run for public office, appealing to women on the basis of policy issues, and carefully manipulating the party image. Once established, the gender gap became commonplace in routine analysis of the results in subsequent elections, with the number of campaign stories using this phrase in major papers dramatically peaking at 547 (about 16 stories a day) in 1996 (Table 4).

But why was the 1982 election framed in terms of the gender gap? This was not the first election with a significant difference in the voting preferences of American women and men. As noted earlier, during the 1950s surveys indicated there was a consistent (although slightly smaller) gender gap, with women leaning towards the Republicans. However, this received little coverage in the news.

As an alternative frame, the story about gender differences in voting during the 1980s could have been framed in terms of a strong Republican shift among men, most evident in changes in party identification from the 1970s into the 1980s. Yet, this explanation was not very common. At least until 1994, stronger Democratic support among women, although one interpretation among several alternatives, rapidly became the conventional way to cover this story.

There is also the question of why the gender gap played a more prominent role in media coverage in some years than in others. The gender gap was the predominant frame for gender-related news coverage in 1982, 1984, 1988, and 1996 (Table 5). Novelty may account for its popularity in 1982. And in 1996 the gender gap in voting in the presidential race was bigger than it had ever been before. However, since the gender gap was fairly consistent in size throughout the 1980s, as noted earlier, it is unclear why the gender gap frame dominated news coverage more in 1994 and 1988 than in 1986 and 1990. The shifts over these election cycles from more attention to the gender gap one election year to less the next gives the impression in news coverage of much more dramatic swings in electoral behavior than actually took place.

The Angry White Male as a Variant on the Gender Gap Frame

The conventional way of understanding the gender gap worked when reporters needed to account for Clinton's victory in 1992. But in contrast, the 1994 mid-term congressional elections were widely framed after-the-fact as the "Year of the Angry White Male," reversing the conventional wisdom about whether changes in women's or men's behavior were more significant in understanding the effects of gender differences on politics. In a Nexis-Lexis search of the major newspapers since 1990, the first reference to this phrase came in an article by Richard Morin and Barbara Vobejda on November 10, 1994, in The Washington Post:

Two years ago, it was the Year of the Woman. This time around, the election may become known as the Year of the Man, or the Year of the Angry Man....

In the next two weeks the phrase "angry white male voter" was rapidly picked by all over the country in the post-mortem of the election results, in The Atlanta Constitution (November 12), The Houston Post (November 14), The Los Angeles Times (November 15), The Boston Herald (November 15), USA Today (November 18), and so on. Interviews rapidly located people who had started to describe themselves as "angry white men." The Washington Post interpreted developments, with considerable dramatic license, as follows:

HEADLINE: WHITE GUYS: THEIR ISSUES ARE GUNS AND BUTTER. THEY'VE HAD IT WITH DEMOCRATS. HERE'S AN EARFUL.

Day dawns gray and wet on the Pulaski Transportation Highway running east out of Baltimore toward Aberdeen. ..Drop into the cozy warmth of the Double T Diner at 6.45 a.m., ..there sits a guy named Michael Dolan, 37, a painting contractor. At the moment, he's hunched intently over a newspaper article about Newt Gingrich. "I feel great that the Republicans kicked ass," Dolan says in a Jersey twang that bounces off the walls. "I'm euphoric about it. I can't stomach what the Democrats stand for," Dolan says. "I don't like Clinton. To me, they took work away. I worked my ass off to achieve middle class, and then it all came crashing down."

Some journalists challenged the framing, but most went along. By the end of November, the following example in The Plain Dealer was typical coverage:

HEADLINE: WHITE MALE VOTERS RECEPTIVE TO GOP MESSAGE

Democratic and Republican leaders, meeting separately to rehash the election, agreed on one conclusion: Angry white men fueled the GOP landslide and could consign Democrats to long-term minority status.

The Democrats, meeting at Walt Disney World in Florida to figure out what went wrong, talked about winning back independents, increasing their margin among union members and finding some way to become competitive in the South. They conceded, with some discomfort, that all of these categories were a manifestation of a problem that has been building for a generation:

White men increasingly don't vote Democratic (November 27, 1994).

The "gender gap" story was turned on its head. The Nexis search revealed no references to "angry white men" voters in the major American papers prior to November 10, 1994. However, by the end of the year, 32 stories had appeared, with the number escalating to 208 stories in spring/summer 1995. Thus, in successive elections first the Democratic edge among women, then the Republican edge among men (angry or not), became the conventional interpretation of gender differences in voting behavior--the dominant frame which could be adopted in a flexible way to cover stories about different regions, candidates, or party strategies.

Scholars with access to the National Election Studies have searched hard for evidence about the angry white male in the 1994 elections, but they have come up empty-handed (Tuchfarber 1995; Reeher and Cammarano 1996). Instead, they have found that the familiar gender gap persisted. Although the gender gap in voting at the congressional level grew in size in 1994 and men did seem to vote more Republican in congressional races in 1994 than they had in previous elections, the thesis that angry men were responsible for either the increase in the gender gap or the election results in 1994 seems to have been a myth.

The Soccer Mom

By 1996, given Clinton's lead in the polls, the "Angry White Male" seemed past its sell-by date. Accordingly, we see the sudden rise to fame of the "Soccer Mom," a revival of the original gender gap story but with a fresh twist. The first reference to the soccer mom was by E. J. Dionne in an article in The Washington Post on July 21, 1996. In this article Dionne quoted Alex Castellanos, a senior media adviser to Bob Dole, suggesting that Bill Clinton, following the advice of his pollster, Dick Morris, was targeting a voter whom Castellanos called the "soccer mom"--"the overburdened middle income working mother who ferries her kids from soccer practice to scouts to school." The phenomenon of the soccer mom, seen as a distinctive demographic voting group, thus appears to have been the creation of consultants involved in the presidential campaigns (Carroll 1997).

The soccer mom story gradually took off, before becoming ubiquitous in coverage in the few weeks just before and after the election (see Carroll 1997). The soccer mom was portrayed as the key swing voter in the 1996 campaign. If Bob Dole were to win the election, according to media reports, he had to win back the votes of soccer moms, who traditionally had voted Republican but who were supporting Clinton in the 1996 election. The Clinton campaign also targeted these women voters according to the press. The New York Times reported that "In Arkansas, [Clinton campaign] workers were sent out to seek those who have become the embodiment of this year's swing voters, the soccer moms, the mission being to hand out campaign fliers to mothers at 51 soccer fields" (October 28, 1996).

The phenomenon of the soccer mom even caught the attention of William Safire, who noted in his New York Times Magazine column:

As the presidential campaign of 1996 enters its final week, its most powerful catch phrase has just been determined.... The phrase is to this campaign what gender gap was to the Bush-Clinton coverage in 1992 and the catch phrases angry white male and Contract With America were to the 1994 Congressional contest.

The envelope, please. And the winner is: soccer mom! (October 27, 1996).

Some reporters criticized the "soccer mom" label as stereotyped, over-simplified, and cliched; others noted that the group "soccer moms," however defined (and "soccer moms" were defined in somewhat different ways in different news stories), represented only a small proportion of women voters or the electorate as a whole. Notably long critiques include those by Amy Schwartz (The Washington Post, November 20, 1996) and Ellen Goodman (The Washington Post, November 16, 1996).

Nevertheless, much of the coverage portrayed the soccer mom in a serious manner. Typical was a front-page article in the Sunday New York Times on October 6, 1996, with the headline: "POLITICS--Political Battle of the Sexes Is Tougher Than Ever; Suburbs' Soccer Moms, Fleeing the G.O.P., Are Much Sought." The article explained:

In fact, the white suburban women... who are fast becoming the most sought-after voters of the campaign season are already known in campaign circles by a... name that is only now gaining currency: soccer moms.

....By breaking with their traditional Republican leanings and spurning Bob Dole in striking numbers, according to early polls, they have become a central focus of campaigners and pollsters who consider them part of the swing vote, without which Mr. Dole cannot win the Presidency.

[Table 6 about here]

Ultimately, however, the soccer mom did not play a particularly important role in Clinton's victory. The pattern of gender voting by subgroups, given in Table 6, shows that the size of the gender gap in 1996 varied in terms of income, work status, marital status, and children. For example, single working women voted for Clinton over Dole by a massive margin of 65 to 26 percent. Poorer women (with an income under $15,000 per year) voted for Clinton by a margin of about 2 to 1, as did unmarried women. In sharp contrast, the category of "soccer moms" (defined as suburban, married, professional women) proved to be an insignificant predictor of voting choice. Like the "angry white male" before her, on further analysis, systematic evidence establishing the importance of the soccer mom proved elusive (Kaufmann and Petrocik 1997). Because of the excessive focus on the soccer mom in 1996, the contributions which single working women, poor women, and other groups of women made to Clinton's lead in the polls and his re-election went largely unnoticed by the press.

First Woman

Just as frames are used by reporters in writing about women voters, gendered frames also are common in coverage of women as candidates. The "first woman" frame is common whether covering breakthroughs in the glass ceiling for women as leaders in politics, science, or business and management. This frame emphasizes the rise of particular women into positions where there have been none before, such as the first female President or Prime Minister in a country (Norris 1997a).

Although this frame occasionally was applied in other circumstances such as the 1986 Nebraska gubernatorial race where for the first time in history two women faced each other in a race for a state's top executive position (New York Times, September 16, 1986), the first woman frame was particularly prevalent in 1984, following Geraldine Ferraro's nomination as the Democratic vice-presidential candidate. This frame was, for example, very much evident in a piece in the Sunday New York Times Magazine written by Betty Friedan, who in the opening paragraph explained:

In July, as a delegate to the Democratic convention in San Francisco, I joyfully helped nominate Geraldine A. Ferraro for Vice President. It was the culmination of the modern movement for equality of women, and I reveled in the sense of limitless possibility it seemed to offer--not only to American daughters, but to women the world over (October 28, 1994).

Similarly, the "first woman" frame is apparent in Judy Mann's column on the day after the 1984 election entitled, "The Gender Barrier Is Gone:"

This election has been a watershed in American politics. In less than a year's time, the notion of having a woman on a national ticket has gone from being the wild-eyed fantasy of feminists... to a historic reality. While polls show Ferraro's nomination for vice president has been at best a mixed blessing to her ticket... she nevertheless has opened the way permanently for women to run for the highest office in the land (The Washington Post, November 7, 1984)
 
 

Record Numbers of Women and a Bad Year for Women

During every election season, organizations such as the Center for the American Woman and Politics and the National Women's Political Caucus release data on the numbers of women running for state legislative, statewide, and congressional offices. These organizations try to draw the attention of journalists to new "breakthroughs" for women candidates as well as to the existence of any record high numbers of candidates at any level of office. Such breakthroughs and record highs have been apparent in virtually every election cycle. Sometimes journalists have been interested; sometimes they have not.

The "record numbers of women" frame is one that reporters have used across elections, but it was particularly prominent in 1986 and 1990 (Table 5). A September 14, 1986, column by David Broder in The Washington Post provides an example of the use of this frame:

From Arliss Sturgulewski in Alaska to Julie Belaga in Connecticut, women candidates (like those two gubernatorial nominees) are on the move in 1986, seeking higher offices in greater numbers than ever before. An accident? A coincidence? Hardly. The smart operators have seen it coming and can explain convincingly why it is happening now....

Most of us were dazzled when returns from last Tuesday's primaries in nine states showed 32 women had been nominated for statewide or federal office. It capped a year in which six women have been nominated for the Senate, nine for governor, 10 for lieutenant governor and 53 for the House of Representatives.

Another example from a different election comes from an editorial in The Washington Post on November 16, 1990:

There were 29 women in the House and two in the Senate during the 101st Congress, and those numbers will remain the same for the 102nd. It's a different story in the states. According to the Women's Campaign Fund, a political action committee supporting pro-choice female candidates, this has been a record year for women running for state-wide office. More than half of the 85 candidates were elected, and when they take office next year, a total of 59 women will be in those high posts.

None of this will come as a surprise in this area. Women were elected city-wide for mayor of the District of Columbia, for two at-large seats on the city council and for delegate to Congress. And adjacent jurisdictions already have women in state-wide office: Mary Sue Terry, the attorney general of Virginia, and Barbara Mikulski, a U.S. senator from Maryland. But around the county and across party lines, voters chose women for important leadership positions.

An opposing frame to the "record numbers of women," and one which also has been employed across election cycles, is the "bad year for women" frame. This frame was most frequently employed in 1994 and 1984 (Table 5). A September 6, 1994, front-page article in The Washington Post, for example, opened with the question, "If 1992 was the Year of the Woman, is 1994 the Year of the Woman in Trouble?" The article continued:

Two years after voters sent a record number of women to Congress, the political landscape has changed and many of the incumbent female lawmakers are in tough reelection fights.

Thirteen of the 48 female House members... are in danger of losing their seats.

The "bad year for women" frame was also applied in 1984 when the addition of Geraldine Ferraro to the ticket failed to deliver a Democratic victory. The following excerpt from an article the day after the 1984 election provides an entertaining illustration:

Three television sets were on, each tuned to a different network... Kathy Wilson, chair of the National Women's Political Caucus, and her staffers gathered around... and waited hopefully for scarce good news, turning down the volume on one television to tune in the other two.

Suddenly a woman's face appeared on the silent set, the word WINNER triumphantly bannered above her head.

"Winner!" cried Rosalie Whalen, executive director of the National Women's Education Fund. "Winner! Who is that? Is that one of ours?" She quickly turned up the volume. The woman, it was announced, was the winner of the lottery.

The half-dozen women in the room roared with laughter.

Of course, you couldn't call it a night of triumph for women activists watching the first woman vice-presidential candidate (and her running mate) go down to swift defeat (November 7, 1984).

The distortion that comes with either frame--record numbers of women or a bad year for women--is that every year is a record year for women candidates somewhere at some level of office and every year is a bad year for women somewhere at some level. For example, while 1994 was frequently portrayed in the press as a bad year for women in politics (Table 5), record numbers of women ran for and were elected to statewide offices generally, and the office of lieutenant governor more specifically (Table 2), in that year. In other words, 1994 really was a record year for women running at statewide levels, and yet the press chose not to emphasize this story. Similarly, 1992, which was commonly framed as the Year of the Woman and a good year for women candidates, was a particularly bad year for women who ran for governor, where only three women won their party nominations and all three lost general elections.

The Year of the Woman and Women as Outsiders/Agents of Change

The 'Year of the Woman' frame was occasionally used in previous elections but the term first came into popular currency in 1990, before taking off like wildfire in 1992 when record numbers of women ran for Congress. We identified 584 stories using this phrase in major papers during this election alone (Table 4). The following article in The Washington Post reflects a typical example of the use of this frame:

HEADLINE: OUTRAGE FUELED WINS BY WOMEN

Tuesday's elections produced record numbers of new congressional women, tripling elected female membership in the Senate and increasing it by two-thirds in the House in one enormous surge.

Despite the triumph and accompanying glitter, however, the "Year of the Woman" may be a rare event, unlikely to repeat itself any time in the near future (November 5, 1992).

Another closely related frame was "women as outsiders/agents of change," which has also been found internationally when covering the entry of new women leaders as heads of state (Norris 1997).

Sometimes the two themes, the Year of the Woman and women as outsiders, were interwoven into the same article. Typical was Guy Gugliotta's post-1992 election story in The Washington Post:

HEADLINE: YEAR OF THE WOMAN BECOMES REALITY AS RECORD NUMBER WIN SEATS

They took their inspiration from Anita F. Hill, saw their opportunity in an electorate hungry for change and cast themselves as outsiders in a year when outsiders could be fashionable.

And it worked. Yesterday, a record number of women won seats in Congress, making the "Year of the Woman" a reality. Four female newcomers won or were projected to win election to the Senate -- tripling the number of elected women currently serving...." (November 4, 1992)

Why was there this shift in 1992 toward the frames of Year of the Woman and women as outsiders, and how do we deconstruct the conventional meaning of the Year of the Woman frame? Four new women did win Senate seats in 1992. Twenty-four new women were elected to the House, raising the number of women there from 29 to 47. Women continued to increase their numbers in state legislatures, as they did in previous and subsequent elections.

But women remained far from parity. Most women who challenged House or Senate incumbents lost, often by large margins. Moreover, it was not the year for women of both parties; only 3 nonincumbent Republicans who ran for Congress won.

Women's representation in Congress remained low compared with legislatures in many established democracies. If we compare the proportion of women in the House with comparable legislatures, there are twice as many women in Canada, Austria, and Germany; three times as many in the Netherlands and Denmark; four times as many in Sweden, Norway, and Finland (Norris 1996). The 1992 increases brought the U.S. up to the world average, no better. In this context, the election could have been framed equally accurately as the year where women's gains in the United States were much as might be expected from worldwide secular trends, or the year where America did much worse than many comparable countries. Instead, the frame overwhelmingly emphasized the positive news of "dramatic" gains.

This is not to dispute that there were important gains for women in office in 1992, but rather to challenge the hype and exaggeration which can leave readers with the impression that the change was more dramatic than it really was. In many ways, the real surprise was that the 1992 increase had not happened earlier or gone further. Similar developments in the 1992 British general election, where women Members of Parliament (MPs) increased from 6.2 to 9.6 per cent, were covered by the press in one or two articles at most and were regarded by women activists as relatively disappointing. It was only when the proportion of women MPs doubled in Britain, surging to 18 percent following the 1997 election, that journalists started to sit up and take notice.

Moreover, although some new women members of Congress were relatively new to politics, others had accumulated a considerable track record of political experience and could not accurately be termed political "outsiders." And while many women candidates might well be agents of change who bring distinctive perspectives to politics that have been previously underrepresented, they bring those perspectives in every election, not just in 1992. In short, the "Year of the Woman" frame for the 1992 U.S. elections, along with the women as outsiders frame that often accompanied it, offered only one of several possible interpretations for the outcome of the elections.



Conclusions

We set out to establish whether we could identify certain predominant frames used by the news media to structure stories about gender politics in election campaigns; to analyze how these frames varied over time; and to compare these frames against a benchmark based on aggregate and survey data to see how far changes in news frames reflect changes in the reality of gender politics.

We can conclude that the news media often tend to use relatively simple frames as a way of conveying to their readers a straightforward narrative, written around a particular 'peg' or story-line. Yet we did find that, much like the fashion industry's use of skirt hem-lines, the process was dynamic. In some elections a predominant gender-related frame was apparent in many stories (eg the 'Angry White Male' frame in the aftermath of 1994). In other campaigns, we identified multiple frames, with the 'gender gap' frame usually providing the richest source of stories. The predominance of shared frames as a way of writing about, and making sense of, different races, candidates and issues is not really surprising. Nevertheless this process disguises both the complexities of gender politics, and the long-term and persistent patterns which are repeated every election year. As we have noted the difference in voting choice between American women and men which first emerged in 1980 has persisted as a long-term pattern in every subsequent general election. Insofar as news should be 'new', the gender gap story does not really qualify after the early 1980s. Nevertheless we have documented how the gender gap story has been retold, with suitable permutations which freshen and dramatize the familiar pattern, (eg the Soccer Mom) in every election. Usually the narrative emphasizes Democratic support among women, although occasionally this is stood on its head as the headlines focus on Republican support for men. But usually the result of the framing process is to make the difference between women and men larger and sharper and brighter and newer than it really is according to the pattern in survey data. In the same way, journalistic attention to stories about women candidates tend to rise and fall in cycles, but often to emphasize similar themes. Moreover, we can speculate that perhaps there is an even broader dynamic whereby press attention cycles between stories about women as voters and women as candidates in successive campaigns. If so, this would suggest that interest in Soccer Moms having subsided, we might well expect some variation of a rediscovery of women as candidates in 1998.

And, it can be argued, that this framing process does have important consequences. The very pervasiveness of gendered frames means that this also starts to invade political discourse among politicians, pollsters, party managers, political scientists, and therefore perhaps the way that we understand the issues symbolized by any electoral choice. 'Angry White Males' and 'Soccer Moms' may be political myths, compared with the hard numbers in surveys, but they take on a reality if candidates start to believe in them, and thereby shape their political strategy accordingly. While journalists probably cannot avoid this process, if they are to communicate with their audience, nevertheless we should become aware of how framing shapes our understanding of gender politics.
 



 
 

Table 1: Proportions of Women and Men Who Voted for the Democratic Candidate for the U.S. House*
% Women  % Men 
1982  57  54 
1984  49  45 
1986  54  51 
1988  48  45 
1990  52  48 
1992  55  52 
1994  54  43 
1996  53  44 
Source: *Data for 1992 through 1996 are from the Voter News Service exit polls; data for 1982 through 1990 are from exit polls conducted by CBS News/New York Times.



Table 2: Numbers of Women Candidates for Various Offices, 1982-1996*
U.S. Senate  U.S. House  Governor  Lt. Governor  State Legislature 
1982  3  55 2 7  1643
1984  10  65 1 6  1756
1986  6  64 8 11  1813
1988  2  59 2 2  1853
1990  8  69 8 19  2064
1992  11  106 3 7  2375
1994  9  112 10 29  2285
1996  9  120 6 9  2274
Source: *Figures furnished by the Center for the American Woman and Politics.


Table 3: Numbers of Women Serving in Legislative Offices, 1983-1997*
U.S. Senate  U.S. House  State Legislature 
1983  2  22  991 
1985  2  23  1103 
1987  2  23  1170 
1989  2  29  1270 
1991  4  28  1368 
1993  7  47  1524 
1995  9  48  1532 
1997  9  50  1598 
Source: *Data from the Center for the American Woman and Politics.

The U.S. Senate and House figures represent the total number of women to serve in each corresponding session of Congress.



Table 4: The Rise and Fall of Gender-Related Phrases
Year  1980  1982 1984 1986  1988 1990 1992  1994 1996
Gender Gap  0  31 44  25 85  63 124  95 547 
Year of the Woman  0  0 3  1 3  30 584  134 43
Angry White Male *  0  0 0  0 0  0 0  19 32 
Soccer Mom  0  0 0  0 0  0 0  0 355 
Total  0  31 47  26 88  93 718  248 977
Note: Nexis-Lexis keyword search of "Major Papers" database using term "_____ and election" from September 1 through November 30 per election year. Table entries reflect number of stories using these terms.

*It should be noted that since the first reference to this phrase seems to have occurred after the 1994 election, the main surge in stories using this frame is not reflected in this table since it occurred in 1995.



Table 5: Frames in The New York Times and Washington Post
Frame  1982  1984 1986 1988  1990  1992 1994  1996 
Gender Gap 

Soccer Mom 

Angry White Male 

Record No. of Women 

First Woman 

Bad Year for Women 

Year of the Woman 

Outsiders/Agents of Change 

33.3 

--- 

--- 

4.8 

4.8 

4.8 

--- 

--- 

24.3 

--- 

--- 

--- 

21.6 

16.2 

--- 

--- 

25.0 

--- 

--- 

37.5 

6.3 

--- 

--- 

--- 

100.0 

--- 

--- 

--- 

--- 

--- 

--- 

--- 

13.3 

--- 

--- 

20.0 6.7 

--- 

6.7 

26.7 

9.1 

--- 

--- 

3.0 

--- 

--- 

15.2 

18.2

6.3 

--- 

18.8 

--- 

--- 

18.8 

6.3 

6.3 

49.0 

8.2 

--- 

--- 

--- 

2.0 

--- 

--- 

Total =  21  37 16 13  15  33 16 49
Note: Proportions of Articles about gender politics in elections in The New York Times and The Washington Post Employing Different Gender Related Frames, 1982-1996

Source: Lexis-Nexis search


Notes:

Bendyna, Mary E. and Celinda C. Lake, 1994. 'Gender and Voting in the 1992 Presidential Election' in Elizabeth Adell Cook, Sue Thomas and Clyde Wilcox (eds.) The Year of the Woman Boulder: Westview Press.

Bennett, W. Lance, 1996. News: The Politics of Illusion (3rd Ed.) New York: Longmans.

Bird, Elizabeth and Robert W.Dardenne, 1997. 'Myth, Chronicle and Story: Exploring the Narrative Quality of News' in Dan Berkowitz (ed.) Social Meaning of News Thousand Oaks, CA: Sage.

Bonk, Kathy, 1988. 'The Selling of the "Gender Gap": The Role of Organized Feminism' in Carol M. Mueller (ed.) The Politics of the Gender Gap Newbury Park, CA: Sage.

Borquez, Julio, Edie N.Goldenberg, and Kim Fridkin Kahn, 1988. 'Press Portrayals of the Gender Gap' in Carol M. Mueller (ed.) The Politics of the Gender Gap Newbury Park, CA: Sage.

Campbell, Angus, Philip E. Converse, Warren E. Miller and Donald E. Stokes, 1960. The American Voter New York: John Wiley & Sons, Inc.

Carroll, Susan J., 1997. "My Kids Play Soccer and I Vote": The Social Construction of the Soccer Mom in the 1996 Elections" Paper presented at the 1997 Annual Meeting of the American Political Science Association, Washington DC August 28-31.

CAWP, 1997a. The Gender Gap Fact Sheet. New Brunswick, NJ: Center for the American Women and Politics.

CAWP, 1997b. Statewide Elective Executive Women 1997. New Brunswick, N.J.: Center for the American Woman and Politics.

Clark, Cal and Janet Clark, 1996. 'Whither the Gender Gap? Converging and Conflicting Attitudes among Woman' in Lois Lovelace Duke (ed.) Women in Politics (2nd ed.) Upper Saddle River, NJ: Prentice Hall.

Cohen, Akiba A. and Gadi Wolfsfeld, (eds). 1993. Framing the Intifada: People and Media. Norwood, NJ, Ablex Publishing.

Cook, Elizabeth Adell and Clyde Wilcox, 1991. 'Feminism and the Gender Gap - A Second Look' The Journal of Politics 53(4): 1111-1123.

Conover, Pamela Johnston, 1988. 'Feminists and the Gender Gap' Journal of Politics 50(4): 985-1010.

Costain, Anne, Richard Braunstein, and Heidi Berggren, 1997. 'Framing the Women's Movement' in Pippa Norris (ed) Women, Media and Politics New York: Oxford University Press.

Creedon, Pamela J., 1993. 'Framing Feminism - A Feminist Primer for the Mass Media' Media Studies Journal 7(1): 68-81.

Duverger, Maurice, 1955. The Political Role of Women. Paris: UNESCO.

Entman, Robert, 1996. 'Reporting Environmental Policy Debate: The Real Media Biases' The Harvard International Journal of Press/Politics 1(3):77-92.

Entman, Robert, 1993. "Framing: Toward Clarification of a Fractured Paradigm" Journal of Communication 43(4): 51-58.

Fico, Frederick and Stan Soffin, 1995. 'Fairness and Balance of Selected Newspaper Coverage of Controversial National, State and Local Issues' Journalism and Mass Communication Quarterly 72(3): 621-633.

Fink, Conrad C., 1995. Media Ethics Boston: Allyn and Bacon.

Gamson, William, 1991. Talking Politics. New York: Cambridge University Press.

Gans, Herbert J. 1979. Deciding What's News New York: Pantheon Books.

Gitlin, Todd, 1980. The Whole World is Watching. Berkeley: University of California Press.

Gitlin, Todd, 1994. Inside Prime Time. London: Routledge.

Graber, Doris, A. 1997. Mass Media and American Politics Washington, DC: CQ Press.

Hermann, Edwards S. and Noam Chomsky, 1988. Manufacturing Consent New York: Pantheon Books.

Hertsgaard, Mark, 1989. On Bended Knee: The Press and the Reagan Presidency New York: Farrar, Straus and Giroux.

Iyengar, Shanto, 1991. Is Anyone Responsible? Chicago: University of Chicago Press.

Kaufmann, Karen M. and John R. Petrocik, 1997. The Revenge of the Soccer Moms? Gender as a Party Cleavage in American Politics. Paper presented at the Annual Meeting of the American Political Science Association, Washington DC.

Lacy, Stephen, Frederick Fico and Todd F. Simon, 1991. 'Fairness and Balance in the Prestige Press' Journalism Quarterly 68: 363-70.

Ladd, Everett Carll, 1997. 'Media Framing of the Gender Gap' in Pippa Norris (ed.) Women, Media and Politics. New York: Oxford University Press.

Lee, Martin A. and Norman Solomon, 1992. Unreliable Sources: A Guide to Detecting Bias in News Media New York: Lyle Stuart.

Lichter, Robert, Stanley Rothman and Linda S. Lichter, 1986. The Media Elite New York: Adler and Adler.

McQuail, Denis, 1992. Mass Communication Theory London: Sage.

Mueller, Carol, 1991. 'The Gender Gap and Women's Political Influence' Annals of the American Academy of Political and Social Sciences 515: 23-37.

Naureckas, Jim and Janine Jackson, 1996. The Fair Reader: An Extra Review of Press and Politics in the '90s Boulder, Co: Westview Press.

Norris, Pippa, 1997a. 'Introduction' in Women, Media and Politics. New York: Oxford University Press.

Norris, Pippa, 1997b. 'News of the World' in Pippa Norris Politics and the Press Boulder: Rienner.

Parenti, Michael, 1986. Inventing Reality: The Politics of the Mass Media New York: St. Martin's Press.

Patterson, Thomas E., 1994. Out of Order New York: Vintage.

Reeher, Grant and Joseph Cammarano, 1996. 'In Search of the Angry White Male: Gender, Race and Issues in the 1`994 Elections' in Philip A. Klinkner (ed) Midterm Elections of 1994 in Context. Boulder, Co: Westview Press.

Sabato, Larry. 1988. Feeding Frenzy. New York:

Shapiro, Robert Y. and Harpreet Mahajan, 1986. 'Gender Differences in Policy Preferences: A Summary of Trends from the 1960s to the 1980s' Public Opinion Quarterly 50: 41-61.

Simon, Todd F., Frederick Fico and Stephen Lacy, 1989. 'Covering Conflict and Controversy: Measuring Balance, Fairness and Defamation' Journalism Quarterly 62: 427-434.

Weaver, Paul, 1972. 'Is Television News Biased?' The Public Interest 27.

Weaver, David H. and G. Cleveland Wilhoit, 1996. The American Journalist in the 1990s Mahwah, NJ: Lawrence Erlbaum.

Wirls, Donald, 1986. 'Reinterpreting the Gender Gap' Public Opinion Quarterly, 50: 316-30.

Zaller, Hjohn, 1992. The Nature and Origins of Mass Opinion. New York: Cambridge University Press.

Notes

This research was completed with the assistance of a Goldsmith Research Award (to Susan Carroll) from the Joan Shorenstein Center on the Press, Politics and Public Policy at Harvard University's Kennedy School of Government. Both authors wish to thank Krista Jenkins, David Jones and Bill Davies for their very able assistance in coding and data entry. \