Article PartyPolitics Party Politics . - _ . . . , _ . 2015, Vol. 21(4) 626-637 To the left, to the right: Representing ©^^^2013 ' Reprints and permission: conservative women's interests ^MtJ™^" ppq.sagepub.com Rosie Campbell University of London, UK Sarah Childs Bristol University, UK Abstract The study of conservative women is expanding to compensate for the historic over-emphasis of gender and politics research on left-wing women. We add to this burgeoning literature and assess the extent to which the modern gender gap in political attitudes - where women have moved to the left of men - is evident among supporters of the British Conservative Party. We find that, like women party members, women Conservative supporters are noticeably to the left of men, but only on economic issues. This sex gap cannot simply be accounted for by women's employment in the public sphere, lesser interest in politics or because they are more morally conservative than male Conservatives. These findings are likely to have serious implications for intra-party discipline, the support for the Conservative Party at the next British general election and, if replicated elsewhere, speak to what it means to 'represent' the interests of right-wing women. Keywords Conservative, gender, ideology, representation Introduction The study of the 'gender gap' in attitudes and party of vote has developed in response to the high profile gender gap in US elections; a greater proportion of women than men have voted for every Democratic presidential candidate since 1980 (Andersen, 1999; Box-Steffensmeier et al., 1997; Burden, 2008; Burns, 2001; Carroll, 1988, 1999; Chaney et al., 1998; Conover, 1988; Duke Whitaker, 2008; Edlund and Pande, 2002; Mueller, 1988; Norris, 2001). Interest in the subject has spawned an international literature that assesses the extent to which the US gender gap can be found elsewhere (Campbell, 2006; Gidengil and Harell, 2007; Ingle-hart and Norris, 2000; Norris, 1999; Wangnerud, 2000). The research has become ever more nuanced and sophisticated, with complex theoretical models employed to explain why men and women might prefer different political parties (Alvarez and McCaffery, 1999; Burden, 2008; Dolan, 2010; Greenberg, 2001). To some extent, the British case remains an outlier in this literature. There is little in the way of a 'gender gap' in political attitudes or behaviour in Britain (Campbell, 2006, 2012) and British research testing Norris and Inglehart's claim that women across the developed world are moving to the left has produced mixed results (Campbell, 2006, 2012; Hayes, 1997; Inglehart and Norris, 2000; Norris, 1999; Steel, 2003). It is apparent that women in Britain have not simply shifted their allegiances from the Conservatives to the Labour Party, and although the Labour Party had some advantage among younger women in 1997 and 2001 there is little evidence to show that this was sustained in 2005 and 2010 (Campbell, 2012). The British case therefore provides an interesting example for furthering the study of gender and political attitudes and behaviour. The absence of an aggregate sex gap among voters does not mean that sex and gender differences do not play out in UK electoral politics. Recent research has shown that the aggregate level similarity between men's and women's political attitudes may mask intra-party differences; a recent study of Conservative Party members established that women party members are significantly to the left of Paper submitted 07 August 2012; accepted for publication 21 April 2013 Corresponding author: Rosie Campbell, Department of Politics, Birkbeck, University of London, Malet Street, Bloomsbury, London WCIE 7HX, UK. Email: r.campbell@bbk.ac.uk Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 Campbell and Childs 627 men on a range of economic issues (Childs and Webb, 2009, 2012; Webb and Childs, 2012). This sex gap in Conservative attitudes is interesting in and of itself, but it is also potentially increasingly politically salient in the current economic climate and in a context of massive planned public expenditure cuts. Should Childs and Webb's findings be replicated beyond the membership to the party's support base, the implications are likely to be more significant not only for future intra-party discipline (the focus of Childs and Webb's interest), but also for the possible re-emergence of a 'modern' gender gap in vote choice at the next British general election. Using the British Election Study 2010^6 asses the extent to which Childs and Webb's observed Conservative sex-gap among party members is also evident among Conservative Party supporters and identifiers; in other words, among the Conservative's core voters. Context The 2010 United Kingdom general election ushered in a Conservative-led Coalition government. Having been out of power for more than a decade, the Conservative Party had engaged in a process of modernization and re-positioning in order to be more electorally competitive (Bale, 2010). One part of this 'detoxification' process involved a conscious decision to feminize the party (Childs and Webb, 2012). The new leader, and now Prime Minister, David Cameron was explicit about his desire to both rectify the 'scandalous' under-representation of women in his parliamentary party and to win back women voters.2 He was focused especially on attracting the votes of younger women and working mothers with a series of what can be categorized as liberally feminist policies, not least inrespect of equal pay, flexible working and maternity and paternity leave and pay (Campbell and Childs, 2010).3 Since the election, however, the Conservatives, as the leading party in the current Coalition, have been on the receiving end of repeated gendered criticism, not least for failing to see how the government's deficit-reduction plans disproportionately and negatively affect women. Theresa May, then Minister for Women and Equality, as well as Home Secretary, first warned the Treasury that it must undertake a gender audit of its emergency budget back in 2010. Labour's Yvette Cooper (Shadow Minister for Women),4 the Fawcett Society and the Women's Budget Group5 have all since produced figures to show that Coalition cuts to public services, benefits and public sector j obs have been overwhelmingly negative for women. Women make up 65 percent of public sector employees in Britain and the Office for Budget Responsibility has forecast that the public payroll will be reduced by 710 000 by 2015 (Office for Budget Responsibility, Autumn 2011).6 The March 2012 budget similarly came under attack for targeting women, with inter alia, cuts to child tax credits, child benefit and pension income tax benefits.7 Defenders of the government refute charges that the Coalition is ideologically 'anti-women'. Rather the differential impacts are said simply to reflect the fact that women are the greater users, receivers and employees of the state - 'collateral damage' perhaps, but with little to do with gender politics perse. In contrast, Coalition critics have raised the questionof how the government - both Conservative and Liberal Democrat - appears not to have noticed that its policies would impact most heavily on one sex. Was this a failure of descriptive and substantive representation, of Coalition women representatives being too few, and of failing to 'act for' women?8 The all-important 'quad' of leading Conservative and Liberal Democrat Cabinet Ministers are all male; women numbered only five in the Coalition's first Cabinet, four in its second, and are all Conservative women MPs; and, on the government benches, women make up only 16 percent of the parliamentary Conservative party and a mere 12 percent of Liberal Democrat MPs are women, compared with more than 30 percent in the parliamentary Labour Party. Adding to feminist academic analysis and Westminster village talk, there has also been extensive media copy suggesting that women voters are turning away from the Conservative Party.9 Although the evidence from the polls is variable, it is clear that within the Conservatives there is much concern, and indeed reaction, over this possibility: a 'women's policy advisor' was appointed (February 2012) to coordinate the ' fight back' - although this was only part of her brief; MPs, mostly but not exclusively women, have established a Conservative Party Women's Forum (2011); and a series of mini policy announcements, for example on forced marriage, have been pushed out, seemingly to ' fill' in the gap in the government programme. Even so, there remains a nagging question in the academic and public realm as to whether the Cameronian Conservative Party has effectively walked away from its commitment to gender equality in the face of an economic context that demands, in the view of the party and coalition leadership, a very particular set of responses: read, austerity economics.10 We do not attempt to answer the normative question as to whether the Conservative Party is failing to recognize women's different political priorities, but we do investigate whether there is a gender dimension to public attitudes to economic issues, particularly cuts in public expenditure among Conservative supporters and identifiers. In so doing, we speak to wider concerns of gender politics scholars and political scientists in the UK and beyond. This article extends analysis that sought to capture the feminization efforts of the Conservative Party between 2005 and 2010 (Childs and Webb, 2012). One part of this earlier study investigated the political attitudes of party members and elites prior to the 2010 general election.11 Informed by existing research that points to intra-party sex differences across all the main parties on the 'women's terrain', Childs and Webb established that Conservative women party members are visibly to the left of men on economic issues - issues that are rarely considered to be directly or explicitly gendered - in addition to their more widely known more feminist positioning on explicitly gendered issues Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 628 Party Politics 21 (4) (Childs and Webb, 2012). Childs and Webb found that on a seven-point self-placement left-right measure, women placed themselves slightly to the left of men, closer to the neutral category; the mean score for women was 5.22 and 5.38 for men (Childs and Webb, 2012: 124). They also found that women were slightly to the left of men on Heath et al.'s socialist/laissez-faire scale (1993) and slightly more authoritarian on the liberal authoritarian scale (Childs and Webb, 2012: 125). The biggest sex differences were in response to two items, censorship and attitudes to 'big business'. In total, 18 percent more women than men agreed with the statement 'censorship of films and magazines is necessary to uphold moral standards' and 5 percent more women than men agreed with the statement 'Big business benefits owners at the expense of workers'. Childs and Webb ran an OLS regression with the tax/ spend item as the dependent variable and found a sex effect even after age, social grade, terminal education age and ideological grouping were controlled for (Childs and Webb, 2012: Table 7.9). Our analysis, jumping off from Childs and Webb's analysis, centres on the extent to which the left/right sex gap on economic issues among Conservative members is also evident among Conservative supporters and identifiers. As such, this article adds to our knowledge about the variation between elite and mass attitudes, an established area of political science (Converse, 1964; Fleishman, 1988; Jennings, 1992; Nie and Andersen, 1974) which has been mostly concerned with ideological consistency and issue constraint. Most studies of elite/mass attitudes focus on whether political elites have more coherent, internally consistent sets of attitudes and beliefs than ordinary members of the public. Here, instead, we are assessing whether the sex differences found among a political elite (Conservative Party members) by Childs and Webb are also evident at a mass level (among Conservative supporters and identifiers). Conceptually, we move away from a two-tier elite/mass comparison and instead conceive of a more complex hierarchy of political commitment combined with power; imagining political commitment and power as a pyramid. The least committed - those people for whom the height of political activity is to vote occasionally - form the base (the party voter), and sitting at the apex are those whose entire lives are consumed by political activity {the elected Members). The hierarchy is also delineated by political power; ordinary voters with the least, and elected politicians with the most. In between these two groups -voters and MPs - are party supporters, party members and parliamentary candidates. Party supporters retain a loyalty to the party between elections but have little direct power over party decision-making; party members sign up and pay their dues and have a vote in the selection of their local parliamentary candidate; parliamentary candidates devote time and money to their selection and election campaigns and are most likely plugged into party networks. Rather than considering voters as a whole, as many elite/mass studies do, we compare two layers from this pyramidal account: party members and supporters/identifiers. Future research will collect data on MPs and parliamentary candidates (in a study of the 2015 British General Election). As such, this research is the first step towards more fully understanding how gender interacts with political attitudes across all of the layers in the hierarchy of the British Conservative Party. Sex differences and possible explanations There is a popular currency to deride Conservative women's politics as the politics of the falsely conscious; to talk of 'Stepford Wives'. Indeed, the title of Beatrix Campbell's analysis of women Conservatives in the 1980s asked the simple question 'Why do women vote Tory?' Although she took the opinions of the Conservative women she interviewed seriously, it is not unfair to suggest that Conservative women's politics are often regarded by feminists and those on the left as suspect. What they need - and we parody here - is some good old fashioned consciousness-raising. A similar approach has been taken by scholars of the gender generation gender gap, who regard right-wing women's 'traditional' values as a remnant from times past when women were largely confined to the private sphere, with restricted access to higher education and paid employment (Campbell, 2006; Inglehart and Norris, 2000; Norris, 1999). From this perspective, Conservative women voters might still be understood to be suffering from a false consciousness that will, over time, slowly die out. The decline of the traditional gender gap in the UK and the emergence of a gender generation gap which broadly, but not completely, maps onto the period in office of New Labour (1997-2010) looked to be some evidence of this; mirroring the emergence of the gender gap found in the US. However, analysis of the 2010 British general election suggests that the UK Conservatives had been rather effective at restoring their traditional advantage among women voters (Campbell, 2012; Campbell and Childs, 2010; Childs and Webb, 2012). Explaining away Conservative support among women as a historic legacy now looks rather premature. In which case, before proceeding to the analysis, it is helpful to think about why we might expect to see such a sex difference and to develop testable hypotheses. First, there is an argument that can be made from the international literature. In a number of Western industrialized countries women's political attitudes have moved to the left of men's (Inglehart and Norris, 2000). British women, too, tend to have more egalitarian attitudes to gender relations, homosexuality and racial discrimination (Campbell, 2011; Campbell et al., 2010; Ford, 2008), and some studies have shown that younger women tend to have more left-leaning attitudes than younger men, although these differences are not persistent over time and across surveys (Campbell, Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 Campbell and Childs 629 2006). Nevertheless, the sex difference among the attitudes of Conservatives found in this article in respect of the economy might possibly be accounted for by a more general trend evident within the whole population. Hypothesis 1: The sex gap in left-right economic attitudes among Conservative supporters and identifiers can be explained by more left-leaning attitudes held by women than men in the general population. There are several reasons why we might expect women to be to the left of men on economic issues. Women are the majority of primary carers of young children and it can be argued that this might shift them to the left, politically; perhaps because the psychological impact of mothering may make women more altruistic (Lehman et al., 1995; Ruddick, 1989), or because they tend to give priority to state services such as health and education for their children rather than reductions in taxation. Hypothesis 2: The sex gap in left-right economic attitudes among Conservative supporters and identifiers can be explained by the impact of parental status on women's attitudes. In addition, women tend to be financially less well off after divorce than men and tend to be over-represented in lower income households, and these conditions may account for any left-leaning tendency among women. Hypothesis 3: The sex gap in left-right economic attitudes among Conservative supporters and identifiers can be explained by women's average lower household income. An alternative approach emphasizes women's authoritarian attitudes relative to men on some issues; particularly on censorship, which women are still considerably more likely to support than men (Campbell, 2006). We might argue, in light of this, that perhaps Conservative women who are economically to the left are at the same time more authoritarian on social issues. This hypothesis reflects accounts that identify women's greater religiosity and moral conservatism as one of the causes of the 'traditional gender gap'(Norris, 1999) - the Conservative's historic advantage among women voters, which was at its peak in the post-war years but declined to negligible levels in the late 1970s. From this perspective women might be drawn to the Conservative Party because of its association with traditional moral values rather than laissez-faire economics. Accordingly, it may be that the subgroups of women who vote for, support, or are members of the Conservative Party are more likely to be drawn from a moral right. If this is the case their views of economic issues might be less important to their identification as Conservatives than their liberal/authoritarian position. If this is the case, the implications for the Conservative Party electorally may be lessened, as women may continue to feel at home in, and supportive of, a Conservative Party that articulates their interests on the liberal/authoritarian axis even as it diverges from them on the economic axis. Hypothesis 4: The sex gap in left-right attitudes towards economic issues among Conservative supporters and identifiers results from more authoritarian values among women than among men. A third potential explanation for the sex difference in Conservatives' political attitudes on the economy might stem from the sex segregation of employment sectors that characterizes the UK. Given that we know women are significantly more likely than men to be employed in the public sector, if women associated with the Conservative Party are also more often employed in the public sector than men, then they might well be more likely to support public spending rather than reduced taxation. In other words, we might be able to explain away the sex difference over tax and spend among Conservatives by reference to another causal factor: public sector employment. Sex, in this case, is not the driver of attitudinal dispositions. Hypothesis 5: The sex gap in left-right attitudes towards economic issues among Conservative supporters/identifiers is an indirect effect of the larger proportion of women employed in the public sector than men and the tendency for public sector workers to be to the left on economic issues vis-á-vis private sector workers. Finally, any sex difference we find might be an artefact of men's generally greater interest in politics than women. It is well established that women report less interest in politics than men and tend to do less well in political knowledge quizzes (Andersen, 1975; Burns, 2001; Campbell and Winters, 2008; Frazer and MacDonald, 2003; Mondak and Anderson, 2004). Consequently, it is important to identify whether any sex gaps we find in political attitudes are generated by women placing themselves disproportionately in the 'neutral' categories because they have less interest in the survey questions. Hypothesis 6: The sex gap in left-right attitudes towards economic issues among Conservative supporters/identifiers results from more women than men placing themselves in the middle or neutral category. Analysis Conservative supporters Turning to conservative supporters, we first examine supporters' self-placement on the tax/spend question in the face-to-face post-election 2010 British Election Study Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 630 Party Politics 21 (4) 20.0%- Strongly agree Tend to agree Nether agree nor disagree Tend to disagree Government should cuttaxesfspending a lot: Your view? Figure I. Responses to the tax/spend item in the Conservative members survey. N = 1690 Differences significant at the 0.001 level chi square test. 60.0%- c Sen •-•Men —Women ri cl J, J, Cut taxes and spending or increase taxes and spending Figure 2. Conservative supporters' responses to tax/spend question by sex, 2010 post-election face-to-face BES. Sex difference significant at the 0.001 level. N = 827. (BES). Respondents to the survey were asked to place themselves on a zero to 10 scale. Zero indicated that 'government should cut taxes and spend much less on health and social services'; 10 that 'government should raise taxes a lot " a>o«- m a. 10.0«- Sex ■ ■ Men — Women 3 4 5 6 7 8 S Rignt 10 Left-Right Self Placement Cases weighted by Weight for mailhack respondents Figure 3. Conservative supporters self-placement on the left/ right scale by sex, 2010 post-election mailback BES. N = 170 The difference between the sexes' responses is not statistically significant. and spend much more on health and social services'. The reference to health in the item is most likely to produce sex differences, as women tend to report greater interest in health policy issues than men (Campbell, 2006; Campbell and Winters, 2008; Wangnerud, 2000). There is a small, but statistically significant difference between men and women Conservative supporters' responses: the mean response for men was 5.09 and 5.83 for women.12 Examination of Figure 2 shows a peak in the responses for both men and women in the middle of the scale and the differences between the sexes are not enormous. However, it is clear that women more often than men favoured increasing services, and men more often than women favoured cutting taxes.13 Furthermore, this sex difference cannot be accounted for by the fact that women in the whole population generally respond to this question to the left of men: among the entire sample, women were on average just 0.37 of a point further towards spending more than men (half the Conservative sex gap). Thus, comparing Conservative supporters against the whole distribution, and to other parties, reinforces the claim that Conservative women supporters are to the left of men on the tax/spend issue and refutes hypothesis 1, namely, that the sex gap among Conservative supporters is an artefact of an overall sex gap in the population. Respondents to the 2010 BES mail-back survey were also asked to place themselves on a left-right scale.14 There was no statistically significant mean difference between Conservative supporting men and women, although women placed themselves marginally to the left of men (see Downloaded from ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 Campbell and Childs 631 Table I. Socialist/laissez-faire scale items by sex, 2010 post-election mailback BES. Sex Agree strongly Agree Neither agree nor disagree Disagree Disagree Strongly Total n Private enterprise is the best way to solve Britain's Men 19.0% 50.9% 23.3% 4.9% 1.8% 100% 163 economic problems*** Women 7.0% 43.0% 36.0% 13.4% .5% 100% 186 There is one law for the rich and one for the poor* Men 9.1% 37.8% 15.9% 29.3% 7.9% 100% 164 Women 9.2% 35.1% 22.2% 31.9% 1.6% 100% 185 Ordinary working people get their fair share of the Men 4.4% 26.3% 15.6% 49.4% 4.4% 100% 160 nation's wealth** Women 1.1% 14.0% 21.5% 56.5% 7.0% 100.0% 186 There is no need for strong trade unions to protect Men 4.9% 25.8% 30.7% 35.6% 3.1% 100.0% 163 employees' working conditions and wages Women 4.4% 16.5% 35.7% 40.7% 2.7% 100% 182 ***Sex difference significant at the 0.001 level chi square test. **Sex difference significant at the 0.01 level chi square test. *Sex difference significant at the 0.05 level chi square test. Table 2. Liberal/authoritarian scale items by sex, 2010 post-election mailback BES. Agree Neither agree Disagree Sex strongly Agree nor disagree Disagree Strongly Total N Young people today don't have enough respect for Men 31.3% 49.1% 8.0% 1 1.7% .0% 100% 163 traditional British values** Women 19.8% 58.8% 12.8% 6.4% 2.1% 100% 187 Censorship of films and magazines is necessary to Men 9.7% 43.0% 20.6% 22.4% 4.2% 100% 163 uphold moral values*** Women 20.0% 53.0% 16.2% 10.8% .0% 100% 185 People should be allowed to organize public meetings Men 1 1.7% 66.9% 16.6% 4.3% .6% 100% 163 to protest against the government*** Women 9.1% 50.8% 29.4% 10.7% .0% 100.0% 187 People in Britain should be more tolerant of those who Men 3.7% 34.6% 37.0% 22.8% 1.9% 100% 162 lead unconventional lifestyles Women 2.2% 33.0% 44.9% 18.9% 1.1% 100% 185 Political parties that wish to overthrow democracy Men 2.5% 17.4% 9.9% 46.6% 23.6% 100.0% 161 should be allowed to stand in general elections*** Women 0.5% 15.8% 25.1% 39.9% 18.6% 100% 183 The government has the right to put people suspected Men 13.4% 37.8% 18.9% 26.8% 3.0% 100% 164 of terrorism in prison without trial Women 14.4% 41.7% 16.6% 21.4% 5.9% 100% 187 ***Sex difference significant at the 0.001 level chi square test. **Sex difference significant at the 0.01 level chi square test. *Sex difference significant at the 0.05 level chi square test. Figure 3). This finding suggests that when asked directly about whether they think of themselves as to the left or to the right, Conservative women do not appear to 'think' of themselves as being as leftist as their answers to the tax/ spend questions suggest they are. Why this is the case is unclear: perhaps women 'mis-label' themselves, denying just how to the left they are. Alternatively, they might be basing their self-placement on issues other than, or in addition to, tax and spend. For example, if you are to the right on issues on the liberal-authoritarian axis, your preference regarding tax and spend might be less relevant to your own understanding of your overall political position. To put it another way, in thinking about being 'left' or 'right' voters might well imagine a '2 by 2' left/right and social authoritarian/libertarian model, and not just a single left/ right economic dimension. The mail-back component of the 2010 face-to-face BES also includes items designed to measure socialist/laissez-faire and liberal/authoritarian positions (Evans et al., 1996; Heath et al., 1993; Tilley, 2005). While the list of items is not identical to those used by Childs and Webb (2012) in the party member survey there is some overlap, and the items are designed to measure the same latent variables (authoritarianism and socialism). The responses to the items are presented in Tables 1 and 2. Starting with Table 1, of the four items that make up the socialist/laissez-faire scale in the 2010 BES there is a statistically significant sex difference in responses to three of the four items.15 But, crucially, most of this difference is accounted for by women more often placing themselves in the neutral category. When asked whether 'Private enterprise is the best way to solve Britain's economic problems', 20 percent more men than women Conservative supporters agreed strongly or agreed with the statement. However, only 8 percent more women than men are to be found in the disagree categories. Instead, women more often placed themselves in the 'neither' category than men (by 13 percent). In response to the statement 'There is one law for the rich and one for the poor' women were also more often found in the 'neither' category than men; 6 percent more women than men selected 'neither' and roughly 6 percent more men than women disagreed strongly with the statement. There is not a statistically significant sex difference in the way men and women Conservative supporters responded to the Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 632 Party Politics 21 (4) Table 3. OLS regression on tax/spend, BES 2010 post-election face-to-face survey. Independent variables Model 1 Model 2 Model 3 B (SE) Beta B(SE) Beta B(SE) Beta Women 0.74***(0.15) 0.20 0.82***(0.22) 0.23 0.53 **(0.I7) 0.15 Private sector employment -0.08 (0.24) -0.02 Interest in politics 0.10 (10) 0.05 Cares for children <5 0.1 1 (0.27) 0.02 Cares for children< 15 0.26 (22) 0.06 Degree holder -0.46 (0.28) -0.08 Age 0.02** (0.01) 0.14 Household income -0.06* (02) -0.13 N 583 271 449 Adjusted R2 0.04 0.05 0.07 ***Sex difference significant at the 0.001 level. **Sex difference significant at the 0.01 level. *Sex difference significant at the 0.05 level. statement 'There is no need for strong trade unions to protect employees' working conditions and wages'. Finally, with regard to the statement 'Ordinary working people get their fair share of the nation's wealth', there is a statistically significant sex difference with 15 percent more men than women agreeing strongly or agreeing with the statement, 6 percent more women than men giving a 'neither' response and approximately 10 percent more women disagreeing or disagreeing strongly with the statement. Thus, overall, there is evidence of a small sex gap among Conservative Party supporters with women responding to the ideological left of men on the Socialist/laissez-faire scale, but the substantive difference largely relates to just one item -' ordinary people'. When the responses to the four items are added together to create a cumulative socialist/laissez-faire scale, there is not a significant sex difference among Conservative supporters. Moving to consider Table 2, we consider liberal/author-itarian position.16 There are statistically significant sex gaps among Conservative supporters in four out of the six items. Again much of the variation in responses between the sexes results from more women placing themselves in the neutral category. In response to the statement 'Young people today don't have enough respect for traditional British values', the sex difference is in strength of feeling rather than direction - with 11 percent more men than women agreeing strongly with the statement, and 10 percent more women than men agreeing. Women more often placed themselves in the 'neither' category in response to two of the items, 'People should be allowed to organize public meetings to protest against the government' and 'Political parties that wish to overthrow democracy should be allowed to stand in general elections', with 13 percent and 15 percent more women than men selecting the neutral response, respectively. There is, however, a large sex difference in the censorship item: 20 percent more women than men agreed strongly or agreed with the statement 'Censorship of films and magazines is necessary to uphold moral values'. Women clearly hold strong opinions on censorship, 4 percent fewer women than men selected the 'neither' response in this case. Added together, the items do not suggest that women hold more authoritarian attitudes than men, with the exception of the sizeable difference in responses to the censorship question which are well documented elsewhere. When the items are added together to create a cumulative liberal/authoritarian scale there is not a significant sex difference in the mean position of Conservative supporters. This finding suggests that hypothesis 4 - the Conservative sex gap in left-right economic attitudes results from women's greater authoritarianism - is unlikely to have much purchase, as overall Conservative supporting women are not significantly more authoritarian in their attitudes than Conservative supporting men. Thus, as there is not a statistically significant sex difference in Conservative supporters' liberal/authoritarian, socialist/ laissez-faire or left/right positions, there seems to be only consistent evidence of a left/right sex gap in regard to attitudes toward taxation/expenditure. In orderto testthe strength of this sex gap further, we ran a series of OLS regressions with tax/ spend as the dependent variable. In Model 1 (Table 3), sex is the only dependent variable, and we can see that Conservative supporting women were on average 0.74 of a point further towards the increased taxes and spending pole of the scale than men and the relationship is highly statistically significant.17 In Model 2, employment in the private sector is added as a control variable. Men Conservative Party supporters are more often employed in the private sector than women, as we expected (78 percent men, 64 percent women), and the employment gap is larger among Conservative supporters than among Labour (70 percent men, 60 percent women) or Liberal Democrat (66 percent men, 59 percent women) supporters. Nevertheless, examination of Model 2 shows that there are no apparent effects of private sector employment on tax/spend attitudes, although the variable has a large number of missing responses - reducing sample size and statistical power, making significant effects more Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 Campbell and Childs 633 Table 4. Attitudes to public sector cuts Conservative identifiers by sex, BES 2010 CMS. Percentage supporting cuts to the item of government spending Spending item Men Women Freeze welfare benefits** 71% 62% Raise retirement age*** 59% 45% Reduce NHS 5%*** 21% 1 1% Freeze public sector pay*** 65% 57% Reduce child tax credit*** 56% 45% Reduce unemployment benefits 49% 51% Reduce student loan subsidies* 31% 26% Reduce winter fuel allowance* 23% 19% Scrap trident missile 32% 31% Reduce top public sector pay 77% 76% Abolish regional dev agencies*** 57% 51% Freeze overseas aid budget** 71% 67% ***Sex difference significant at the 0.001 level chi square test. **Sex difference significant at the 0.01 level chi square test. *Sex difference significant at the 0.05 level chi square test. N = I 198. difficult to pick up. It would seem, nonetheless, that Conservative supporting women do not more often favour increased taxation and spending than men simply because they are more often employed in the public sector - a finding which refutes hypothesis 5. Private sector employment is excluded from Model 3 in order to preserve sample size:18 added to the model are: interest in politics, age, caring for children under 5 and under 15, household income and holding a degree. The inclusion of these items did not remove the gender effect and thus we have evidence to refute hypotheses 2 and 3. Both age and household income did have a significant effect on responses to the tax/spend question. For every increase of one point on the household income scale there is a reduction in support for increasing taxes and public spending by 0.06 of a point, and for every increase of one year in age there is an increase in support for greater taxation and public spending of 0.02 of a point. However, neither women's greater longevity nor their greater representation in lower income households can account for the sex gap in Conservative supporters responses to the tax/spend measure: their inclusion reduces mean difference between men and women on the tax/spend item by 0.21 of a point from 0.74 to 0.53, but the effect remains and is significant at the 0.01 level. Furthermore, examination of the standardized coefficients demonstrates that age, household income and sex all have roughly similar impacts ontax/spend attitudes, with sex achieving the largest standardized coefficient of 0.15 followed by 0.14 for age and -0.13 for household income. Clearly, Conservative supporting women are more inclined to support increased taxation and spending than Conservative supporting men, and the effect does not seem to be an artefact of variations in the personal circumstances of men and women. It is clear from this analysis that the resilient sex gap in tax/spend attitudes identified by Childs and Webb among Conservative Party members is also evident among Conservative Party supporters. The interest in politics variable does not have a statistically significant effect on responses to the tax/spend measure and does not diminish the sex effect. Thus, we have further evidence (besides the fact that women did not disproportionately place themselves in the middle of the tax/spend scale) to refute hypothesis 6 that the left/right sex gap in Conservative supporters' economic attitudes is an artefact of lower levels of interest in politics.19 Conservative identifiers Alongside the standard face-to-face surveys the British Election Study team has developed an Internet continuous monitoring survey (CMS). The CMS, while not including a party supporter variable, does have a party identifier item enabling us to undertake analysis on Conservative identifiers. The survey does not contain the tax/spend measure, but does include a series of questions relating to public expenditure cuts (Table 4).20 Examination of Table 4 shows sizeable sex gaps in support of a number of potential cuts to public expenditure, with the biggest supporting raising the pension age (12 percent), reducing child tax credit (11%) and reducing the NHS by 5 percent (10 percent). While there were negligible sex differences in support for reducing unemployment benefit, scrapping Trident (the UK's nuclear missile), and in support of reducing top public sector pay, overall quite an array of measures designed to reduce public expenditure receive more support from Conservative identifying men than women. This suggests that there are fairly stark sex divisions among Conservative identifiers in respect of spending cuts. The responses to the items were added together to create a cumulative scale, with responses ranging from zero (supports none of the cuts to the public spending list) to 12 (supports all of the public spending cuts listed). The mean response for men Conservative identifiers was 6.10 and for women 5.41. The difference in the mean scores of men and women was significant at the 0.001 level. This sex gap of 0.69 of a point among Conservative supporters outstrips the 0.33 sex gap among Labour supporters, and is further evidence that the Conservative Party faces a particularly stark sex gap in left/right economic attitudes. We conducted an OLS regression on the cuts scale (Table 5) and added year of birth, having school age children, terminal educational age, household income and attention to politics as independent variables (Model 2). Note that there is no measure of public sector employment in the CMS and therefore we cannot test hypothesis 5 on attitudes to spending cuts. The sex effect remains even after the addition of the controls (although the size of the coefficient is reduced by about one-third - from 0.93 to 0.66). Younger Conservative identifying respondents and those Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 634 Party Politics 21 (4) Table 5. OLS regression on Conservative identifier's attitudes to cuts by sex, BES CMS. Independent variable Model 1 Model 2 B(SE) Beta B(SE) Beta Women 0.93*** (0.14) 0.19 0.66*** (0.14) 0.13 Year born -0.04*** (0.01) -0.23 Has school-age -0.01 (0.14) -0.02 children Age completed 0.08 (0.05) 0.05 education Household income -0.01 (0.02) 0.02 Attention to 0.24*** (0.03) 0.21 politics Adjusted R2 0.04 0.14 N 1 198 1 135 who paid less attention to politics were less supportive of the cuts listed than the other respondents. However, the sex gap in support for the cuts remains and cannot be accounted for by women's lower levels of interest/attention in politics nor their greater longevity.21 Thus, there is little evidence to support any of the hypotheses; the sex gap is surprisingly robust to the addition of statistical controls. Conclusions Childs and Webbs' observation that women Conservative Party members are to the left of men on economic issues extends beyond party members: to Conservative party supporters and identifiers. This is an important finding in and of itself. Hitherto, sex differences in voting behaviour in the UK have been of a smaller magnitude than those found elsewhere. Studies of voting behaviour and gender in the UK have failed to show a realignment of women - which would require a shift from the traditional to the modern gender gap. Here, though, on issues of tax and spend, we find clear and large differences among Conservative women and men. Establishing that within the UK's rightist party, women and men - at the level of members, identifiers and supporters - are divided over whether tax and spend has the potential, as Childs and Webb suggested, to become significant for the Conservative Party (and the Coalition) if and when tax and spend comes to dominate the political agenda. We suggest that this time is now. There are three main possible effects. First, the party cannot afford electorally to forget about women voters, particularly Conservative women supporters. Against the backdrop of critical media coverage of the Party and the government's policies' effects on women, Cameron should be minded that Conservative women supporters are simply less in favour of his cuts programme than men Conservatives. If he ignores this difference, it is likely that he will lose Conservative women's votes at the next election. Secondly, there are issues of intra-party management, with the likelihood that women and men party members - and the party's organizations 'for women' (if we presume they are in line with women Conservative supporters and party members) - are likely to conflict over the party's priorities in respect of tax and spend. Thirdly, our findings have implications for the political representation of right-wing women in terms of descriptive representation. Elsewhere, researchers have argued for the greater representation of women in politics on the basis of greater attitudinal congruency between elite women and women in society over gender issues, arguing that while the latter may not want women representatives, they nonetheless need them so that their interests are (or rather have a greater chance of being) better represented substantively (Campbell et al., 2010). This research strongly reinforces the earlier claim, and does so in our view at a time when the descriptive representation of sex appears to be increasingly questioned within the academy - calls for the representation of class are sometimes presented as a zero sum game, and seem to omit that women are members of the working class too (Childs and Cowley, 2011; Kenny, 2012).22 Yet, if the sex gap evident among Conservative Party members is also apparent among its supporters and identifiers, then the Party may well face a strong backlash along sex lines if it fails to address these issues. In turn, this raises questions of intra-party management, electoral competition and electoral strategy. To extend the phrase: the Conservative Pparty on the ground may not want women MPs, but it might well need them to ensure that the Party's platform represents Conservative women supporters' views. More conceptually speaking, we might question how a disproportionately male Conservative Party can substantively represent Conservative women when their attitudes (perceived interests) appear to diverge significantly. Acknowledgement Many thanks to Paul Webb for sharing data and to Participants at the 2012 ECPR Joint Session workshop on Conservative Women for their useful feedback. Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Notes 1. The British Election Study was conducted by Harold Clarke, David Sanders, Marianne Stewart and Paul Whiteley and funded by the ESRC. http://bes2009-10.org/. 2. It has been argued that without women voters there would have been continuous Labour governments in the post-45 period (see Harmen and Mattinson, 2000). 3. Even if his commitment to recognize marriage in the tax system smacked more of social conservatism. 4. http://www.g^ardian.co.uk/politics/2010/oct/22/cuts-women-sp-ending-review. 5. http://www.wbg.org.uk/RRB.htm. Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 Campbell and Childs 635 6. Economic and Social Data Service, Quarterly Labour Force Survey Household Dataset, April to June, 2010. 7. http://fawcettsociety.org.uk/index.asp?PageID=1268; http:// www.guardian.co.uk/lifeandstyle/2012/mar/30/women-paying-price-osborne-austerity. 8. Of course such a claim remains underpinned by an uncritical assumption that Conservative and Liberal Democrat women Ministers and Members of Parliament are ideologically predisposed to want to act for women in this direction (Celis andChilds, 2011). 9. http://www.spectator.co.uk/coffeehouse/7691068/camerons-pitch-to-women-voters.thtml; http://www.guardian.co.uk/co mmentisfree/2011/oct/02/david-cameron- women-voters?utm_ source=feedburner&utm_medium=feed&utm_campaign=Fe ed%3A+theguardian%2Fcommentisfree%2Frss+%28Comm ent+is+free%29; http://www.guardian.co.uk/politics/interac-tive/2011/sep/13/leaked-memo-women-coalition-government. Consideration of the Liberal Democrats and women's representation lies beyond the remit of this article (see Evans, 2011). 10. The failure of the party to publish its candidate selection data, as per the 2010 Speaker's Conference Recommendations, means that we cannot determine whether they have walked away from the first dimension either. 11. Chapter 8 also looks at voters of all partie s at the 2010 GE; in this article we look just at conservative identifiers and supporters. 12. The mean difference is significant at the 0.001 level ANOVA. The data are weighted by the post-election face-to-face weight 'postwtgť. 13. At first glance this sex difference seems pretty marginal. Yet, when we compare the mean responses of Labour and Conservative supporters (6.14 and 5.48), respectively, we can see that the 0.7 of a point sex difference is actually slightly larger than the 0.65 of a point ideological difference between supporters of the Labour and Conservative parties. In other words, while the sex difference within the Conservative Party is not large, it is bigger than the gap between Labour and Conservative supporters, when we might have expected the latter gap to be the larger as the ideological gap between the parties is usually much greater than the gap between the sexes (Campbell et al., 2010). 14. The self-placement left-right scale was coded from l=left to 10=right. 15. The Cronbach's alpha for the scale was 0.526. A principle components factor analysis with varimax rotation indicated that there were two factors with eigenvalues above 1. The first factor explained 41 percent of the variation in the items and the second 25 percent. Examination of the factor loadings for the rotated solution suggested that the two items, 'ordinary people get their fair share' and 'there is one law for the rich' contributed most to the first factor, and the items relating to trade unions and private enterprise contributed most to the second. The items were nonetheless used in one cumulative scale due to the relatively high Cronbach's alpha and the fact that the scale is well established in the academic literature. 16. The Cronbach's alpha for the scale is 0.346 and a principal components factor analysis with varimax rotation indicated that there were two factors with eigenvalues above 1. The first explained 30 percent of the variance in the items and the second 20 percent. Examination of the rotated factor solution suggests that the items related to tolerance of unconventional lifestyles and permitting protest made the greatest contribution to the second factor and the remaining items to the first. 17. The adjusted R2s are low. However, we are not seeking to maximize model fit as we are interested in the sex effect not providing a complete explanation of attitudes. 18. When Model 3 was run with the private sector employment variable included, none of the independent variables reached statistical significance; this is not surprising given the reduction in sample size. The variable which was closest to having a statistically significant effect on the dependent variable was the respondent's sex with a lvalue of 1.83. The coefficient for the sex variable was 0.45 and the standardized coefficient 0.132 (the largest of the standardized coefficients). 19. Model 3 was run again with socialist/laissez-faire and liberal/ authoritarian position included. These items are included in the mailback part of the survey and have a reduced sample size. The household income variable was removed from the re-run model, as this also has a low sample size and it was impossible to include this item and the scales in the same model. Neither scale had a statistically significant effect on the tax spend item when included in Model 3 and the coefficient for the sex variable was 0.64**. 20. The items were included in the BES CMS between June and September 2011. 21. When the regression was repeated for Labour identifiers the sex gap was statistically significant, but the coefficients were smaller (the unstandardized coefficient for women Labour identifiers was 0.35 in Model 2, roughly half the size of the effect among Conservative identifiers). Overall, then, there is a significant sex difference in support for public expenditure cuts among Conservative identifiers. 22. http://www.telegraph.co.uk/news/politics/conservative/917364 6/David-Cameron-needs-friends-in-the-North-heres-how-to-wi n-them.html. References Alvarez RM and McCaffery E (1999) Gender and Tax. POLMETH Working Papers. Available at: http://web.pol- meth.ufl.edu/papers.html. Andersen K (1975) Working women and political participation, 1952-1972. American Journal ojPolitical Science 19: 439-453. Andersen K (1999) The gender gap and experiences with the welfare state. P.S. 32: 17-19. Bale T (2010) The Conservative Party: From Thatcher to Cameron. Cambridge: Polity. Box-Steffensmeier J, DeBoef S and Lin T-M (1997) Microideol- ogy, macropartisanship and the gender gap. American Political Science Association. Burden B (2008) The social roots of the partisan gender gap. Public Opinion Quarterly 72: 55-75. Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016 636 Party Politics 21 (4) Burns N (2001) Gender: Public opinion and political action. In: Katznelson I and Milner H (eds) Political Science: The State of the Discipline. Washington, DC: APSA, 462^187. Campbell R (2006) Gender and the Vote in Britain. Colchester, Essex: ECPR Press. Campbell R (2011) The politics of diversity. In: Heffernan R, Cowley P and Hay C (eds) Developments in British Politics 9. London: Palgrave Macmillan. Campbell R (2012) What do we really know about women voters? Gender, elections and public opinion. Political Quarterly 83: 703-710. Campbell R and Childs S (2010) Wags', 'wives' and 'mothers'... But what about women politicians? In: Geddes A and Tonge J (eds) The UK Votes: The 2010 General Election. Oxford: Oxford University Press. Campbell R, Childs S and Lovenduski J (2010) Do women need women MPs. British Journal of Political Science 40. Campbell R and Winters K (2008) Understanding men and women's political interests: Evidence from a study of gendered political attitudes. Journal of Elections, Public Opinion and Parties 18: 53-74. Carroll S (1988) Women's autonomy and the gender gap: 1980 and 1982. In: Mueller C (ed.) The Politics of the Gender Gap. Los Angeles, CA: Sage. Carroll S (1999) The disempowerment of the gender gap: Soccer moms and the 1996 elections. P.S. 32: 7-11. Celis K and Childs S (2011) The substantive representation of women: What to do with conservative claims? Political Studies 60: 213-225. Chaney C, Alvarez MR andNagler J (1998) Explaining the gender gap in U.S. Presidential elections. Political Research Quarterly 51. Childs S and Cowley P (2011) The politics of local presence: Is there a case for descriptive representation? Political Studies 59: 1-19. Childs S and Webb P (2009) The feminisation of the Conservative Parliamentary Party: Party members' attitudes. Political Quarterly 80: 204-213. Childs S and Webb P (2012) Sex, Gender and the Conservative Party. London: Palgrave Macmillan. Conover P (1988) Feminists and the gender gap. Journal of Politics 50: 985-1010. Converse P (1964) The nature of belief systems in mass publics. In: Apter D (ed.) Ideology and Discontent. New York: Free Press, 206-261. Dolan K (2010) The impact of gender stereotyped evaluations on support for women candidates. Political Behavior 32: 69-88. Duke Whitaker L (2008) Voting the Gender Gap. Urbana and Chicago, IL: University of Illinois Press. Edlund L and Pande R (2002) Why have women become left-wing? The political gender gap and the decline in marraige. Quarterly Journal of Economics 111: 917-961. Evans E (2011) Gender and the Liberal Democrats: Representing Women? Manchester: Manchester University Press. Evans G, Heath A and Lalljee M (1996) Measuring left-right and libertarian-authoritarian values in the British electorate. British Journal of Political Science 47: 94-112. Fleishman J (1988) Attitude organisation in the general public: Evidence for a biodimensional structure. Social Forces 67: 159-183. Ford R (2008) Is racial prejudice declining in Britain? British Journal of Sociology 59: 609-634. Frazer E and MacDonald K (2003) Sex differences in political knowledge in Britain. Political Studies 51: 76-83. Gidengil E and Harell A (2007) Network diversity and vote choice: Women's social ties and left voting in Canada. Politics and Gender 3: 151-177. Greenberg A (2001) Race, religiosity and the women's vote. Women and Politics 22: 59-82. Harmen H and Mattinson D (2000) Winning for Women. Fabian Society. Hayes B (1997) Gender, feminism and electoral behaviour in Britain. Electoral Studies 16: 203-216. Heath A, Evans G and Martin J (1993) The measurement of core beliefs and values: The development of balanced socialist/lais-sez faire and libertarian/authoritarian scales. British Journal of Political Science 24: 115-158. Inglehart R and Norris P (2000) The developmental theory of the gender gap: Women and men's voting behaviour in global perspective. International Political Science Review 21: 441-462. Jennings KM (1992) Ideological thinking among mass publics and political elites. Public Opinion Quarterly 56: 419-441. Kenny M (2012) The political theory of recognition: The case of the 'white working class'. British Journal of Politics and International Relations 14: 19-38. Lehman K, Burns N, Verba S, et al. (1995) Gender and citizenship participation: Is there a different voice. American Journal of Political Science 39: 267-293. Mondak J and Anderson M (2004) The knowledge gap: A reexamination of gender based differences in political knowledge. Journal of Politics 66: 492-512. Mueller C (1988) The Politics of the Gender Gap. Sage Yearbooks in Women's Policy Studies. Beverly Hills, CA: Sage, 316. Nie N and Andersen K (1974) Mass belief systems revisited. Journal of Politics 36: 540-591. Norris P (1999) Gender: A gender-generation gap? In: Evans G and Norris P (eds) Critical Elections: British Parties and Voters in Long-Term Perspective. London: Sage, 146-163. Norris P (2001) The gender gap: Old challenges, new approaches. In: Carroll S (ed.) Women and American Politics: Agenda Setting for the 21st Century. Oxford: Oxford University Press. Ruddick S (1989) Maternal Thinking, Boston, MA: Beacon Press. Steel G (2003) Class and Gender in British General Elections. Paper prepared for the presentation at the Midwest Political Science Association Annual Meeting. Chicago. Tilley J (2005) Libertarian-authoritarian value change in Britain, 1974-2001. Political Studies 53: 442-453. Downloaded from ppq.sagepub.com al UNIV OF VIRGINIA on June 4, 2016 Campbell and Childs 637 Wangnerud L (2000) Testing the politics of presence: Women's representation in the Swedish Riksdag. Scandinavian Political Studies 23: 67-91. Webb P and Childs S (2012) Gender politics and conservatism: The view from the British Conservative Party grassroots. Government and Opposition 47: 21-48. Author biographies Rosie Campbell is a Senior Lecturer in Politics at Birkbeck College, University of London. Rosie has research interests in political behaviour, representation, political careers and women and politics. Her book Gender and the Vote in Britain was published in 2006 and she has recently published articles in the British Journal of Political Science, Political Studies, the British Journal of Politics and International Relations and British Politics. Sarah Childs is Professor of Politics and Gender at the University of Bristol, UK. She writes on women's political representation, gender and political parties, and the UK parliament. Her latest book, Sex, Gender and the Conservative Party was published in 2012, with Paul Webb. She has recently published in Politics and Gender and Political Studies. Downloaded Irom ppq.sagepub.com at UNIV OF VIRGINIA on June 4, 2016