INTRODUCTION
The recent push for more women parliamentarians around the world via positive action measures such as gender quotas naturally begs the question of whether the increased descriptive representation of women in parliament is making a substantive difference, in terms of the types of policies passed, behavioral norms, and procedures. How does the substantive representation of women take place? One way of measuring gendered impacts in politics is to look at the participation of Members of Parliament (MPs) in formal, parliamentary debate on key policy issues. It is reasonable to assume that one (though certainly not the only) measure of a good MP is his or her ability to contribute his or her voice and thoughts to public policy debates in parliament. Furthermore, if we consider women MPs representative of women in particular as a marginalized group, making their voices heard in parliamentary debates is perhaps one of the best ways of acting for women. As Melissa Williams (Reference Williams1998, 138) argues, the unique voice of women is a necessary component for women's political equality:
It is not enough that women communicate their perspectives to men; a full respect for women's agency, and a full protection of their interests, requires that they act on their own behalf. Only through their legislative presence can women be sure that the “woman's point of view” on each policy issue is formulated and expressed.
Women as a group share no exclusive set of political interests aside from the eradication of gender-based discrimination. Traditionally, gender has not been regarded as one of the main voter cleavages (e.g., social class, race, region, and religion) because men and women are thought to experience many of the same, cross-cutting forces (Lipset and Rokkan Reference Lipset, Rokkan, Lipset and Rokkan1967). The gender gap in Britain has recently shifted away from conservatism, with 2005 marking the first election when more women voted for Labour (38%) than Conservative (32%) (Cowling Reference Cowling2005). It is understandable, given the fact that discriminatory laws still exist and the structure of most parliaments is still based upon the universal disembodied male (Puwar Reference Puwar2004), that women parliamentarians might be expected to contribute a “woman's point of view” to certain issues that affect women in particular (e.g., breast cancer care in the National Health Service). It is not so understandable that women are often expected to participate in public discourse on whole categories of legislation (e.g. education, health care) and not to participate in others (e.g. foreign policy, finance). In a study of Norwegian MPs, Hege Skjeie found a consensus across party lines and sex that gender does and should make a difference in political interests. A majority of representatives saw women as particularly concerned with policies on welfare, the environment, equality, education, and disarmament, while men were said to be more interested in the economy, industry, energy, national security, and foreign affairs (Skjeie Reference Skjeie1991). If such assumptions were made by representatives in one of the world's most equal and socially progressive parliaments (of the time, and today), then there is good reason to think that they are widespread.
The British case offers some evidence that women MPs are thought to have a particular set of political issues and interests because they are women. Research on parliamentary candidates and MPs in the 1992 and 1997 elections found that, compared to men in each party, women were more supportive of feminist and left-wing values and expressed stronger concerns about social policy issues. However, in all cases the gender gap was small, and political party proved the strongest predictor of a representative's values and attitudes (Norris Reference Norris, Lovenduski and Norris1996, Reference Norris and Hay2000; Norris and Lovenduski Reference Norris and Lovenduski1995). In all likelihood, politicians are reflecting the preferences of their (or their party's) constituents. In an analysis of the British Election Studies (BES) and British Social Attitudes Surveys, Rosie Campbell (Reference Campbell2004) finds some evidence that there is a “women's interest” in the British electorate, albeit with important subgroup differences. Women are more likely to prioritize education and health care issues, and men are more likely to select the economy as their most important election issue (Campbell Reference Campbell2004: data from BES 2001).
Anecdotal evidence further contributes to the stereotype. When the 1997 British general election resulted in record numbers of women MPs, one male Conservative MP commented that he feared that the women “will start meddling in defence policy, increasing the aid budget and deploying peace-keeping troops everywhere” (cited in the Spectator, 24 May 2007, as quoted in Puwar Reference Puwar2004, 54). Such comments are startling because they occur in what many people consider to be gender-neutral institutions (e.g., electoral systems and governments). They are useful because they encourage one to think about how one might operationalize and analyze gender and political participation across various issue areas.
This study undertakes a quantitative analysis of second reading bill debates in the British House of Commons. The analysis considers debates on health care (a “soft” policy area considered “women's interest”) and finance (a “hard” policy area considered more masculine) over a two-year period of the current legislature (May 2005–August 2007). It addresses two questions. First, does sex makes a difference in the likelihood of an MP contributing his or her voice to the formal policy discussion of different issue areas? Second, if sex does make a difference, does this indicate that women are using debate as a forum to achieve greater substantive representation in areas of perceived women's interests?
The hypothesis tested is that women will be more likely to participate in health care debates than men because they feel a particular, gendered duty to represent women's perspectives on issues popularly construed as “women's issues.” In finance debates, the hypothesis is that women MPs will be less likely to participate than men, both because male MPs may feel a gendered duty to participate in issues often construed as masculine and because women are often perceived to have less expertise than men in this issue area, and thus may be sidelined.
The article proceeds as follows: First, a theoretical section provides an essential framework, situating gender and debate participation within the wider context of women's political representation. A review of the secondary literature on women's descriptive and substantive representation, along with a brief explanation of the British parliamentary system, gives the reader a greater understanding of why certain methodological choices were made. Some hypotheses are then put forward about the possible effects of sex as well as six other (control) variables on participation in health care and finance bill debates. Next, data and methods are described, including operationalization of the variables. Finally, the analysis and models themselves are set forth, concluding with a discussion of the findings and implications for future research.
WHY DOES GENDER MATTER?
The question of how gender affects debate participation is important because it is part of the broader discussion of women's representation and participation in politics. The dominant understanding of political representation in competitive party systems is that politics is about ideas (substance). In this model, voters make decisions about which candidate to support on the basis of policies and programs—not on personal characteristics such as sex, race, class, attractiveness, or family background. These individual characteristics are thought to be irrelevant (Phillips Reference Phillips1995). In practice, however, women are underrepresented in nearly every parliament in the world.
As of February 29, 2008, the percentage of women in single or lower houses of national parliaments worldwide stood at 18.0%, more than double the rate of 8.1% in 1965 (Inter-Parliamentary Union 2008). While this progress is significant, compared to economic opportunities, education, and legal rights, political representation is the area in which the gap between men and women has narrowed the least (Norris and Inglehart Reference Norris and Inglehart2003). Today, the United Kingdom ranks 59th in the world with 19.5% (126/646) women in national parliament, behind such countries as Rwanda, China, and Iraq. Within Europe, the UK is roughly on a par with the regional averages of 20.9% (Organization for Security and Co-operation in Europe member countries, including Nordic countries) and 19.0% (excluding Nordic countries), respectively (Inter-Parliamentary Union 2007).
The underrepresentation of women in parliaments worldwide indicates that political values and behaviors are, in fact, gendered. As an historically marginalized group, women traditionally have been left out of political life. Women achieved full suffrage in the UK in 1928; however, British political parties did not select women for winnable seats in significant numbers until the mid-1990s (Lovenduski Reference Lovenduski2005). While a detailed discussion of why women remain underrepresented today is beyond the scope of this article, entrenched patterns of social norms, combined with direct and indirect discrimination, will serve as an (apologetically) simplified explanation. Women today continue to bear the main caretaking responsibilities at home and in the workplace. So long as the sexual division of labor remains unbalanced, it is difficult to imagine women feeling able to manage a full-time career in politics and running for political office (Phillips Reference Phillips1999). Women face indirect discrimination because the conditions under which women-friendly policy changes, such as more nurseries or family-oriented working hours, are likely to occur are not those in which predominantly male MPs determine the political agenda. Furthermore, when women do buck the social norms and run for office, they may face subtle or direct discrimination, as has been the case, for example, in British selection committees (Shepherd-Robinson and Lovenduski Reference Shepherd-Robinson and Lovenduski2002).
Women's underrepresentation is related to central issues of how to ensure equal and fair representation for all citizens; namely, in what ways, and to what extent, does the fair representation of women depend upon their physical presence within legislative bodies? The concept of descriptive representation has emerged as a model for addressing the question of how to ensure political representation for disadvantaged groups. In descriptive representation, representatives are, in their own persons and lives, in some sense typical of a larger class of persons whom they represent (Griffiths and Wolheim Reference Griffiths and Wolheim1960). The idea, applied to gender, is that women representatives can represent women constituents better than men can. As Anne Phillips (Reference Phillips1995) postulates, a “politics of presence” is needed to supplement the lofty (and unrealized) ideal of the politics of ideas.
Women Acting for Women? The Complicated Relationship between Descriptive and Substantive Representation
Theorists disagree about the extent to which descriptive representation leads to substantive representation, if at all. Some proponents of descriptive representation build the case for greater representation for women without referencing substantive effects at all. These theorists claim that women should be equally represented in parliament either for justice reasons—because women compose 50% of most populations (e.g., Scott Reference Scott1997)—or for symbolic reasons—because the historical absence of women in political institutions may be associated with perceptions of women's second-class citizenship and the notion that politics is a “male domain.” Such theorists make no assertions about women acting differently or having a different set of policy preferences than do men. Others contend that beyond the justice or symbolic arguments for greater physical representation, women contribute something different to politics (e.g., Childs Reference Childs, Geddes and Tonge2005; Lovenduski Reference Lovenduski2005; Mansbridge Reference Mansbridge1999). This notion, that women MPs will act for women as a group, serves as the major theoretical underpinning of this article.
The notion of women's feminizing of politics is controversial because it runs the risk of essentializing women, implying that as a group, they inherently share a set of policy preferences or ideology. Women are not a monolithic entity with a collective set of interests and beliefs. One could cite many examples of female political figures who represent the antithesis of what many regard as “women's interests” (e.g., Margaret Thatcher). In order to make the claim that the fair representation of women requires their legislative presence, one must adduce something that members of women as a group share because of their historical marginalization, rather than their essential (biological) nature (Williams Reference Williams1998). Furthermore, it is not a guarantee that women representatives will act for women, only that they will be more likely to act for women than will men (Phillips Reference Phillips1995). Any carefully considered study of the connection between women's descriptive and substantive representation should take into account the inevitable problems that arise when one conceives of gender as a spectrum but sees practical value in categorizing it as a dichotomy. Keeping in mind the complexities and limitations of this method, this study uses sex as a proxy for gender in order to make some empirical claims about the gendered behavior of women and men MPs.
Finally, it is worth noting that the simple increase of women's numbers in parliament is not exclusively responsible for furthering women's interests. The substantive representation of women occurs on many interacting levels. Positional power, strategic alliances, coalitions, and relationships with women's movements, not to mention the numerous political roles outside parliament, are all crucial in making women-friendly policy change (Celis and Childs Reference Celis and Childs2008). Even within the universe of a politician's role, parliamentary activities (such as introducing legislation, debating, and voting) are only some of the ways in which women MPs can act for women. While not the only or even necessarily the best mechanisms to make substantive change, parliamentary activities are particularly high profile—and thus “high stakes.” As such, the floor of the parliament is perhaps the best forum for women literally to make their voices heard.
LITERATURE REVIEW: THE COMPLEX RELATIONSHIP BETWEEN DESCRIPTIVE AND SUBSTANTIVE REPRESENTATION
Traditionally, the assessment of claims that women act for women has focused upon the who (descriptive) and the what (substantive), rather than the how (Squires Reference Squires2005). Studies have relied upon the concept of “critical mass theory,” that once women achieve a certain proportion in parliament, political culture, behavior, and policy will be changed or feminized (Childs and Krook Reference Childs, Krook and Childs2008). Addressing the who, empirical studies have analyzed the impacts of greater numbers of women in parliament. Roll-call vote analysis is perhaps the most obvious empirical method of assessing sex differences in legislative behavior. Studies of roll-call voting in the United States have analyzed whether women were more liberal than their male colleagues, with mixed results (e.g., Burrell Reference Burrell1994; McCarty, Poole, and Rosenthal Reference McCarty, Poole and Rosenthal1997; Welch Reference Welch1985). A study of the voting behavior of Labour MPs during the 1997 Parliament found that newly elected women were less likely to rebel against the party whip as the rest of the parliamentary party, even when controlling for a range of other factors (Cowley and Childs Reference Cowley and Childs2003). A classic but increasingly criticized method, roll-call voting can only begin to explore how gender differences in legislative participation might affect substantive transformations. The indiscriminate use of critical mass theory cannot take into account how representatives identify with gender, the impact of minority members, the other activities that MPs engage in, and “critical acts of change” (Dahlerup 1988; Dovi Reference Dovi2003).
Addressing the what of substantive change for women, qualitative efforts to examine women's political and gendered motivations and intentions have found that women MPs see themselves as acting for women and empathize with them in particular. The record number of women ushered into the House of Commons after the landslide Labour victory of 1997, in particular, has provided a “rich laboratory” for scholars of gender and politics. Sarah Childs finds that many women MPs expressly seek to represent women, and that women MPs believe that they have raised women's concerns in Parliament, having a positive effect on policy addressing violence against women, child care, and equal opportunities (Childs Reference Childs1999, Reference Childs2001, Reference Childs2004). Many of Labour's new women MPs claim to have acted for women since their election (Childs and Withey Reference Childs and Withey2004). Fiona MacKay's Reference MacKay2001 work on Scottish councillors mirrors work at Westminster, finding that elected women believe that they have a responsibility to act for women by raising certain issues like child care, domestic violence, and equal opportunity issues in policy debates.
Connecting the who and the what of descriptive and substantive representation in meaningful ways involves addressing the how, or behavioral mediation. Recently, a call has been made for scholars to make a shift in their research question from “when women make a difference” to “how the substantive representation of women occurs” (Childs and Krook Reference Childs and Krook2006). Studies addressing this question narrow the focus of analysis to how sex and gender play a role within particular loci or mechanisms of change. For example, Alisa Henderson (Reference Henderson2005) examines the plenary behavior of MPs in the Scottish Parliament's first year of debates and finds that women participate differently from men; women participate much later than men in almost all aspects of plenary debate, are more likely to participate in Members' Business Debates, and less likely to prompt or engage interventions. In an examination of more than 300 plenary debates in the National Assembly for Wales, Paul Chaney (Reference Chaney2006) finds that women parliamentarians have a greater propensity than their male colleagues to initiate and engage in political debate on women's issues. In the British case, work on the signing of Early Day Motions (EDMs) in the 1997 Parliament shows that Labour women were more likely than Labour men to sign women's and feminist women's EDMs in particular (Childs and Withey Reference Childs and Withey2004). Karen Bird (Reference Bird2005) finds that women speak more often than men on issues of sex and gender.
This article agrees with the proposed need to focus new research on how change takes place and to acknowledge the complexities that multiple sites and actors bring to women's substantive representation. It takes debate participation in the House of Commons as its point of inquiry. The following brief explanation of the British parliamentary system gives the reader a general understanding of the model, as well as insights as to why debate participation was chosen as the focus of analysis over other parliamentary or political activities.
THE BRITISH PARLIAMENTARY SYSTEM
Various ways that British MPs might express their views and those of their constituents were taken into account in order to develop a logical methodology for operationalizing women's political participation across different issue areas. Firstly, the form of participation had to be public, due to the stated purpose of this study to focus on women's voice. Four forms of quantifiable public participation in Westminster politics were considered for the analysis: debate on the floor, debate in committee, written questions, and Early Day Motions (EDMs). Of these, debates on the floor seemed most fitting because they require direct, vocal, and public participation. Politicians will expect their participation in debates to have a higher degree of public scrutiny (Barnett Reference Barnett2002). By focusing upon participation in plenary sessions, this study assesses whether a representative's voice is being heard and registered in the most public proceedings of the British legislative process.
Having said this, it is important to note that excluding other forms of participation, such as committee meetings, leaves out many elements that might uncover different levels and styles of political contribution according to gender or other factors. Particularly because of the highly adversarial style of debate in the British system, women MPs may prefer to act for women via other forms of political participation. Focusing on gendered behavior in committee work or even work in the constituency could be a natural next step in this line of research, and findings could be compared and contrasted with plenary behavior.
A few key points about the process of debate in the House of Commons inform the research. Although debate, or formal discussion, in the House of Commons may include legislation, general topics of interest, or issues selected by the major parties, this article focuses only upon legislation. In any session, roughly 50% of parliamentary time will be spent scrutinizing proposed legislation (Barnett Reference Barnett2002). Debates on legislation have three key stages in the House of Commons: the first reading, when the bill is introduced as a formality; the second reading, which is generally seen as the substantive stage and the one in which the most debate is heard; and the third reading, when no further amendments may be made and debate is usually very short (“Parliamentary Stages of a Government Bill” 2007). This article analyzes debate in the second reading stage of legislation.
Debate in the House of Commons is famously adversarial. Unlike most legislative bodies, which sit in a semicircle, in the Commons the chamber is rectangular, pitting government benches against opposition benches—separated by a distance of two sword lengths (Barnett Reference Barnett2002). Members must be called upon to speak in debate by the Speaker of the House. MPs may write in advance to indicate their wish to speak in a certain debate, but this is no guarantee. In any debate, an MP must get the Speaker's attention (called “catching the Speaker's eye”), usually by half-standing from his or her seat.
The gendered aspects of debate in the House of Commons have been well documented. It has been said that men are more likely to gain and hold the floor in such formal arenas and to speak for longer periods of time than women, who tend to “leave the floor to men” (Holmes Reference Holmes1995, 193). Sir George Young MP, Conservative Leader of the House for much of the 1997 Parliament, claimed that the influx of new Labour women MPs after the 1997 elections did “make less noise in the chamber” (Young 2000, 8). Work on the gendered dynamics of political discourse has shown that men are more likely to interrupt other MPs via illegal interventions, whereas women tend to follow the Commons' legislative procedure quite stringently (Shaw Reference Shaw2000). Women MPs have also reported a range of constraints and sexist treatment in the House of Commons (see, e.g., Jackie Ashley, “Women MPs bullied and abused in Commons,” The Guardian, 7 December 2004 [Politics section]). While parliamentary debate is a crucial forum for women MPs to speak for women as a group, it is also important to remember that women's participation may be mitigated by the subtle or direct sexism of the institution.
EXPLANATORY VARIABLES AND HYPOTHESES
This study is primarily concerned with analyzing the effects of sex on policy debate participation in two different issue areas, finance and health care Sex is considered to be a proxy for gender, though it is acknowledged that a range would be more appropriate than a dichotomous variable for measuring the latter (Withey Reference Withey2003). In addition to sex, six additional explanatory variables are included in the analysis: seniority, relevant committee membership, political party, leadership position, attendance rate, and rebellion rate. Following is a brief discussion of why each explanatory variable is included and what its expected effects are.
Seniority has long been correlated with the likelihood of representatives' participation in formal debate. D. R. Matthews's classical 1960 study of the folkways of the U.S. Senate found an “amateur” effect in which new members are expected to be “seen and not heard.” Confirming this insight, more recent studies have shown that the amount of legislative activity, including participating in debates, significantly increases between the first and second terms of a U.S. representative's career (Hibbing Reference Hibbing1991), and that seniority has a positive effect on debate participation in the Honduran Congress, with the most frequent speakers having at least one term of experience (Taylor-Robinson and David Reference Taylor-Robinson and David2002). From these findings, it is reasonable to assume that more senior MPs participate more frequently in debate than those with fewer terms under their belts.
Relevant committee membership has also been found to be a significant and positive predictor of debate participation. Matthews (Reference Matthews1960) found that relevant committee members, particularly the chairmen of the committees, were (informally) expected to lead floor debate on related bills. In his 1996 study of the U.S. Congress, R. L. Hall found that the debate of a bill tends to be dominated by the relevant committee members. In the British context, it is logical to assume that the expertise accrued by membership of a select committee (or perhaps the previous expertise indicated by selection for a committee) might make an MP with relevant committee membership more likely to participate in debate than others.
Positions of leadership in the governing or opposition party often require speaking as part of the formal role, for example, clarifying the official position or responding to questions about what action the executive is taking on a certain issue. Empirically, some work has shown that leadership position positively affects the likelihood of a representative participating in debate. Michelle Taylor-Robinson and Sky David (2002) find that holding a leadership position modestly increases the likelihood of participating in debate. Thus, one might assume that leaders of the government and the Official Opposition might be more likely to participate in debates than would backbenchers.
In British politics, the focus for the party in government (in the period studied, Labour) is to implement the party's platform of national-level policies (Kavanagh Reference Kavanagh1987). The role of the party of the Official Opposition (in the period studied, Conservative, though the Liberal Democrats are also included in the analysis as the third largest party) is (at least formally) not to prevent these policies from being passed, because to do so would be to ignore the mandate given by the people to the governing party, but to question these policies as thoroughly as possible. From the point of view of the opposition, it is particularly important for backbencher Members to participate in debate, lending support to the opposition's claims about the government (Barnett Reference Barnett2002). Thus, Conservative MPs may be more likely to participate in debate than Labour or other party members.
An MP's attendance rate in the House of Commons is perhaps a good proxy for the value that the MP places on legislative debate, as opposed to other forms of legislative participation, such as constituency or committee work. Moreover, an MP who is present much of the time has more opportunities to participate in debate than an MP who is not present as often. Members with high attendance rates may participate in debates more frequently than those with low attendance rates.
Rebellion rates are another potentially significant predictor of MPs' debate participation. Because party discipline is quite strict in the UK, an MP might feel the need to participate in debate to justify his or her opposition to the party line if he or she is planning to vote against the party. This sense might be particularly pronounced for women who do not wish to upset the party because of their electoral situations. Many of the women in this analysis were first elected in 1997, partly due to the Labour use of all-women's short lists. Interviews with Labour women MPs in 2000 found that several of the women themselves felt this pressure, with one commenting that “most of us come from key seats, they [men likely to rebel] are marginal seats, and to be frank we are not stupid“(Cowley and Childs Reference Cowley and Childs2003, from interviews on June 13 and May 10, 2000). One might expect Members with high rebellion rates to participate more often in debate than those with low rebellion rates.
Finally, the critical variable for this study, an MP's sex may affect his or her likelihood to participate in debate, or in certain types of debate. Though the study makes no normative claims, the stereotype is that women are more likely to be interested in “soft” policy issues, such as health care, education, and family affairs, than “hard” policy issues like finance or foreign policy (Lovenduski Reference Lovenduski2005). Because of the reasons previously discussed, women MPs may have a greater interest in speaking in policy debates on issues that they view—or understand their constituents to view—as disproportionately affecting women. Women MPs may see health care as a women's issue because of the traditional caretaking role of women, because nearly all women go to the hospital to have children, and because women take their children and others they care for to the hospital. Although financial affairs certainly affect women as well, the masculine association may be due to the historical, structural division of labor. Capital markets traditionally have been and are still predominantly controlled by men. In addition, financial matters have always served as one of the core policy areas of governance, whereas issues like education and health care have taken up a bigger part of the legislative agenda relatively more recently, following the rise of the European social model after World War II. Financial matters may be perceived to be more masculine because of the long-term, strong association with traditional (male) governance.
DATA AND METHODS
The data for this study were collected from the parliamentary sessions of the 2005–2007 British House of Commons. Data collection involved reading all transcripts of bills related to health care and finance within this time period and taking note of Members who participated, in speech or (spoken) intervention, and how many times that Member participated. The source of these transcripts was Hansard (the Official Report See http://www.publications.parliament.uk/pa/pahansard.htm), the edited verbatim report of proceedings in both Houses. There are nine second reading debates related to health care and eight second reading debates related to finance in this period. See Table 1, “Debate Participation in Health Care and Finance-Related Bills in the May 2005–August 2007 UK House of Commons,” for a summary of MPs' participation in all health care and finance-related second -reading debates during this period.
Table 1. Debate participation in health care and finance-related bills in the May 2005–August 2007 UK House of Commons
aNumber of speeches made is not equal to number of MPs who participated in debate, as many MPs speak more than once. The number of MPs (634) does not include those with 0% attendance rate in the House of Commons (Gerry Adams, Pat Doherty, Michelle Gildernew, Gordon Marsden, Shona McIsaac, Conor Murphy), the deputies (Sir Alan Haselhurst, Sir Michael Lord), or the Speaker of the House (Sylvia Heal).
bThree health-related bills included here had begun but were not yet completed at the second reading debate.
Source: Hansard (the Official Report) 2005–2007, “Volume Indexes to the House of Commons Parliamentary Debates,” parliamentary copyright. Available at http://www.publications.parliament.uk/pa/cm/cmbvindx.htm.
The dependent variables in the two models, health care debate participation and finance debate participation, are operationalized as the number of interventions (regardless of length) made by each Member during health care debates and, separately, during finance debates. The Speaker of the House and deputies were taken out of the analysis as their participation in debate is of a different nature. Additionally, Members whose attendance rates in the House of Commons were 0% were taken out of the analysis (i.e., Sinn Fein Members). Interruptions were included in participation if the Member was recorded as speaking; if the Member was noted in Hansard as standing, gesturing, or otherwise non-audibly participating in debate, this was not included in the data.
The independent variables of “Female,” “Leadership Position,” “Health Committee,” and “Finance Committees” are binary (where female = 1, leadership position (Cabinet or Shadow Cabinet) = 1, and membership on a relevant select committee = 1). Two binary variables, “Conservative” and “Liberal Democrat,” in which Members from the named party = 1, measure the effect of the main political parties in opposition. “Seniority” is the number of previous terms that each Member served in the House of Commons, ranging from 0 for newly elected MPs to 12 (Sir Peter Tapsell, Conservative). “Attendance” and “Rebellion” are both operationalized as percentages, ranging from 0 to 100, though in the analysis these were translated from percentages into rates. The source for these data was Dod's Parliamentary Communications' “Selective Biographies for Members of Parliament” (2007), except for attendance and rebellion rates. The source for data on attendance and rebellion rates was The Public Whip (2007). See Table 2, “Summary Statistics for Debate Participation Regression Models,” for a complete list of dependent and independent variables used in the analysis.
Table 2. Summary statistics for debate participation regression models, 2005–2007 UK House of Commons
Two separate models are used in the analysis, completely equivalent except that Model 1 tests participation in health care-related debates and includes the independent variable “Health Committee,” and Model 2 tests participation in finance-related debates and includes the independent variable “Finance Committees.” The statistical models are given as:
- Model 1:
Health Debate Participation =
+ 1 Seniority + 2 Female + 3 Health Committee + 4 Leadership Position + 5 Conservative + 6 Liberal Democrat + 7 Attendance Rate + 8 Rebellion Rate + - Model 2:
Finance Debate Participation =
+ 1 Seniority + 2 Female + 3 Finance Committees + 4 Leadership Position + 5 Conservative + 6 Liberal Democrat + 7 Attendance Rate + 8 Rebellion Rate +
Because the dependent variable is a count, the number of times a given Member participated in debate within the categories of either health care or finance, an ordinary least squares (OLS) regression model would not be appropriate to fit the data (particularly as the data are discrete rather than continuous). The negative binomial model was chosen as the best method for analyzing the debate participation models of the British House of Commons. STATA v.9 was used for all analyses.
ANALYSIS AND FINDINGS
Please refer to Table 3, Negative Binomial Regression of Debate Participation Models, for the results of the fitted regression models. The table gives coefficients and standard errors for Model 1, where the response variable was participation in health care debates, and Model 2, where the response variable was participation in finance debates. Incidence rate ratios, that is, exp(β1) rather than β1, are also given with similarly transformed standard errors. Rate ratios describe the multiplicative change in the expected count when the explanatory variable increases by one unit, while all the other explanatory variables remain unchanged.
Table 3. Negative Binomial Regression of debate participation models, 2005–2007 UK House of Commons
Notes: Dependent variable: Debate participation = number of times MP spoke in health care (Model 1) or finance-related (Model 2) debates in the 2005–2007 UK House of Commons. The sign following each variable is the predicted sign of the coefficient. Coefficients and incidence rate ratios are followed by standard errors, in parentheses.
* p < 0.10, **p < 0.05, ***p < 0.01.
The results of the two regression models were not identical, indicating that separate patterns characterize health care and finance debates. In health care debates, the analysis suggests that the most significant predictors of debate participation are seniority (–), Conservative party (+), attendance rate (+), and female sex (+). Leadership position also had a borderline significant effect (+), as did Liberal Democrat party (+), all other variables held at constant. In finance debates, findings show that the most significant predictors of debate participation are relevant committee membership (+), Conservative party (+), and seniority (–), holding other explanatory variables at constant. Sex was not significant, holding all other explanatory variables at constant. The following detailed analysis of results analyzes the effects of sex and the other explanatory variables on health care and finance debates.
Findings for the key variable for this study, sex, show that female sex is significant and a positive predictor of participation in health care debates, but not significant at standard levels in finance debates. Being a female increases the likelihood of participation in health care debates by a factor of 2.224, or 122%, as compared to men, holding all other explanatory variables at constant. It is interesting to note that being a female does not affect likelihood of participation in finance debates, despite the descriptive statistics that show disproportionate male interventions (82.5% of all interventions in finance debates are made by men, while men compose a slightly lesser 80.5% of all MPs; see Tables 1 and 2) when all other variables are held at constant. A chi-square analysis of gender and finance debate participation finds that, in fact, 23% of women MPs participated in finance debates, compared to 20% of men (not significant at standard levels). The analysis suggests that men may participate more prolifically than women, with those men who speak making a greater number of interventions on the average than those women who speak, though women are just as likely to participate. This would coincide with some women MPs' criticisms of male MPs' debating style, specifically claims that men have a tendency toward repetition: “a lot of men, you know, like the sound of their own voice” (Childs Reference Childs2004, 6).
Two variables, Conservative Party membership and seniority, were consistently significant across both models. Both variables carried similar coefficients across the two models, indicating similar effects on the two response variables. Confirming initial hypotheses, Conservative Party members were more likely to speak in debate than other MPs, holding other explanatory variables at constant. Being a Conservative MP increases the likelihood of participation in both health care and finance debates by a factor of about 1.85, or 85%, compared to MPs from other parties. This provides some evidence in support of the intuition that opposition MPs use parliamentary debate as an opportunity to question and state their party's disagreements with government policy proposals. In contrast to original hypotheses, seniority, or number of terms in office, had the effect of decreasing the likelihood of an MP participating in debate, holding other variables at constant. If an MP were to increase his or her number of terms in office by one unit, his or her participation in both health care and finance debates would be expected to decrease by a factor of 0.76, or 24%, while holding the other variables constant in the model. This finding suggests that the traditional theory that Members with greater experience and seniority are more likely to speak in debate merits additional research.
Many of the remaining initial hypotheses about the qualities that make an MP more likely to speak in debate are confirmed in the health care model—namely, leadership position (+), Liberal Democrat party (+), and attendance rate (+). MPs with leadership positions were more likely to participate in health care debates. Members of the Cabinet or Shadow Cabinet are expected to have a rate ratio of 2.218 times greater, or 120%, for health care debate participation compared to other MPs (significant at 0.10 level), while holding the other explanatory variables at constant. Members of the Liberal Democratic party are also more likely to speak in health care debates than other MPs. Being a Liberal Democrat MP increases the likelihood of speaking in health care debates by a factor of 2.384, or 138%, compared to other party MPs (significant at 0.10 level), holding all other explanatory variables at constant. Finally, Members with greater attendance rates are more likely to speak in health care debates. Increasing attendance rate by 1 point has the effect of increasing an MP's expected health care debate participation by a factor of 1.025, or 2.5%, holding other explanatory variables at constant. The effect may seem small, but if attendance rate is increased by 10 points, an MP's expected participation increases by 25%, holding other variables at constant. Relevant committee membership and rebellion rate were not significant at standard levels for the health care model.
In contrast with the health care model, findings for the finance model suggest a more constrained debate in terms of the types of MPs likely to participate. Leadership position, Liberal Democrat party, attendance rate, and rebellion rate are all not significant at standard levels for finance debates. Relevant committee membership, however, is significant and positive. All other variables held at constant, finance committee membership increases the likelihood of participating in finance debates by a factor of 2.714, or 170%, as compared to those without relevant committee membership. Perhaps committee membership affects finance but not health care debate participation because there are more finance-related committees (n = 4) than health committees (n = 1). The allocation of greater political resources to finance (number of select committees) and the tendency for committee members to participate more frequently in finance debates suggests that finance may be seen as a subject in which only experts should participate, whereas health, as a soft policy issue, may not be treated the same way.
The dissimilar results of the two regression models indicate that separate patterns of participation characterize health care and finance debates. Different types of MPs are likely to speak in health care debates and finance debates. A more diverse group of MPs is likely to speak in health care debates compared to finance debates. This may be because finance is perceived as a subject in which only experts should contribute their views (i.e. select committee members). Gender is likely a factor that interacts many of these variables, and it affects an MP's perceptions about whether his or her voice is appropriate and relevant to the debate. The findings of many of the explanatory variables are interesting in and of themselves and merit additional discussion, though a full analysis of what affects debate participation in general lies outside the scope of this study.
DISCUSSION AND IMPLICATIONS
This study has attempted to contribute some empirical evidence for examining the notion of “women acting for women.” The results of analyzing representatives' participation in health care and finance second-reading bill debates in the British House of Commons 2005–2007 give additional, empirical proof to the growing consensus that women MPs are making a difference for women's substantive representation. Sex makes a difference in the likelihood of an MP contributing his or her voice to the formal policy discussion of different issue areas. Women MPs participate disproportionately in debates on issues of health care, suggesting that women MPs are using debate as a forum to achieve greater substantive representation in areas of perceived “women's interest.” Overall, what do these findings say about the relationship between descriptive and substantive representation?
Beyond the numbers showing that women participate more frequently than men in health care debates, some preliminary qualitative analyses suggest that women are voicing concerns for women in particular. Taking a closer look at the nine bills used in the health care analysis, one can see that many of them address issues that concern women specifically and disproportionately (see Table 1). In the debate on the Safeguarding Vulnerable Groups Bill (June 19, 2006), for example, women made 72 of the 116 interventions. “Vulnerable groups” is highly gendered, both as a term often used as a proxy for “women and children” (or even phrased, “Women, children, and other vulnerable groups,” e.g., in NATO and the UN), and as a stark reality: Women face greater threats of economic insecurity, sexual exploitation, and domestic and other forms of violence than do men. Looking at the discussion in greater detail, women from both sides of the floor highlighted the background of the bill, the 2002 murders of the two young girls, Holly Wells and Jessica Chapman. The outcry over the failure to stop Ian Huntley from working at the Soham school where he went on to murder the girls led directly to the bill. The bill is now an act, which in 2007 introduced a new vetting/barring scheme to prevent unsuitable people from working with children and vulnerable adults, among other provisions. Women MPs' strong participation in the debate of this bill may have had an impact on its passing into law and the subsequent protection of more women from harm.
Women also contributed disproportionately to the debate on the Mental Health Bill (April 16, 2007), making some 53 of the 119 interventions. There are gender differences with respect to men's and women's experience of mental health in England: 20% of women are reported as having some form of mental illness, compared to 14% of men (Economic and Social Research Council 2007).Footnote 1 In her introduction of the bill, Minister for Health Patricia Hewitt spoke about caring for her sister, who suffers from mental ill health. Although women MPs' disproportionate participation rates in the debate indicate that mental health is an important issue for women MPs and women in general, the contentious debate of the bill's provisions shows that agreement does not extend to the way in which the issue ought to be addressed. This confirms previous findings that women representatives share a broad concern for certain women's issues but that this is intertwined with party identity, as one female MP asserted: “It's when you come to the next stage about proposed policy and solutions then [you] get the divergence” (Childs Reference Childs2006, 13; see also Norris and Lovenduski Reference Norris and Lovenduski1995). The divergence in this case was about whether the strict measures set out in the Mental Health Bill would protect the vulnerable people whom a severely mentally ill person could potentially harm, including themselves, or whether the measures were too severe, even draconian.
This article confirms the findings of many qualitative analyses, that women MPs see themselves as representing women in particular (but not exclusively). The evidence suggests that the presence of women in representative politics is likely to push policy debate toward issues that women give greater importance. The findings add to the growing consensus across the literature that women representatives often act differently from men, if not exclusively “for” women. This study might be enhanced by analyzing sex and debate participation in additional policy areas, such as education and skills, foreign affairs, and defense. It would be interesting to see if standard perceptions of the gendered nature of each issue area hold up in analysis (e.g., is education a “women's issue?”). The study would also be improved by undertaking the analysis over a greater period of time, for example, over an entire legislature, or several.
