Female representation and legislative behavior in the Italian Parliament ( )

Transcription

1 Female representation and legislative behavior in the Italian Parliament ( ) Licia Claudia Papavero Università degli Studi di Milano DRAFT VERSION - PLEASE DO NOT QUOTE WITHOUT AUTHOR S PERMISSION Abstract: While dynamics in female legislative recruitment have been considered in several studies and researches about the Italian politics less attention has been paid to the legislative activity of women elected in the Italian parliament, and to whether it differs somehow from that of their male colleagues. Once elected in the parliament, do women MPs display, vis à vis men MPs, distinctive legislative interests? What type of priorities do they display in their legislative work? How do women s priorities change as the number of women MPs increases over time? And if any, what is the impact of women MPs activity on the policy priorities embraced by men MPs? Through a quantitative analysis of the bills introduced in the Italian Chamber of deputies in six legislatures, over a time span of twenty years, this paper tries to answer these questions. Paper delivered at the International Conference on democracy as Idea and Practice, Oslo 14th-15th January

2 Female representation and legislative behavior in the Italian Parliament ( ) 1 Licia Claudia Papavero Università degli Studi di Milano licia.papavero@unimi.it After years of oscillation around the 10% threshold, the 2006 Italian general elections seem to have outlined a new pathway towards a more steady and significant increment of the number of women MPs in the Italian Parliament. Indeed, in the Italian lower Chamber, in the XV legislature ( ) women accounted for more than 17% of the 630 elected deputies, and today they account for almost 21%. There are several variables that have contributed to this achievement, and among the others especially the reform of the electoral system that has reintroduced the principle of proportional representation and has favored a contagion among all the main Italian parties regarding their patterns of female legislative recruitment (Papavero 2008). For the first time in the Italian democratic parliament, the gap between left-wing parties and (center-) right parties has started to reduce. The most striking innovation has occurred in the center-right area, where women, very underrepresented in the past legislatures (below 10%), have increased their presence dramatically, reaching today almost the 20% threshold. While dynamics in female legislative recruitment have been considered in several studies and researches about the Italian politics (see for example Guadagnini 1987 and 1993; Cotta, Mastropaolo,Verzichelli 2000), less attention has been paid to their political consequences. In particular, less attention has been paid to the legislative activity of women elected in the Italian parliament, and to whether it differs somehow from that of their male colleagues. This issue, more than being a simple curiosity, is grounded on theoretical bases, and is relevant to the theory of representation. It stays, at a certain extent, at the intersection between descriptive and substantive representation theories. The theories that stress the importance of descriptive representation of historically disadvantaged social groups, like that of women, argue that the shared experience imperfectly captured by descriptive representation may facilitate the relationship of accountability between representatives and citizens by improving the communication between those elected and voters. In particular, it may help representatives to act according to not yet explicit preferences, expectations and values of their constituents, and to give a public relevance to some specific demands and interests that otherwise would be overlooked in the decision-making (Mansbridge 2000). Feminist theory, on the other hand, suggests that there is a set of issues that can be identify as specific interests of women, which may be supported by women politicians in their activities as representatives. In particular, in their institutional activity women representatives not only may 1 I wish to thank Gabriele Restelli for valuable research assistance. I wish also to acknowledge the financial support given to this study by the national research project PRIN 2007 I luoghi del legislativo, i luoghi delle politiche. Giochi, veti, reti nell Italia dell alternanza, coordinated by prof. Marco Giuliani. A preliminary version of this paper was presented at the ECPR Conference of Postdam, in September I am grateful to the participants in the conference for their useful questions and comments. 2

3 express different types of values and attitudes, but also they are very likely to give priority to those issues that are closer to their particular life-experiences in the home, workplace and public sphere, such as childcare, health, education and welfare (Lovenduski and Norris 2004; Phillips 1995). The increased number of women elected in the Italian parliament in the last twenty years and the progressive reduction of the gap between left and right parties in women s legislative recruitment allow to empirically address questions and hypotheses related to those theories, some already explored for other political and cultural contexts. This is what this paper does, in a partial and limited way, focusing on the Italian women representatives policy priorities, in order to control whether women tend to represent substantively women. Once elected in the parliament, do women MPs display, vis à vis men MPs, distinctive legislative interests? What type of priorities do they display in their legislative work? How do women s priorities change as the number of women MPs increases over time? And if any, what is the impact of women MPs activity on the policy priorities embraced by men MPs? Through a quantitative analysis of the bills introduced in the Italian Chamber of deputies in six legislatures, over a time span of twenty years, this paper tries to answer these questions. Literature review The behavior of female legislators has been at the center of an almost huge literature in the last forty years. The studies on this topic have usually confirmed that women display values, attitudes, styles and policy priorities in their legislative behavior different from their male counterparts, although those differences may change over time and according to the policy at stake. Some studies on the US Congress highlight, for example, that female legislators usually display more liberal values than their male colleagues, voting in a more liberal direction than men. However, these patterns of voting become less significantly different over time and they seem to be almost largely based on the different constituencies of men and women MPs (Welch 1985, Vega and Firestone 1995). Lovenduski and Norris (2004) find that women and men MPs in the British parliament, no matter their party affiliation, do not differ significantly in their support towards freemarket left-right economic values, their attitudes towards Britain s role in the EU, and moral traditionalism. Yet, significant differences raise when issues directly related with women interests such as affirmative action and gender equality are at stake. Several other studies confirm that women tend to think of themselves as representatives of women and to consider women as a constituency group with specific concerns (Thomas and Welch 1991; Thomas 1991; Reingold 1992; Bratton and Haynie 1999; Wängnerud 2000). These studies point out that women tend more than men to list as prior in their legislative agenda those policies related with women s rights, children and family issues, such as for example reproductive rights, sexual rights, women s illnesses prevention, parental leave, domestic violence, marital status, children care and so forth. Comparing the US and Argentine low Chambers, Jones (1997) finds that in both countries women MPs assign much higher priority than men MPs to policies in the areas of women s rights, children 3

4 and family, while in other policy fields considered traditional areas of interest of women, such as health, education, welfare and environment, there are no significant differences between men and women's policy priorities. Similar findings are stressed in the study of the 35 th legislature in Canada (Tremblay 1998). Here, although women MPs are more likely than man to devote their parliamentary activity to issues concerning women, however this activity is mainly focused on issues related to family and children. Moreover, in general, for both man and women MPs, the parliamentary activities in favour of women remain marginal from a quantitative point of view. While stressing a different legislative behavior between man and women when we consider policy areas of interest for women, this literature also points out that we should expect variations in the policy priorities displayed by women representatives as their number increases in the representative institutions. Using the theoretical framework proposed by Kanter (1977) in her study of corporate sale forces about the effects of minority status on behavior of those in minority, Thomas and Welch (1991) argued that when women representatives are still a skewed group (about 15% or less), they may feel constrained in their behavior, and it is more difficult for them to translate feminist concerns into their policy activity. The fear of being labeled as too narrow or only interested in women s issues may be, for example, at the origin of such a difficulty. However, as the size of this minority group increases, it should be easier for women to make alliances, and to act in a more assertive way, eventually transforming the institutional culture, norms and values. They found that, in the US, women officeholders in the lower houses of twelve state legislatures were more active than women serving as state legislators ten years before, and that their priorities were more focused on issues dealing with women, children and family than they were a decade before. More interestingly, Thomas (1991) found that, in some US state legislatures, the different proportions of women among legislators had an impact not only on women representatives policy agenda, but also on that of the entire legislature. Women officeholders were found to gave higher priority to issues traditionally considered of interests of women in those state legislatures with a more balanced presence of men and women among legislators rather than in those state legislatures where women were still a skewed group. In the same legislatures with higher proportions of women, moreover, the degree of success of women s legislative proposals was higher than in the other legislatures, and also men legislators tended more to prioritize legislations dealing with women, children and family. Hypotheses, data and method Within the pathway of the stream of studies above reviewed, in the next sections I will try to test and discuss two main hypotheses: H1. Women are more likely than men to introduce bills concerning women s rights, children and family, while for policies concerning health care, education, welfare and environment women do not differ significantly from men. 4

5 There are several reasons to expect that the second part of this hypothesis is confirmed for the case of the Italian parliament. The peculiar structure of the social welfare in Italy, which is basically particularistic and oriented to sustain particular social categories; the profound social and economic unbalance between Northern and Southern regions of Italy, with a strong intervention of the central government on economy and social life especially in the South; the consequent strong influence of political parties, especially those in the cabinet, on economic and social life; all these reasons make it plausible to expect that other factors rather than gender may predict the tendency to prioritize welfare policies in the legislative agenda. We can have a similar expectation also for health care and education. Italy has an health care system publicly funded whose employees are basically public employees and whose managers are sometimes selected according to political affiliation. So, it is a policy sector in which the interests of several influential categories ask for representation and it is not plausible to think those interests are not as prior for men legislators as they are for women. Education in Italy is mainly public, and still it is a strong basin of public employees recruitment. Moreover, it is also one of the field in which the traditional cleavage Church/secularization is still lively, with the Church asking for public funds to sustain private Catholic schools. Also in this case, several different constituencies are implicated in this policy field, so that it is not plausible to think men and women legislators consider them with different intensity in their legislative activity. The second hypothesis I consider is about the influence of changing proportions of women legislators on the legislative agenda. Following Thomas (1991), I expect that H2. The increase of the number of women in the legislature has implications on the legislative behavior of women: the higher the number of women MPs, the higher will be the priority they give to issues traditionally considered of interests of women. In order to identify the policy areas relevant to this study, the bills introduced by every single MPs at the Chamber of Deputies were gathered for six legislatures, from the X ( ) to the XV ( ). The bills were downloaded from the web site of the Italian Chamber of deputies using a technique of web scraping, that translated them directly into excel files. Those file were then formatted in SPSS. I considered only bills introduced by MPs 2, and not those sponsored by the national government, by the regional assemblies, by the CNEL (a constitutional body of experts in economics and law) and by citizens (the so called leggi di iniziativa popolare ). The huge amount of bills sponsored by MPs in each legislature (see tab 1) testifies that this activity is free and not particularly demanding for MPs: no parliamentary rule requires a minimal number of sponsors or the political group s allowance. Therefore sponsoring bills reflects quite faithfully the sincere MP s preferences. The parliamentary bills were then classified according to their titles and to the classification codes made by the parliamentary offices into seven categories: women s rights (WR), 2 Almost the majority of bills have co-sponsors, which however have not been considered in this study. 5

6 children and family (C&F), health (H), education (E), welfare (Wel), environment (Env) and a category other, comprising bills dealing with issues different from the former ones 3. Following Jones (1997), I considered also a category in which women s rights, children and family were collapsed together (WCF), as several issues falling in the two categories are very often overlapping. Each policy area was re-coded as a dummy, with value 1 for the specific area and 0 for the others, and put as dependent variable into seven different logit models for each of the six legislatures considered in this study. The models measure what is the impact of the explanatory variables on the probability that a bill concerning one of the specific policy areas is introduced. Tab 1. Number and percentage of bills by policy area and legislature Legislatures X XI XII XIII XIV XV Policy Area (N) Spons.by Spons.by Spons.by Spons.by Spons.by Tot W. Tot W. Tot W. Tot W. Tot W. Tot Spons.by W. Other Women's rights Children&Famil y Welfare Education Public Health Environment TOT Policy Area (%) Other Women's rights Children&Famil y Welfare Education Public Health Environment TOT The explanatory variables Together with the main explanatory variable, that is, gender (male =1; female=2), I included in the models (also for control purposes) the age and the tenure in office of the MPs, the constituency of election and the partisan affiliation of MPs. Age is the MPs age in years at the beginning of each legislature. Tenure is coded as a dummy variable in which new elected MPs are coded as 0 and MPs with one or more legislatures of parliamentary experience are coded as 1. The constituency of election as well is a dummy in which the electoral districts of South Italy are coded as 1, and the other districts 4 are coded 0. This 3 For details on the classification criteria, see the appendix. 4 As the electoral system has changed three times in the time span considered in this study, changing also somehow the design of the electoral districts, I considered the region in which the MP s district of election is comprised rather than the actual district. So, regions of the South were coded 1. Code 0 was assigned to the Center and Northern regions, but also to the nation-wide district (the so called Collegio Unico Nazionale) in which in the X and XI legislatures the seats were 6

7 variable should account for the impact of cultural/political differences on the legislative agenda. Southern regions are less economically developed; they are the regions in which the parties of the center-right and of the right have usually gained more votes, and in which women elected have always been very few (at least until 2006). So, if this variable may have an impact on the priority MPs give to policies related to women s concern, I expect that being elected in a southern electoral district is negatively correlated with the dependent variable. Partisan affiliation is a variable that accounts for the parliamentary group of which an MP is member when introducing the bill. An MP s parliamentary group does not necessarily overlap with his/her party of election. As the standing orders of the Italian Chamber of deputies give to MPs a large amount of freedom to move from one parliamentary group to another, and also to create new groups with a minimum number of deputies, it is possible that the same deputy introduced some bills under a certain label, and some under another one. This phenomenon, almost insignificant in the X and XI legislatures, has become a little bit more relevant from the XIII legislature onwards. Moreover, as a reflex of the increased fragmentation of the Italian party system and of its continuous transformation over the period here considered, also the number of parliamentary groups has increased. In order to give to the analysis over time a certain degree of homogeneity, and to simplify a very complex political and ideological scenario, when it was possible, I grouped together different parliamentary groups, according to their ideological or political proximity, and I singled out basically six political families 5 : 1) the new left ; 2) the traditional left ; 3) the moderates ; 4) the Catholic center ; 5) the Right ; and 6) Others. These groups have different numbers and proportions of women as members (see Tab. 2). The legislatures Before running the statistical models, some words should be spent about the characteristics of the legislatures considered in this study. The six legislatures cover an historical period of important changes in the Italian political system. The X ( ) and XI ( ) legislatures are the last two legislatures of the so called Italian First Republic. MPs were elected with a highly proportional electoral system, with preference vote. The party system presented still some characteristic of the first legislatures of the Republic: the pivotal role played by the large center party, the Christian Democratic party (DC), which had stably governed either alone or in coalition with (or with the external support of) the parties to its left (PSI, PSDI) and right (PRI, PLI), while the main left-wing party, the Communist Party (PCI), remained permanently in opposition, and the presence of an important extreme right party (MSI) 6. By the end of the 1980s, the signs that the old assigned according to the rests of the percentages of votes collected by parties at national level, and to the abroad district that was introduced in the XV legislature by the new electoral reform. 5 See the appendix for a summary description of the composition of each group. 6 These features have been synthesized in the model of polarized pluralism proposed by Sartori (1970 and 1974) to define the Italian party system. Other authors (Hine 1993) classifies the last phase of the first Republic as characterised by the syndrome of bargained pluralism, that is, by the permeability of governmental institutions to the pressures of (all kinds of) interest groups and the bargaining character (at all levels) of the decisional process ( ). 7

8 mass parties had begun to decline had become more evident. In the case of the PCI, this process was to lead to the transformation of the party into the Democratic Party of the Left in 1991, and the break way of one section which went on to form the Party of Communist Re-foundation (PRC). In the case of the parties within the area of government, including the smallest ones, this process was to develop into an electoral weakness in the Northern areas of the country (but not in the South), where the Northern League and the Veneto League began to erode their electoral consensus. The ensuing corruption scandals were to lead to the disappearance of the entire political elite of those parties after The legislatures XII ( ), XIII ( ) and XIV ( ) are usually considered the legislatures of the so called Second Republic. The early 1990s witnessed the beginning of the radical transformation of the Italian political system. The changes made at that time resulted in the collapse of the existing political order that had governed Italy since the dawn of the Republic, and in the advent of a new political order. The 1994 elections have been seen as one of the great turning points in Italy s political history (Cotta, Verzichelli 1996). The organizational crisis of the old mass parties, with the disappearance of some of the main protagonists of political life since the beginning of the Republic, and the emergence of new political actors, together with the introduction of new electoral rules, are only some of the elements which characterised the process of transition to what is referred to as the Second Republic. In 2006 the new legislature, the XV ( ) was elected with a new electoral system, which re-introduced the proportional representation, but with closed lists. Tab. 2 Percentage of women MPs in each parliamentary group in the Chamber of deputies X XI XII XIII XIV XV New Left 32,3 8,3 New Left 15,1 18,8 33,3 24,1 PCI 26,9 18,1 PDS 29 19,1 23,2 PSI 5,1 4,2 Margherita 22,1 8,1 6,7 Moderates 0 1,6 Ulivo 20,7 DC 4,5 4,7 Catholics 11,9 2,1 6,3 9,4 MSI 2,6 5,9 Right 10,7 7,7 6,9 16,4 Other 0 11 Other 0 10,6 18,2 9,7 Total % W MPs 12,4 8, ,3 11,5 17,6 (Total N MPs) (670) (652) (638) (652) (644) (649) Data analysis and discussion 7 The statistical models confirm partially the expectations of hypothesis 1. Also in the Italian parliament women tend more than men to prioritize bills concerning issues directly connected with women s life experience, those that here were classified under the labels WR and C&F, while for bills concerning Wel and Env gender differences are less pronounced and more variable over time. Contrary to my expectations, however, in all the legislatures here considered, education is 7 The statistical models are in the Appendix 2. 8

9 confirmed to be another policy area of special concern for women MPs, and in some legislature also health policy area results significantly prioritized according to MPs gender. Hypothesis 2 seems disconfirmed: if the number of women MPs plays a role, then it seems to be opposite to the expected one. In fact, the results of the statistical analysis seem to suggest that when the number of women MPs increases in the legislatures, the priority they display on women's issues decreases. In order to make as clear as possible the data analysis and discussion, I discuss the models concerning WR, C&F and WCF separately from the other models. Women s rights, children and family policies In all the six legislatures considered in this study, gender is the explanatory variable that most accounts for the probability that a bill about those issues is introduced in the legislature. Once controlled for the other variables, gender is always very significant, and its predictive power (as measured by the Exp(B)) is the highest also among the other variables whose impact is significant. This happens in all the three models that have as dependent variables respectively WR, C&F and WCF. Moreover, the models about WR and WCF show almost always a predictive power considerably higher than that of the models about C&F. In fact, WR models predictive power shows a range of variation that goes from a maximum of almost 21% of the explained variance in the X legislature to a minimum of almost 12% in the XIII legislature. C&F models, on the other hand, show a predictive power that only in the X legislature overcomes the 10%, and riches its lowest value in the XV legislature (2.3%). These differences parallel those registered in the Exp(B) for the variable gender in the same models in each legislature, and they seem to stress once again that, when issues related to women s rights are considered, descriptive and substantive representation of women tend to strongly overlap. These statistics seem to suggest that, without the presence of women in the parliament, it is very likely that those issues would have been disregarded, while issues related to C&F would have received a certain (although little) amount of attention anyway. The variations of the intensity of the predictive power of gender over the six legislatures deserve additional attention as they seem to contradict the Hypothesis 2. In the WR models (and in WCF models) the highest intensity is displayed in the X legislature, when being a woman increases more than 20 times and 14 times the probability that a bill on WR or WCF is introduced. After the X legislature a general trend of decrement of the intensity starts, reaching the lowest score in the XV legislature (Exp (B)= 6.9 for the WR model and Exp (B)=4.023 for the WCF model). How to interpret this decreasing trend in the legislatures in which the number of women MPs is stable or increases? Some hypotheses can be considered. One hypothesis is that those differences reflect a different degree of ideological commitment of women MPs towards the issues WR and WCF. Maybe, it is not by chance that the highest score is displayed in the X legislature. During 1987 general elections, in fact, a specific campaign was conducted by a group of women within the 9

10 Communist Party, aimed at the election of more women to the parliament. This group was led by Livia Turco, head of the PCI s Secretariat for Women, and included some women from the Italian feminist movement, which together proposed a political document (the Carta itinerante delle donne (Itinerant Charter for Women) that attempted to place the issue of female representation at the centre of the PCI s political agenda. However, the relationship with this section of Italian feminism had been a conflictual one from the very beginning, and tension arose mainly over the different meanings given to the concept of female representation. So, after the 1987 elections no repetition of this cooperative experience took place. Yet, the 1987 campaign, accompanied by the slogan woman votes for woman, led to a substantial increase in the number of women elected on PCI tickets, and to a small increase also in the governing parties and in the party of the extreme right. That slogan implicitly suggested that, by voting female candidates within any party list, women voters would have been also substantively represented in the parliament. It is likely that the legacy of this feminist campaign had lasted for the entire legislature, and partially for the following one, giving to women elected in all the Italian main parties a major sense of being the representatives of women s interests in the parliament. After that legislature, this legacy has probably attenuated, and probably a less feminist-minded new generation of women MPs has entered the parliament. Unfortunately, at the moment we do not have empirical data that may help to corroborate this interpretation. Another possible explanation of the different intensity these statistical models report may be that a sort of labor division mechanism is at work, which varies according to the variations of the number of women MPs in the legislatures. According to this interpretation, the increase of the number of women MPs would have consequences on the policy agenda of women MPs and also of men MPs but in a direction opposite from the predominant expectations in the literature. Since the XII legislature, there is a general trend of growth of the absolute number of women MPs (despite some oscillation) in comparison with the X and XI legislatures. This general trend may have contributed to change the priority ranking in the policy agenda of women MPs. When they were less numerous, they tended to specialize more their policy agenda, giving more priority to those issues related with WR and C&F. It is very likely that a sort of labor division was at work within each party, which led women to have a lexicographic ordering of their policy priorities: the few women present probably felt to have to deal first and foremost with women s issues, and then with other issues. As the number of women MPs started increasing, women could start differentiating their policy agenda: WCF issues were always at the top of their lexicographic ordering, but now also other issues ranked closer to the top. This is particularly clear in the last legislature, the XV, which is also the legislature with the highest number of women MPs. On the other hand, also men MPs policy priorities started to change: as the fig. 1 shows, there is a general increasing tendency of men MPs to deal with WCF issues, a trend that had accelerated since the XIII legislature. Whether this tendency can be seen as a simple by-product of the increasing number of women MPs, as suggested by some previous studies, remains, in my view, still an open question, as I will discuss 10

11 in the conclusions of this paper. Anyway, this tendency may account (at least partially) for the decreasing intensity of the predictive power of the variable gender in my statistical models. Fig. 1 Percentage of bills on WCF introduced by women and men MPs between X and XV legislature % bills on WCF M W 5 0 X XI XII XIII XIV XV Legislatures Another interesting finding emerging from the statistical models is that in the models about WR and WCF (and also in C&F), the affiliation to some parliamentary groups displays for almost all the legislatures a significant impact on the probability that a bill on those issues is introduced. In the X legislature, if an MP belongs to a group like the Socialist, the secularized Moderates and the Fascist, coeteris paribus, he/she seems to be more likely to introduce a bill on WR or WCF than an MP belonging to the Communist party (which is considered as basis for the statistical analysis). In the XI legislature, being an MP belonging to the MSI increases significantly the same probability. Still, since the XII legislature, being member of a Catholic centrist group, or of the moderate group of the left coalition (which has a significant presence of Catholic MPs within), increases significantly the probability that a bill on WR and WCR is introduced rather than being member of the group of the new left (which is considered the benchmark group in the statistical analysis). Does it mean that ideology counts in defining policy priorities? It is plausible to argue that Catholic MPs (especially those members of Center parties) and MPs belonging to traditionalist/right oriented parties (like the fascist MSI) are more likely to prioritize policies about family and some issues (like abortion) that are very important for the doctrine of the Catholic Church, from different perspective (conservative the center/right oriented Catholics and more progressive the left-oriented ones). At the same time, parties more secularized and liberal oriented (like the moderates and the Socialist in the X and XI legislatures) are more interested in the same issues from a reformist perspective. 11

12 Health, education, welfare, environment policies As written in the introduction of this section, the models concerning health, education, welfare and environment policies are less significant than those on WR and WCF. In the models concerning environment, gender is slightly significant only in the XIII legislature, but with a negative sign: being a woman would decrease the probability that a bill concerning environment is introduced. Moreover, the only variable that has a significant influence on the dependent variable is partisan affiliation: not surprisingly, being member of the group new left increases significantly the probability of introducing a bill concerning environment in almost all the legislatures (the only exception is the XI legislature). Also the statistics for welfare offer a sometime contradictory picture. Gender is significant in the X, in the XII and in the XIV legislatures, but the sign of the relationship with the dependent variable is sometimes positive (X and XIV) and sometimes negative (XII). This means that in some legislatures women MPs put welfare among their top policy priorities, and in other not. Health and education models revealed themselves surprising. For health models, gender is not significant only in the XI and XII legislatures. In all the other legislatures, being a woman increases significantly the probability that a bill concerning health is introduced in the legislature. For models with education as dependent variable, gender is always significant: the Italian women MPs give higher priority than men also to this policy area. This finding, actually, may be interpreted as a reflex of a further way women MPs articulate their policy agenda and represent substantive interests of women. The Italian job market is characterized by an almost high level of gender segregation, with women strongly concentrated in public employment, and in primary and secondary school especially. Education is, therefore, a sector of the job market of particular interest for Italian women. Moreover, teachers have been one of the most represented professional category among the Italian women MPs, with a decrement occurred only in the most recent legislatures (Papavero 2006). Obviously, this policy area does not include only policies concerning school or university s personnel. However, the proportion of bills related to it is almost high within this group of bills. There are also other elements in the statistical models that suggest that public employment is a central issue when dealing with this policy area. First of all, the variable constituency of election significantly affect the dependent variable in all the legislatures but the last one (XV). This means that being elected in a constituency of South Italy, which is economically depressed and traditionally more dependent on government's intervention, increases significantly the probability that an MP introduces a bill concerning education. Moreover, the statistical models show that political affiliation to parliamentary groups that have stronger electoral roots in the South of the country has a significant impact on the dependent variable as well. 12

13 Summary and provisional conclusions In this paper, following a long tradition of studies on the subject, I tried to test whether women officeholders in the Italian parliament tend to substantively represent women. The statistical analysis of the data I pooled for six Italian legislatures has provided a further confirmation to the hypothesis that women representatives give higher priority than men in their policy agenda to those policies that are closest to their life experience, that is, women s rights, children and family. Contrary to other studies (Jones 1997), I found that women MPs in Italy prioritize also other traditional policy areas of interest for women, such as education and health care. I suggested that the characteristics of the Italian job market may mostly account for this result. As several studies highlighted especially for Anglo-Saxon and Northern European countries, also this study suggests that variations in the numbers of women legislators may have consequences on the legislative behavior not only of women but also of men representatives. The analysis made on the Italian legislators in this paper is obviously not conclusive, and further refinements are needed. However, at the present, the empirical evidence gathered for the Italian case suggests some interesting considerations and raises new questions on this point. Previous studies, following Kanter s theoretical framework, focused especially on one possible strategy women representatives adopt when they are very few in the legislature, that is their tendency to limiting their visibility, to adopt a low profile in the legislature, which means that they tend to limit their action for women s interests. As the number of women increases, this attitude tends to disappear, women display different values and priorities from men, and once a certain critical mass has been overcome, the differences between women and men's legislative behavior usually start to decrease. The Italian case suggests that also when women legislators are very few, they tend to give very high priority to women s issues. As the number of women increases, the policy agenda of women legislators tend to be more diversified and similar to that of men. The analysis of the data about the Italian legislators activity have also shown that in the Italian parliament also male representatives have changed over time their way to consider the issues traditionally of higher interest for women. We have seen that progressively (although very slowly), and with an acceleration in the last two legislatures, men have tended to converge with women legislators, giving more attention in their policy agenda to issues related with women s rights, children and family. Moreover, the number of bills concerning these issues introduced in the Italian parliament has progressively increased over time. Sue Thomas (1991) suggested in her work on US state legislators that such a new tendency could be interpreted as a by-product of the increasing attention that women s activity brought towards those issues as their number in the legislature increased. My conclusions are more cautious. This could be a possible interpretation, but we should be able to control whether those changes are also correlated with more profound social changes, in the family, in the workplaces, and more in general in the relationships between men and women. Moreover, as the data I used in this study are basically based on a classification of the bills introduced in the various legislatures only according to their subject matter, nothing can 13

14 be said about the substantive content of those bills, about what they propose to do. It is possible that the increasing attention to issues related to women s rights may be the signal of a major sensitiveness towards the discriminations and difficulties that still women face. However, it could also be the signal of a tendency to use more regulative power in order to rule aspects of private and public life that often are very divisive and controversial in the modern political communities, and which can be treated in more than one way. This caution in the interpretation of the data is suggested also by the findings about the role of political ideology in determining policy priorities in the Italian parliament. As we have seen, especially since the XII legislature, the members of Catholic parliamentary groups have shown a significant activism on issues concerning women, children and family. This is an interesting finding, which differentiates the case of the Italian national legislature from the cases most studied in the literature (Anglo-Saxon national and subnational legislatures, Scandinavian parliaments, Argentine parliament), where partisan affiliation, when considered, has not been found significantly affecting policy priorities. Controlling data for a number of legislature greater than that considered in this study, and using a comparative research design may shed more light on these differences. 14

15 References Bratton, K.A. and Haynie, K. (1999). Agenda Setting and Legislative Success in State Legislatures: The Effects of Gender and Race. The Journal of Politics, 61(3): Cotta, M., Mastropaolo, A., Verzichelli, L. (2000). Parliamentary Elite Transformations along the Discontinuous Road of Democratization: Italy In Best, H., and Cotta, M. (Eds). Parliamentary Representatives in Europe Legislative Recruitment and Careers in Eleven European Countries. Oxford: Oxford University Press. Guadagnini, M. (1987) Una rappresentanza limitata: le donne nel parlamento italiano dal 1948 ad oggi. Quaderni di Sociologia, 8: (1993). A Partitocrazia without Women: the Case of the Italian Party System. In Lovenduski. J. and Norris, P. (Eds). Gender and Party Politics. London: Sage Publications. Hine, D. (1993). Governing Italy. The Politics of Bargained Pluralism. Oxford: Oxford University Press. Kanter, R. (1977).Some Effects of Proportions on Group Life: Skewed Sex Ratios and Responses to Token Women. The American journal of Sociology, 82(5): Jones, M.P. (1997). Legislator Gender and Legislator Policy Priorities in the Argentine Chamber of Deputies and the United States House of Representatives. Policy Studies Journal, 25(4): Lovenduski, J. and Norris, P. (2004). Westminster Women: The Politics of Presence. Political Studies, Mansbridge, J. (2000). What Does a Representative Do? Descriptive Representation in Communicative Settings of Distrust, Uncrystallized Interests, and Historically Denigrated Status. In Kymlicka, W. and Norman, W. (Eds.). Citizenship in Diverse Societies. Oxford: Oxford University Press. Papavero, L. (2006). Revisiting the Contagion Theory :Female Parliamentary Recruitment in Italy and Spain ( ). Ph.D Dissertation thesis Reingold, B. (1992). Concepts of Representation Among Female and Male State Legislators. Legislative Studies Quarterly, XVII, 4: Sartori, G. (1970). The Typology of Party Systems. In Allardt, E., and Rokkan S. (Eds). Mass Politics: Studies in Political Sociology. New York: Free Press. (1974). Rivisitando il pluralismo polarizzato. In Cavazza, F.L. and Graubard, S.R. (Eds). Il caso italiano. Vol. 1. Milano: Garzanti. Thomas, S. (1991). The Impact of Women on State Legislative Policies. The Journal of Politics, 53(4): Thomas, S. and Welch, S. (1991). The Impact of Gender on Activities and Priorities of State Legislators. The Western Political Quarterly, 44(2):

16 Tremblay, M. (1998). Do Female MPs Substantively Represent Women? A Study of Legislative Behaviour in Canada's 35 th Parliament. Canadian Journal of Political Science/Revue canadienne de science politique, vol. 31, n. 3: Wängnerud, L. (2000). Testing the Politics of Presence: Women's Representation in The Swedish Riksdag. Scandinavian Political Studies, vol. 23, n. 1: Welch, S. (1985). Are Women More Liberal Than Men In The U.S. Congress? Legislative Study Quarterly, 10(1):

17 Appendix 1 Policy areas 1) women s rights, comprises bills dealing with issues such as equal opportunities in politics and in the work place, maternity leave, abortion, the regulation of the medically assisted fecundation, domestic and sexual violence, and so forth; 2) children and family, comprises bills dealing with all the issues related to family, marital status, parental leave, the protection of children from any kind of abuse and so forth; 3) health: comprises bills dealing with medical care and medical instruments standards, public health, medical care institutions and their organization. It comprises also bills related to the medical personnel and its carrier 4) education: comprises all issues dealing with school and university and the personnel working in these public sectors. 5) Welfare policies are considered in this paper pensions, measures against unemployment, specific provisions towards handicapped people etc. 6) environment comprises bills dealing with pollution and environmental quality, natural parks and preservation of animal and plants. Political families 1) the new left, which includes the Greens, the Italian Radical Party, some post-communist and post-socialist small parties. These are small parties that, although divided by some ideological differences, give great emphasis on civil rights, women s rights and environment, and are (or try to be) closer to social movements interested in those issues; 2) the traditional left, which is represented by the Communist Party until the XI legislature ( ), and afterwards by its heirs, the Democratic Party of the left/democratics of the left (PDS/DS), and by the Socialist Party until the XI legislature. In the statistical models for the X and XI legislatures I considered the two parties separately, basically because of the relevance of their political differences. First, the Socialist Party at that time had already abandoned the Marxist ideology, while the Communist party had not yet, although a process of revision was at work. Second, the Socialist Party since 1963 had been part, or external supporter, of the Italian coalition governments with the Christian Democratic party, while the Communist Party had always been at the opposition. After the turmoil of the 1994 general elections the Socialist Party disappeared, run over by the clean hands investigations, and what was left of it split into different parties of both the left and the right. So, from the XII legislature onwards the Democratic Party of the left, heir of the dead Communist Party, remained the only significant representative of this political family. 3) the moderates. This group (or political family) includes different parties in the legislatures of the first and of the second republic. In the X and XI legislatures it was formed by small parties that had been part of the coalition governments with the Christian Democratic party: the Social Democratic party and the Republican party (at the left of the DC) and the Liberal party (at the right of the DC). Experts of the Italian politics usually refer to them as secularized moderates, in order to distinguish them from the big catholic party of the center (the DC). Although some ideological differences, they basically had shared a pro-free market position and a more liberal approach to 17

18 such issues as divorce and abortion 8. After 1994 all these parties disappeared, following the same sort of the Socialist party. From the XIII legislature onwards, the moderate group includes some very small secularized new parties and the leftist wing of the former Christian democratic party, which from the XIV legislature collapsed into a new party, the Daisy (la Margherita). This is basically a liberal party with a strong component of non-conservative Catholics. 4) The Catholic center. This is a political area that during the first republic was occupied only by the DC, while after 1994 it has become quite crowded. After the split of the DC, this area is represented by those parties or groups, sometimes small but relevant from an electoral and political point of view, that emphasize their commitment to Christian-catholic values. They usually present themselves as conservatives on moral issues and traditional values, but supporters of welfare and of the intervention of the State in economy. Some of these parties or groups are actually built around some political leaders, local bosses that were second or third backbenchers in one of the many factions in the former DC. They comprise the party Christian-Democratic Center/Union of the Center (CCD, then UDC), which supported the governments of the centerright, the UDEUR leaded by Clemente Mastella, which supported the governments of center-left, and the Pact for Italy, a party formed in 1994 and headed by the promoter of some referenda that gave a fundamental impulse to some institutional reform of the Italian political system in the mid of the 1990s. 5) The Right. This is a political family that includes several parties. For the X and XI legislatures it is represented by the extreme-right, fascist party MSI (Italian Social Movement). Since 1994, this party, after the split of its right wing, was reformed and turned into a new post-fascist, conservative party, National Alliance (AN). However, the most important party in this family is Forza Italia, leaded by Silvio Berlusconi, who has been the prime minister of the Italian governments in different legislatures. This party is completely a new one in the Italian party system, although several members comes from the benches of the disappeared parties that governed Italy until It appeared immediately as a catch-all party, with an ideological fund quite lightened, able to keep together representatives of all those political and electoral areas that the disappearance of the old governmental parties left uncovered. The third party relevant in this family is the Northern League. When it emerged in the Italian party system, it was basically an anti-system, ethno-regionalist party, mainly interested to represent a cleavage that opposed the Northern Italian regions, with their economic productivity and their sense of civic-duty, to the central government, in their view wasteful and corrupted, and to the South, unable to produce economic welfare for the country and lacking of civic-mindedness. Since the very beginning supporter of the free-market and against taxes to finance national welfare state, as its importance in the Italian political system has increased, this party has started to emphasize its commitment with Christian identity, traditional values, although often it has found itself in conflict with the hierarchies of the Catholic Church. This 8 Indeed, the bill that in 1974 became law and made divorce legal in Italy was introduced by two MPs: the Socialist Loris Foruna and the Liberal Antonio Baslini. 18