The aim of this article is to provide new insights into theorizing on “critical mass” and “critical acts,” a widely studied topic within feminist political science. From a more comprehensive viewpoint, these concepts are linked to the descriptive and substantive representation of women (hereafter referred to, in short, as DRW and SRW).Footnote 1 The aim is to apply and test a key element—the recruitment hypothesis—with empirical data from Finnish parliamentary standing-committee hearings. The project simultaneously serves to pinpoint theoretical and methodological problems in the way we conceive and study DRW and its potential effects regarding SRW. The overall implications of the study represent a profound challenge to theorists and empiricists alike: Perhaps we should reconsider some of our beliefs concerning the accumulation of knowledge in this field from the perspective of “strong objectivity”Footnote 2 and start reviewing fundamental assumptions that underpin our concepts of “critical” mass and acts.
Considerable attention has been paid over the years to the question of whether an increase in women's presence (DRW) has any impact on their substantive representation, and if so, what this impact is. Empirical research, in particular, has focused on theories of critical mass or critical acts (Dahlerup Reference Dahlerup1988). According to the former approach, an increase of women in politics begins to affect the content and quality of decision making only after their proportion reaches a certain “critical mass.” The threshold is often set at 30% (Dahlerup Reference Dahlerup1988), although other figures have been proposed, too, from 15% to 40% (Kanter Reference Kanter1977a). The fundamental, underlying assumption is that an increase in women's presence in politics affects the contents, style, and mechanics of politics in some manner and that women's presence changes politics as compared to earlier, more male-dominated contexts or to bodies with fewer women. The critical-acts approach emphasizes that it is not numbers but, rather, whether and in what manner female politicians act for the interests of women's constituencies that counts.
This study investigates the effects of women's presence by posing two questions. First, does the proportion of female actors in a parliamentary standing committee increase its selection of female experts? Second, does a higher proportion of female actors in a parliamentary standing committee increase the frequency of consultation with representatives of women's collective-group interests (here defined as women's nongovernmental organizations (NGOs), governmental gender-equality agencies, and gender scholars)? I design a dual test for exploring the effects of women's presence on the two dimensions of representation (DRW and SRW). First, I investigate whether there are any effects of DRW for the increase of women in the expert pool. This is basically a straightforward application of the “recruitment hypothesis”—the idea that an increased proportion of women in an organization will affect its recruitment in a positive manner and result in an increase of women employees. Second, I perform a stricter test and assess whether the presence of women affects the procedural inclusionFootnote 3 of SRW in the expert pool—that is, whether it secures institutional access for gendered concerns and interests to be taken into account. The data consist of information gathered from all expert hearings in the Finnish parliamentary standing committees in 2005.
The results indicate that women's presence exerts different impacts on DRW and SRW. More women members of Parliament (MPs) in a standing committee may have a positive effect on the selection of more female experts, but the finding is inconclusive. By contrast, the increase of female MPs does not increase, but rather decreases, the selection of representatives of women's collective-group interests in the expert hearings. A much more significant result, perhaps, is the finding concerning conceptual confusions surrounding the relationship and boundaries between critical mass and critical acts. In feminist scholarship, recruitment of more women tends to be defined as a “critical act” by default and, thus, an instance of SRW per se. I argue that the very idea that increasing DRW (as referred to by these critical acts approaches) is/becomes SRW is deeply paradoxical and fundamentally faulty.
“CRITICAL” CONCEPTS IN THEORIES OF WOMEN'S REPRESENTATION: FRAMEWORK OF ANALYSIS
DRW, Critical Mass, and the Place of the Recruitment Hypothesis
The concept of critical mass was introduced to political science by Dahlerup (Reference Dahlerup1988). Taking into account the later history of critical mass, it is quite ironic that Dahlerup herself was very skeptical about the concept and the whole idea of the “automatic” political change it seemed to imply. Indeed, she suggested replacing the idea of critical mass with the concept of critical acts, which she regarded as a better tool for analyzing processes of social change.
Dahlerup points out that the idea of a critical mass originated in the discourse of feminist activists in the 1980s (Dahlerup Reference Dahlerup1988, 276). As the proportion of women in politics began to grow incrementally in Western industrialized countries, the concept was deployed—as is still the case today—to explain the slowness of the anticipated change and delays in it. The main argument was that it is not enough to have one or a few women in representative bodies because they will suffer from the disadvantages caused by their position as a small minority and will be able to affect neither the contents of politics nor political culture as a whole (cf. Kanter Reference Kanter1977a; Reference Kanter1977b). Rather, what is needed is a sufficient proportion of women—a critical mass. It is only after this has been achieved that the anticipated positive outcomes will become visible.
Notably, both Dahlerup and feminist activists who first adopted the concept of critical mass drew from an influential study by Kanter (Reference Kanter1977a; Reference Kanter1977b) that analyzes the negative effects emanating from women's minority position in organizations. Kanter divides them into four categories on the basis of their gender composition: uniform (gender proportions 100–0), skewed (minority gender under 15%), tilted (minority gender about 15%–35%), and balanced (women and men compose 40%–60% of the staff). Kanter's empirical analysis (1977a, 206–42, 248–49) targeted the negative impacts of skewed gender proportions for the members of a minority group and organizational culture more generally. She did not, however, regard gender as the foundational cause. Rather, the effects of a minority position were assumed to be similar, regardless of whether the minority was composed of women, men, blacks, whites, handicapped, or nonhandicapped people in a context where organizational power lies with another group (Childs and Krook Reference Childs and Krook2008, 729). In addition, Kanter's empirical data did not include an analysis of actual changes resulting from an increase in the size of a minority composed of women, nor an analysis of what will occur when the organization reaches gender balance. Instead, Kanter inferred a series of presumptive hypotheses concerning the potential consequences from a growing proportion of a minority group. They include an increase in mutual alliances among the members of a minority group, as well as impacts on organizational culture and decreased stereotyping, which lead to greater individuation of the members of a minority (Kanter 1977a, 248–49). In addition, elsewhere Kanter (Reference Kanter1977b, 966) made reference to the idea that a relatively large proportion of people “of one's own sort” also contributes to one's career opportunities and an increased credibility in an organizational context.
Kanter's theories have been applied diligently in organization research, though with somewhat inconclusive results, partly due to the fact that scholars have tended to operationalize the impacts of the size of a minority in very different ways. For the purpose of this study, the so-called recruitment hypothesis based on Kanter's theory is of major interest. It assumes that an increase in the size of a minority group affects organizational recruitment—that is, it increases the likelihood of the recruitment of employees from that minority. An empirical finding has been that an increase of members of a minority group in the leadership of a company affects the recruitment of employees of this minority group positively (Chambliss and Uggen Reference Chambliss and Uggen2000; Chused Reference Chused1988; Hammond Reference Hammond1990). For instance, in their study of U.S. law firms, Chambliss and Uggen (Reference Chambliss and Uggen2000) observed that the more women and African-American partners in these companies, the more they hired female and African-American employees. But this effect does not endure when we move from leadership to the personnel structure of an organization as a whole.
Dahlerup's (1988) explicit aim was to apply Kanter's theory to women's position in politics. She highlighted six areas where the increase in the relative proportion of women might be presumed to transform politics. But she concluded that it would be difficult to separate its impacts from societal change more generally, as well as from other possible factors. Moreover, according to Dahlerup, it was not possible to establish a numerical threshold value after which the change would occur. Consequently, she recommended that the idea of the critical mass be abandoned.
Dahlerup's advice notwithstanding, U.S. scholars of gender and politics became deeply engaged with “theories” of critical mass in the 1990s–2000s. Many projects tested them empirically with quantitative methods, studying, for example, the impact of the proportion of female legislators on “women's issues” (Bratton Reference Bratton2005; Bratton and Ray Reference Bratton and Ray2002; Saint-Germain Reference Saint-Germain1989; Thomas Reference Thomas1991, Reference Thomas1994); on the passage of bills concerning women's issues (Saint-Germain Reference Saint-Germain1989); and on women's policy priorities (Wängnerud Reference Wängnerud2000) and women's alliances. Some studies found support for the critical-mass hypothesis while others did not. The negative results show that an increase of women does not necessarily affect the content of politics in the manner one expects (Childs Reference Childs2004; Grey Reference Grey2002; Towns Reference Towns2003). Rather, it may increase men's resistance (Kathlene Reference Kathlene, Duerst-Lahti and Kelly1995; Towns Reference Towns2003; Yoder Reference Yoder1991), thereby posing a further barrier to change. Some studies indicate that women in a small minority in an organization may, in fact, be more effective advocates for women's concerns than those composing a larger minority group (e.g., Carroll Reference Carroll2001).
Nevertheless, political scientists applying critical-mass theories have seldom paid explicit attention to the recruitment hypothesis. Davis (Reference Davis1997) shows that the more women there are in parliaments, the more often they are selected to cabinets. In their study of 20 industrialized democracies, Studlar and McAllister (Reference Studlar and McAllister2002) investigated whether a critical mass of women increases the number of women MPs and whether it accelerates their recruitment to these positions. Their results showed that the electoral system and cultural attitudes toward gender equality explained the change better and that the effect of the existence of a critical mass of women was very small.
Some critical remarks arise when these political analyses are compared with the results of organization studies with regard to the recruitment hypothesis. Taking the latter into account, for example, Studlar and McAllister's research design appears somewhat far-fetched, inasmuch as in their role as parliamentarians (versus their role as party activists), women MPs hardly play a direct role in the recruitment of more women into politics. Moreover, as both Kanter and Dahlerup observed, attitudes toward female politicians and political culture may serve as intervening variables (see also Childs Reference Childs, Sawer, Tremblay and Tremble2006, 154; Opello Reference Opello2008). On the basis of these observations, therefore, we should test the recruitment hypothesis in empirical cases where women (either as politicians or in other influential positions) have a direct and explicit role in “recruiting” or selecting more women to various positions. The selection of experts to be heard by parliamentary standing committees provides such a case.
An Assessment of the Conceptual Foundations of the Recruitment Hypothesis
Another problem highlighted in this paper deals with the conceptualization of the main ingredient of the recruitment hypothesis: the increase of women (as employees, politicians, committee experts, etc.) as the outcome and effect of the presence of women in decision-making positions. According to Dahlerup's original formulation, a critical act is “one that will change the position of the minority considerably and lead to further changes” (Dahlerup Reference Dahlerup1988, 296). Significant criteria for it include “the willingness and ability of the minority to mobilise the resources of the organization or institution to improve the situation for themselves and the whole minority group” (p. 296). Dahlerup gave three examples of possible critical acts: recruitment (women politicians recruiting other women into politics), gender quotas, and gender-equality reforms and the establishment of permanent organs for gender equality (pp. 296–97).
Notably, Dahlerup and Kanter differ in how they view recruitment. For Kanter and other organization researchers, change in recruitment was a potential consequence caused by the increase in size of a minority group (i.e., “critical mass”). It was linked to transformations in organizational culture and in the position of dominant in-groups, possibly caused by intervening factors. For Dahlerup, on the other hand, women recruiting women into politics is a critical act. The problem here is that for many of those who conceptually draw from Dahlerup's definition, recruitment becomes a critical act par excellence. For instance, Childs and Krook (Reference Childs and Krook2008, n. 4, p. 735) criticize Studlar and McAllister's political application of the recruitment hypothesis. According to them, the scholars misnamed their object of study as “critical mass” when it was really about “critical acts.”
Compared with the attention paid to critical mass, the academic silence concerning critical acts—and the controversy indicated here—has been noteworthy. Although the concept of critical acts continues to be applied as an alternative to critical mass, as well as a solution to the problems caused by it, it has not been subjected to serious interrogation. In most instances, scholars tend to resort to Dahlerup's original formulations and examples, which, however, already include a series of possibilities for questioning them. They include, for example, the following problems: Where does the boundary between critical and noncritical acts lie? Is it to be found in the inability or unwillingness of the actors to mobilize the resources of the institutions and bring about change in favor of women? Is it that this boundary can only be discovered ex post facto when the impacts of actions and possible change (“the success of critical acts”) later on become visible? Is the presence of “wrong” motivations or “wrong” groups of actors sufficient for distinguishing critical acts from noncritical acts? This series of questions suggests that the concept of critical acts is inexorably linked to concepts such as “feminist consciousness,” “feminist agency,” and “feminist politics.”Footnote 4
Innovative attempts to theorize the boundaries between critical acts and SRW would, in themselves, be a welcome contribution to the field, but they would also have major implications for the study of women's representation (e.g., Mansbridge Reference Mansbridge1999; Phillips Reference Phillips1995; Young Reference Young2002). Such interventions might also help us solve a profound foundational paradox embedded in the idea of defining women's recruitment of more women as a critical act and, thus, also an instance of SRW. Scholars unanimously agree on the fact that DRW—women's presence—is not the same thing as SRW. But increasing DRW becomes SRW (or, in Dahlerup's term, a critical act). This paradox is bound to cause a conceptual mess in feminist research. The assumptions that underpin it should be assessed critically: Is increasing the descriptive representation of women always, by necessity, SRW or a critical act?
In this deceptively simple question, much is at stake for feminist scholarship. In order to highlight my point in the clearest terms available, I rewrite my criticism in simple, logical terms:
Assumption 1: DRW → More DRW (the recruitment hypothesis)
Assumption 2: DRW → SRW (critical-mass theories)
Conclusion (erroneous): More DRW = SRW
My argument can thus be summarized as follows: The idea that the recruitment of more women into various societal areas, such as politics, equals SRW is based on an erroneous application of logic to the underlying premises.
This observation also gives rise to disquieting notions concerning the genealogical heritage and development of feminist theory. Should we also start asking whether some of the original connotations embedded in “critical mass,” which was borrowed from nuclear physics, have unwittingly been repeated and reproduced in the field of feminist political research. Namely, the end result appears so very similar: Enough/increase of A (one type of matter, here: DRW) is in the end stage transformed to something else, namely, B (other types of matter, here: SRW). As I have argued, it is precisely this kind of a “transformation” of one type of representation, namely, DRW, to another, SRW, that takes place when we conceptualize an increased recruitment of women by women as a critical act.
THE RECRUITMENT HYPOTHESIS IN A POLITICAL CONTEXT: THE SELECTION OF EXPERTS BY PARLIAMENTARY STANDING COMMITTEES
Expert hearings are a crucial element in Finnish parliamentary standing-committee work. “Experts” include all organizations or persons who are consulted by parliamentary standing committees in their legislative scrutiny, including not only professional policy experts, such as civil servants and academic scholars, but also representatives of interest organizations, NGOs, private business, and even ordinary citizens. Expert duty in the Finnish Parliament is considered to be a privilege by all the invited consultees. Although it is obligatory for some, all think it is an honor. From the Finnish standing-committee point of view, hearing experts is considered to be the way of fulfilling the constitutional “duty to prepare.” Moreover, in a political context with a strong state-corporatist tradition, the choice of experts is also a method for exerting influence on the information received, as well as on the contents and quality of decision making (Ahonen Reference Ahonen1980; Helander and Pekonen Reference Helander, Pekonen and Helander2007; Mattson and Ström Reference Mattson, Ström and Döhring1995; Pekonen Reference Pekonen2011, 195–208).
On the basis of earlier research, we can pinpoint some crucial positions of power in the selection of experts by Finnish parliamentary standing committees. At the outset of the reading process, the committee chair and the committee secretary (i.e., a bureaucrat of the Parliament) play a central role in that they jointly sketch an initial proposal concerning which experts are to be consulted in conjunction with a specific bill. The committee deliberates on this proposal and approves it, possibly with additional names suggested by individual members (MPs). During the different phases of the reading, the committee can complement its list of experts when deemed necessary. Notably, the members' suggestions for additional experts are almost always accepted as they stand (Ahtonen, Keinänen, and Kilpeläinen Reference Ahtonen, Keinänen and Kilpeläinen2011, 117–21; Holli and Saari Reference Holli and Saari2009a, 45–46; Svanström Reference Svanström2010).
The criteria affecting the choice of committee experts have been studied on the basis of qualitative interview data (Holli and Saari Reference Holli and Saari2009a, 47–54). Some institutions, such as the ministries that participated in the initial preparation of a governmental bill, are regarded as “obligatory” to consult, whereas there is more room for choice in the selection of other experts. Selection is also strongly influenced by established committee practices—that is, the organizations or persons the committee usually consults. Moreover, the choice of experts depends on the values of individual MPs, their party affiliations, professional backgrounds, ties to constituencies, and knowledge concerning the field in question. As far as the experts are concerned, their title and position, visibility in the media, earlier experience as committee experts, and personal characteristics play a role. In short, the selection of standing-committee experts in Finland displays many informal practices of inclusion and exclusion. On the other hand, some groups of actors are granted access to the policy-preparation process in all of its various stages from the very beginning, and some in none.
On the basis of this information, we can now adapt the recruitment hypothesis in a more contextualized manner (cf. Celis et al. Reference Celis, Childs, Kantola and Krook2008) to parliamentary standing committees, and ask: Do the committees with more women in crucial positions of power regarding the selection of experts select women as experts more often than do other committees? Notably, organization studies indicate that it would be necessary for an adequate test of the recruitment hypothesis to focus on the number of women in crucial decision-making positions (“leadership”). Consequently, I include in the independent variables to be studied not only the proportion of the female members (women MPs) but also the presence of women in two other significant positions: chairs and committee secretaries. The latter is a bureaucratic and administrative office, not a political or elected one.
Some caveats are in order. First, in Finnish parliamentary standing committees, any choice concerning the gender of an expert is actualized only when selecting individual persons as experts. When the committee wishes to consult a specific organization or an interest group, it sends a consultation request and the organization selects the expert. The gender division of the committee—or any other committee characteristic—is, therefore, limited in the extent to which it can affect the gender division of the experts. Second, potential differences in outcome can also be caused by differences in the “supply” of experts. For example, social affairs, health, and culture tend to be female dominated, whereas defense issues are very male dominated.
An increase in the proportion of female experts is a case of improved DRW in the pool of committee experts. For studying the effects of the proportion of women in committees for increasing SRW, however, we need to operationalize SRW in some other manner. In parliamentary standing committees, there exist many opportunities for women MPs (as well as slightly different ones for female chairs and secretaries) to “act for” women; they can propose an expert, discuss and argue for an important issue, lobby members of one's own party group and those of the others for modifying a bill, make a counterproposal as an alternative to the government's proposal in the committee proceedings, or submit a dissenting opinion to the committee's final report to the plenary. Deciding whether and when these activities are SRW depends on how we define and operationalize the concept. It would, however, be extremely difficult to study many of them empirically, since most of them would require large-scale participatory observation data gathered from a closed political arena.
Instead, taking into account the prominent role played by institutions and interest organizations in the hearings, we can take it as a starting point and ask: Do the committees with more women in crucial positions of power regarding the selection of experts select representatives of women's collective-group interests as experts more often than do other committees? For example, Weldon (Reference Weldon2002) has pointed out that scholars should pay more attention to institutional sources of representation for women, such as the autonomous women's movement and gender-equality agencies. In her opinion, they may better represent women's collective voices and interests in the democratic process than relying on the presence and “female bodies” of individual women MPs only. In addition, Halsaa (Reference Halsaa, Nijeholt, Vargas and Wieringa1998) and Woodward (Reference Woodward, Christiansen and Piattoni2003) have argued for the significant role of feminist researchers in the adoption of more women-friendly policies.
Previous research also indicates that we might regard three groups of actors as institutional representatives of women's group interests in the committee hearings: women's NGOs, gender-equality agencies, and gender scholars (Halsaa Reference Halsaa, Nijeholt, Vargas and Wieringa1998; Lovenduski and Guadagnini Reference Lovenduski, Guadagnini, McBride and Mazur2010; Mazur Reference Mazur2002; Woodward Reference Woodward, Christiansen and Piattoni2003). This is the definition utilized for operationalizing the second dependent variable in this study.Footnote 5
Notably, the number of female experts measures DRW in the pool of experts, and the number of representatives of women's collective-group interests measures the procedural, institutional inclusion of SRW within it. Whereas selecting female experts for the most part links to the ability of female committee actors to mobilize institutional resources for the improvement of women's position in committee hearings, women's collective-group interests as a dependent variable also may measure some willingness to do so, not to mention possible feminist motivations in the background (cf. Dahlerup Reference Dahlerup1988).
RESEARCH DESIGN AND DATA
The first question is whether, as the recruitment hypothesis suggests, an increase in the proportion of female actors in parliamentary committees increases the selection of female experts. The second question is whether it influences the selection of experts who represent women's collective-group interests. I operationalize hypotheses 1–4 to examine these questions:
H1:
The higher the proportion of female MPs in committees, the more often they select female experts/experts who represent women's collective-group interests.
H2:
Committees with a female chair select female experts/experts who represent women's collective-group interests more often than other committees.
H3:
The higher the proportion of female committee secretaries in committees, the more often they select female experts or experts who represent women's collective-group interests.
H4:
The higher the proportion of women in positions of power (MPs, chairpersons, committee secretaries) in committees, the more often they select female experts or experts who represent women's collective-group interests.
In addition to sex, we may assume that other committee characteristics influence the selection of female experts directly or indirectly. These factors include “party domain,” a gender-equality mandate, a committee's function, and its established practice of selecting experts.
The party domain of a committee affects both the gender composition of a committee and its selection of experts.Footnote 6 Finnish parliamentary standing committees have typically been situated within a certain party's “domain of interest” in a relatively stable manner. The overall party composition of the committees reflects proportionally the composition of the Parliament. Chairperson positions in the committees, however, are negotiated between the parties and divided among them at the beginning of a parliamentary term, thus reflecting the strength of interest of the parties in different policy sectors. Previous studies (e.g. Norris Reference Norris1987; Reference Norris, Lovenduski and Norris1992) demonstrate that support for gender equality and the recruitment of women as candidates is more prevalent for the Left than for the Right, leading us to expect that committees situated in the left-wing domain would select more female experts (or representatives of women's collective-group interests) than those in the right wing.
A gender-equality mandate, the function of a committee, and established practices of drawing experts from a specific sector can also influence the selection of experts. Committees that have an explicit responsibility for gender-equality issues may attract more female MPs as members, as well as being more conscious of gender-equality concerns in their selection of experts (or representatives of women's collective-group interests). Further, committees with sociocultural functions tend to have more women in all positions of power, as well as a larger supply of female experts at hand and, consequently, use women's expertise more often than other committees do. Finally, Finnish parliamentary standing committees have rather established practices of expert selection that draw more on certain sectors of society (public, private, third [sector], scientific experts) than others. We may expect that committees that deploy more public and third-sector experts invite more women as experts, and committees that rely more on private-sector or academic experts invite fewer female experts than on average.
The empirical data consist of expert hearings in parliamentary standing committees in 2005 (the second complete calendar year of the 2003–7 parliamentary term). During the first calendar year of a parliamentary cycle, the work of the committees begins and there are usually fewer bills; the last year of the term sees more work than on average. Previous studies on expert hearings in Finnish parliamentary committees have made the same choice (Helander and Pekonen Reference Helander, Pekonen and Helander2007). Data on the experts, their background organization, and their sex were coded manually into a database from 260 reports (on government bills) and 340 statements (given to other committees) by the 15 committees functioning at that time. The database consists of 5,187 organization consultations, including 4,630 hearings by individuals. It was completed with information from other sources, especially statistics produced by the Parliament and earlier studies on other relevant factors (Helander and Pekonen Reference Helander, Pekonen and Helander2007).
I run an analysis for each dependent variable: selecting a female expert (0 = man, 1 = woman), and selecting an expert representing women's collective-group interests (0 = no, 1 = yes), regardless of the expert's sex. Women's NGOs, gender-equality agencies, and gender scholars are defined as such experts. I include the following independent variables:
Gender: the proportion of women as committee members and committee secretaries (the gender of the committee chair is dichotomous).
Party domain: the party affiliation of the committee chair.Footnote 7
Gender equality mandate: a dichotomous variable (0 = no, 1 = yes). Only the Committee for Employment and Equality has such a mandate.
Sociocultural function: a dichotomous variable defined on the basis of the standardized BEIS-categorization.Footnote 8 Three committees have such a function (Education and Culture, Employment and Equality, and Social Affairs and Health).
Established expert practices of committees: the proportion of third-sector expertise, calculated from information provided by Helander and Pekonen (Reference Helander, Pekonen and Helander2007, 94–95).Footnote 9
The effect of the independent variables on both of the dependent variables is examined with logistic regression analysis in seven models (a–g). To control for the nested structure of the data, I use a robust-clustered standard-errors approach that allows for variance within clusters.
EMPIRICAL ANALYSIS
Selection of Female Experts
Table 1 demonstrates that, when we examine the effects of female committee actors separately, the proportions of women as MPs and committee secretaries both have a positive impact on the selection of female experts (models a and c), whereas the gender of the committee chair does not have an influence (model b). When all the gender variables are included in the same model (d), none of them at a first glance seems to be significant. When testing whether any of these variables has a coefficient that differs from zero, however, we find (at p < 0.01) that one or several gender variables indeed do have an effect; they simply cancel each other out when modeled jointly.
Effects of committee characteristics on the selection of female experts, with robust standard errors (nested in committees) in parentheses
Notes: Because of a missing value in Helander and Pekonen's (2007, 94–95) data set, the number of observations (N) is smaller in Models f and g.
***p < 0.001, **p < 0.01, *p < 0.05, †p < 0.1.
Model e illustrates that Left or Green party domain increases the selection of female experts. Contrary to expectations, a gender-equality mandate exerts a negative effect on the selection. This result might be linked to the twofold nature of the Committee for Employment and Equality. On one hand, this committee examines issues of gender equality, and usually a good number of female experts are heard in conjunction with this category. On the other hand, the committee also handles issues dealing with work and employment, where the experts heard and the organizations they represent are quite different from those who participate in conjunction with gender-equality issues. Work and employment issues also make up a much larger part of the activities and the consultations of the committee than do equality issues. Model e also shows that a committee's function within the sociocultural sector is positively associated with the selection of female experts. What is noteworthy here, is the finding that the gender effect disappears after adjustment for the other committee variables. The overall conclusion is simple: It is the other variables, not the gender of the committee actors, that affect the selection of female experts.Footnote 10
Model f also shows that a Left/Green party domain increases, and a gender-equality mandate decreases, the selection of female experts. It also demonstrates that established committee practices have a strong effect on the selection: The more the committee draws from the third sector, the more often it selects women as experts. By contrast to results from Model e, the proportion of women MPs is highly significant for the selection of female experts.Footnote 11
The final model (g) combines the independent variables. Once again, a Left/Green party domain and a high use of third-sector experts increase, and a gender-equality mandate decreases, the selection of female experts. The committee's function in the sociocultural field becomes insignificant. In the final analysis, the proportion of women MPs almost reaches the conventional level of significance (p = 0.053) for the selection, suggesting somewhat ambiguous, at best weak, support for H1. By contrast, the proportion of female secretaries and chairs does not play a role. Indeed, the effect of female chairs is negative. These results, however, should be treated with caution due to high multicollinearity.
As logistic coefficients are difficult to interpret, the impact of women MPs is illustrated graphically in estimated probabilities.Footnote 12Figure 1 demonstrates that an increase in the proportion of female MPs in a committee increases its probability of selecting female experts. The widening confidence intervals, however, show that the estimates become less reliable when the proportion of women MPs deviates from the optimal point, around which most of the observations in the data are concentrated.
Estimated probability of female experts when the proportion of women as committee members increases.
In order to form a more comprehensive picture of the magnitude of the joint effects of gender and other committee characteristics, Table 2 presents the estimated probabilities of selecting female experts from Model g by varying the values of the other impacting factors. Table 2 shows that the gender effect is strongly influenced by the other committee characteristics. There is a substantial difference in a probabality of selecting a female expert even at the initial stage when a committee has few female members. As the proportion of women MPs in committees increases, the overall propensity of committees to select a female expert increases. The magnitude of the gender effect, however, varies considerably, depending on other committee characteristics.
Estimated probabilities of female experts based on minimum/maximum values of female committee members, party, equality mandate, and use of third-sector experts
*Minimum value in the data was 12% of female members of all committee members.
**Maximum value in the data was 65% of female members of all committee members.
***Minimum value in the data was 1% of all experts consulted by the committees.
****Maximum value in the data was 44% of all experts consulted by the committees.
*****Mean values given to all other committee characteristics except the proportion of women as committee members.
On the whole, contrary to findings in organization studies, women's overall presence in power positions in committees does not increase the propensity for selecting female experts. The weak positive effect is limited to female MPs only, suggesting that the behavior of women as secretaries and chairs of standing committees is shaped by committee culture rather than gender concerns. Notably, other committee characteristics than the proportion of women MPs play a stronger role for the selection of female experts. The results also reinforce the importance of controlling for the supply side when testing the recruitment hypothesis. They also show that the gender effects are not robust; rather, they are sensitive to the selection of control variables.
Selection of Experts Representing Women's Collective-Group Interests
When we turn to the selection of experts representing women's collective group interests in Table 3, we see in Models a–c that, contrary to expectations, none of the groups of female committee actors has an effect on the selection of experts. When we control for other women's presence, though, one group — namely female MPs — has an almost significant effect (model d). Notably, there were very few experts representing such interests in the data (38 consultations out of 5,187 organizations consulted). Models e and f demonstrate the significance of a committee's gender equality mandate. Party, sociocultural function, and area of practice are all insignificant. The gender of committee actors does not matter for the outcome of selection either. And the effect of female committee secretaries is actually negative.
Effects of committee characteristics on the selection of representatives of women's collective group interests as experts, with robust standard errors (nested in committees) in parentheses
Notes: Because of a missing value in Helander and Pekonen's (2007, 94–95) data set, the number of observations (N) is smaller in Models f and g.
***p < 0.001, **p < 0.01, *p < 0.05, †p < 0.1.
The final combined model (g) indicates that the gender-equality mandate of a committee is the most important factor for selecting women's collective-group representatives as experts. An increased proportion of female MPs in a committee does not increase the selection of women's collective-group representatives; rather, it diminishes it, although the negative effect does not quite reach the conventional level of statistical significance. What stands out here again, though, is that it is the other committee characteristics, not gender, that account for the selection of representatives of women's collective-group interests. Notably, several of the committee characteristics that affect the selection of female experts positively appear to have an adverse effect for the selection of representatives of women's group interests, and vice versa.
The finding concerning women MPs' negative effect on the selection of representatives of women's group interests contradicts the basic assumption of critical-mass theories, which argue that an increasing number of women MPs would change politics and induce improved SRW. There are two interpretations on offer. The first explanation points to the socializing effect of committees: Female MPs are probably not any readier than are their male counterparts to deviate from their committee's culture and established practices of consulting “the usual suspects.” Each committee has a number of established experts on whom it relies, and “new” or alternative experts from a different “field” are rarely consulted. Moreover, especially in the case of women's NGOs, the established selection criteria for experts, namely, a committee's answers to the question “Whom does this issue concern?” are defined rather narrowly (Holli and Saari Reference Holli and Saari2009a, 51–52). Previous studies show that in the Finnish Parliament, consulting women's NGOs was, for the most part, deemed relevant only in the case of legislation dealing explicitly with gender equality (pp. 51–52). This was generally the case regardless of the fact that nearly all policies and legislation affect both sexes and might have different outcomes for women and men.
An alternative explanation is that Finnish female MPs already often view themselves as representatives of women's interests and gender-equality concerns, at least to some degree (cf. Holli Reference Holli, Hellsten, Holli and Daskalova2006; Holli and Saari Reference Holli and Saari2009b). Accordingly, they may consider themselves capable of bringing women's concerns into the committee arena by themselves. Many of them are also active in their own party's women's organization or women's cross-party coalition networks nationally or within the Parliament (Holli and Kantola Reference Holli, Kantola, Outshoorn and Kantola2007). In addition, because of the gendered division of policy sectors, higher proportions of both female MPs and female experts tend to concentrate in the same committees and policy fields. Having plenty of “female bodies” present, can, mistakenly, be interpreted as sufficient “representation of women's collective-group interests.” For these reasons, women MPs might overlook the role played by experts in shaping the agenda for deliberations and decision making in standing committees.
CONCLUSION
This study has offered an assessment of the foundations of “critical mass” and “critical acts” and proposed a revised research agenda and methodology. On the basis of such, the recruitment hypothesis, which forms the crucial link between critical mass and acts—and DRW and SRW—was tested in a political context: in the selection of experts by standing committees in the Parliament of Finland. The empirical results point to the following conclusions: First, in contrast to results achieved by the study of private-sector organizations, at least in Finnish parliamentary committees, it is not women in leadership positions in general we should look to as potential agents of change. Rather, it is the women MPs who are the agents for change, as also claimed by both critical mass and critical acts approaches. Second, the proportion of women MPs in a committee has a weak and inconclusive positive effect on the selection of female experts (DRW). Nonetheless, it is the other committee characteristics that play a more important role. Nevertheless, this finding does not, as such, tell us anything about actors' feminist motivations: Perhaps female MPs just happen to know more women with the needed knowledge in the relevant field and, as a consequence, propose them as experts more than their male counterparts do. Finally, it is not the gender of the committee actors but, again, the other committee characteristics that increase the selection of representatives of women's collective-group interests as experts. In fact, higher proportions of women MPs in committees rather seems to diminish it.
The methodological and theoretical insights gained by this study should, however, be of even greater interest to feminist scholars. In addition to pointing out the need to take context better into account when testing the recruitment hypothesis, this article has revealed some serious problems and logical fallacies in the fundamentals of the concepts on which feminist political research and theorizing rely. My criticism targets the fact that increasing DRW (as in recruiting more women to politics) tends to be defined as SRW, or critical acts, which, basically, is to claim that DRW transforms into SRW at some point. On grounds of conceptual clarity, I disagree with this line of thought. Methodologically, my standpoint implies that DRW — also an increase of DRW (as in the recruitment hypothesis) — must be kept conceptually distinct from SRW and critical acts when studying the impacts of women's presence. This result is relevant not only for theorists, but also to empiricists seeking to operationalize these concepts.
From the perspective of feminist scholarship, my findings and the implications we can draw from them may seem quite perturbing. The results question several assumptions we have taken as “truths” and evidence of a progressive accumulation of knowledge concerning certain central concepts, theories, and empirical evidence on women's political representation. They may also imply that there are other problems in the background, namely, a deep segregation between theoretical and empirical strands of these types of studies. In the final analysis, what this article calls for are serious attempts to bridge this gap, an enterprise that would demand increased reflexivity and collaboration by both strands. In order to verify or falsify my empirical findings concerning the impact of women's presence, we will need not only better data and better models and ways of testing them but also better ideas about how to solve the problem of “unboundedness” in existing theories and their conceptual foundations.
From the viewpoint of feminist politics, my results may seem equally troublesome, though for a different reason. The findings indicate that increasing the proportion of women MPs in a committee may increase the selection of female experts, but it does not guarantee that gendered needs, concerns, and impacts are taken into account in legislative scrutiny. Not all female experts “represent women” substantively. In order to have information about gendered concerns and viewpoints, the standing committees must also consult groups, institutions, or individuals with a mandate for representing women's interests or knowledge concerning gender and gendered issues. In the Finnish Parliament, it is only the Committee for Employment and Equality that fulfills this requirement to some degree. Women MPs have not paid attention to the issue. Notably, should they wish to do so, women MPs always have the opportunity to “act for women” within committee work in a variety of ways (Holli and Saari Reference Holli and Saari2009b). Regardless, there remains the need to include more fully the representation of women's collective-group interests in committee hearings.
