Cracking the glass ceiling The representation of women and men in political and public decision making in the Council of Europe s member states Åsa Ekberg Fredell and Drude Dahlerup Department of Political Science Stockholm University Working Paper Series 2006:1 The Research Program on Gender Quotas Prepared for the Council of Europe May 2006.
Contents 1. Introduction...1 1.1 Aim and data collection...3 1.2 Main work in this field by the Council of Europe...3 1.3 Why balanced representation?...4 1.4 Historical background...4 2. Legislative power: women in parliaments...14 2.1 National parliaments...14 2.2 Regional parliaments...17 2.3 Local Councils...17 2.4 International Assemblies...18 3. Measures and practices for promoting women s political representation in legislative bodies...20 3.1 Electoral gender quotas: A global trend and maybe a future European trend 20 3.2 Legal gender quotas in Europe...23 3.3 Voluntary party quotas in Europe...25 3.4 Countries without quotas...28 3.5 Are gender quotas discrimination of men?...31 3.6 Quotas and plurality-majority systems a bad equation?...32 4. Executive power...34 4.1 Heads of state...34 4.2 National governments...35 5. Diplomatic service...39 6. Women in the Judiciary (High/Supreme and Constitutional Courts)...41 7. Challenges for increasing women s representation...46 7.1 Why women s under-representation?...46 7.2 A new global challenge for Europe...47 8. Conclusions and recommendations...50 8.1 The glass ceiling...50 8.2 Cracking the glass ceiling - recommendations...51 Bibliography...56
Charts Chart 1. Women in the national parliaments, s/l house, according to electoral system.9 Chart 2. Electoral systems in the Council of Europe s member states....10 Chart 3. Women and men in national parliaments, single / lower house...15 Chart 4. Women and men in national parliament, upper house...16 Chart 5. Women and men in regional parliament....17 Chart 6. Women and men in local councils....18 Chart 7. Head of state, percent...34 Chart 8. Head of Government, percent....34 Chart 9. Women and men in government....36 Chart 10. Women and men in diplomatic service, percent...39 Chart 11. High / supreme court judges, women and men, percent....42 Chart 12. Women and men in constitutional courts. Percent...44 Chart 13. The representatives, percent...50 Chart 14. The leaders, percent....50 Tables Table 1. Women in parliament, single/lower house, 1990-2005....5 Table 2. Women s enfranchisement and parliamentary representation...8 Table 3. Women in government, 1980-2000....11 Table 4. Women judges 1980-2000. Percent...12 Table 5. Countries in Europe with constitutional quota and/or election law quota regulation for candidates to national parliaments, lower/single house...23 Table 6. European countries with voluntary party quotas. Lower/single house...27 Table 7. European countries with no type of quotas for parliamentary elections to lower/single house....28 Table 8. Women s parliamentary representation...48 Table 9. The top of the world rank order of women in parliament...49
1. Introduction Two contrasting discourses on women s representation can be identified. According to the optimistic discourse, the representation of women in political and public decisions-making will gradually and constantly increase until gender balance is achieved. Thus, it is seen primarily as a matter of time until gender balance is reached. In contrast, the more pessimistic discourse rejects the idea of constant progress and points to historical examples of backlash in women s representation, like the situation in Russia and Central and Eastern Europe after the change. Further, in some of those countries where women s representation in recent decades has constantly increased and reached a high level of about 30 percent, the development seems to have come to a standstill. In countries like Denmark, Norway and Finland an invisible class ceiling seems to prevent further advancement of women in public life. Consequently, in Europe as well as in the rest of the world, active measures to change women s underrepresentation are on the agenda today. Women s access to decision-making arenas around the world is limited. As an example, women s representation world-wide in lower/single houses has increased from 3 percent in 1945 (Pintat 1998) to 16.8 percent in April 2006 (IPU 2006), i.e. an increase of less than 14 percentage points in 60 years. Globally, the present share of legislative seats, lower and upper houses combined, is 16.6 percent for women and 83.4 percent for men (IPU 2006). In recent years, however, many countries have started to take action to increase women s representation. One such measure involves different types of affirmative action, and today legal and/or voluntary party quotas have been adopted in a total of 97 countries around the world (IDEA and Stockholm University 2006). In 2003 the Council of Europe s Committee of Ministers made the recommendation that the representation of each sex in any decision-making body in political or public life should not fall below 40% (Council of Europe 2003:7). As this report will show, most of the institutions in the member states examined have a long way to go before realizing this ambition. In spite of European women s high level of education and extensive labour market participation, an invisible glass ceiling seems to prevent women s political empowerment. Gender balance in politics does not seem to occur by itself. 1
This report is entitled Cracking the Glass Ceiling. A glass ceiling is an analogy for the structural (and hence not legal) barrier that brings women s careers to a standstill. This concept is usually applied in research on women s (lack of) promotion and career opportunities. However, it is also an appropriate concept in the study of women s under-representation in elected political institutions. The glass ceiling represents the almost invisible structural barriers and indirect discrimination, which stops women from gaining access to some positions as well as becoming more than a minority in decision-making positions. The concept of glass ceiling first appeared about 20 years ago, most often ascribed to two journalists at the Wall Street Journal examining the lack of corporate women in high positions. In the last 20 years extensive research have made use of this metaphor to describe how structural barriers secure women s under-representation and men s numerical dominance in decisionmaking and in high positions. The concept has been used in a variety of situations to describe this vertical segregation. As an example, in the 1990 s the US Department of Labor Glass Ceiling Commission studied the barriers that keep both women and minorities from reaching higher positions in different workplaces, organizations and government (Cornell University 2006). A recent example is a bill in the House of Commons introduced in March this year seeking to remove the glass ceiling that keeps women from becoming bishops in Church of England (Ekklesia 060322). Almost all formal, legal barriers against women have now been removed. But various types of structural barriers still set limits on women s opportunities. Consequently, it is necessary to use active measures in order to reach the goal of gender balance (FWCW 1995; Dahlerup and Freidenvall 2005:29). Apart from the concept of glass ceiling, this report will apply some additional central concepts from research on women in politics and leadership positions: the time lag thesis, the law of increasing disproportion (the higher up, the fewer women), the acceptable minimum of women, the leaking pipeline, and the fast track versus the incremental track discourse and practice. Such concepts are helpful when describing and explaining women s under-representation. 2
1.1 Aim and data collection The aim of this report is to describe the current situation in the Council of Europe s 46 member states 1 as regards the representation of women and men in political and public decision making and to make recommendations in this regard. Areas covered in this study are the national and regional parliaments and the local councils as well as the national governments, the high/supreme and constitutional courts and the diplomatic service. The basis for the presentation and analyses is empirical material gathered in the form of the Council of Europe s Questionnaire on gender segregated data on the participation of women and men in political and public decision-making, answered by 36 of the 46 member states 2. The questionnaire was answered by the Council of Europe s member states during September 2005, and additional data was collected in the following months. The additional data has been gathered through official websites of the member states as well as Electionworld (2005), International IDEA & Stockholm University (2005), Inter-Parliamentary Union (2005) and Reynolds et al (2005). In cases of inconsistency between the questionnaire and the official websites, information from the official websites has been used. Data from the questionnaire about political representation and electoral systems have been checked against the websites mentioned above. 1.2 Main work in this field by the Council of Europe Previous recommendations from the Council of Europe in this matter are listed in Recommendation (2003)3 on balanced participation of women and men in political and public decision-making. In the same year, the Parliamentary Assembly adopted Resolution 1348 (2003) Gender-balanced representation in the Parliamentary Assembly, where the new rules of procedure include regulations regarding the composition of national delegations: National delegations should include the underrepresented sex at least in the same percentage as is present in their parliaments and in 1 As of November 2005: Albania, Andorra, Armenia, Austria, Azerbaijan, Belgium, Bosnia &Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Georgia, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Monaco, Netherlands, Norway, Poland, Portugal, Republic of Moldavia, Romania, Russian Federation, San Marino, Serbia and Montenegro, Slovakia, Slovenia, Spain, Sweden, Switzerland, TFYR of Macedonia, Turkey, Ukraine and UK. 2 The following member states did not answer the questionnaire: Albania, Andorra, Bulgaria, Georgia, Malta, Poland, Republic of Moldavia, Romania, Russian Federation and San Marino. Some of the member states that answered the questionnaire only answered a few of the questions. In general, the quality of the answers varied a great deal and, consequently, all data used were double checked with other sources whenever possible. In a few cases, however, problems derived from unclear formulations in the questionnaire. 3
any case one representative of each sex. In 2004 Resolution 1360 (2004) Contested Credentials of the parliamentary delegations of Ireland and Malta suspended the voting rights of the two countries in the Assembly due to non-compliance with the previously mentioned Rule 6.2. 1.3 Why balanced representation? There are a number of common arguments in favour of making efforts to raise women s share of political and public decision-making positions. Three frequent sets of arguments in this regard are connected to the areas of justice, interests and democracy. The justice argument proposes that a society does not attain real equality as long as women only have limited access to positions of power. Here, gender balance in politics is a goal in itself. The argument regarding interests claims that women s presence in legislatures is needed when making policies on the basis of women s interests. (Squires 2004:51) Two different positions are included in the interest argument. Either women are seen as a natural identity group, or as an interest group constructed via economic or other positions in society (Squires 1996:77-80). Arguments based on the democracy concept propose that a balanced share of power between women and men should vitalize democracy (Squires, 2004:51). Another democracy argument suggests that women s exclusion from political institutions is in itself a basis for questioning the legitimacy of democratic governance (Montgomery 2003:4). The interest argument as well as the democracy argument are based on equal representation as a means of achieving certain aims. 1.4 Historical background Table 1 shows the development of women s representation during the past 15 years. 4
Table 1. Women in parliament, single/lower house, 1990-2005. Member state Women in parliament (%) 1990 Women in parliament (%) 1995 Women in parliament (%) 2000 Women in parliament (%) 2005 Change 2005-1990 Albania 29 6 5 6-23 Andorra.. 4 7 29 +25 Armenia 36.. 3 5-31 Austria 12 24 27 33 +21 Azerbaijan.. 2 12 10 +8 Belgium 9 12 23 35 +26 Bosnia& Herzegovina 5 4 29 14 +9 Bulgaria 21 13 11 21 0 Croatia 6 6 8 21 +15 Cyprus 2 4 5 10 +8 Czech Republic 30 10 15 16-14 Denmark 31 33 37 37 +6 Estonia 6 13 18 20 +14 Finland 32 34 37 38 +6 France 7 6 11 12 +5 Georgia 6 6 7 10 +4 Germany 21 26 31 33 +12 Greece 7 6 6 13 +6 Hungary 21 11 8 9-12 Iceland 21 25 35 33 +12 Ireland 8 13 12 14 +6 Italy 13 15 11 7-6 Latvia 15 15 17 18 +3 Liechtenstein 4 8 4 24 +20 Lithuania 10 7 18 21 +11 Luxembourg 13 20 17 20 +7 Malta 3 2 9 9 +6 Monaco 11 6 22 21 +10 Netherlands 21 31 36 35 +14 Norway 36 39 36 37 +1 Poland 14 13 13 20 +6 Portugal 8 9 19 25 +17 Republic of Moldova 2 5 9 22 +20 Romania 34 4 7 11-23 Russian Federation 9 13 8 10 +1 San Marino 12 12 13 17 +5 Serbia & Montenegro.. 3 5 10 +7 Slovakia 12 15 13 15 +3 Slovenia 13 14 8 13 0 Spain 15 16 22 36 +21 Sweden 38 40 43 46 +8 Switzerland 14 18 23 27 +13 TFYR Macedonia 4 3 8 20 +16 Turkey 1 2 4 4 +3 Ukraine 3 4 8 5 +2 United Kingdom 6 10 18 20 +14 Average 14 13 16 20 +6 Note: Regarding the figures for 1990: Andorra held its first election under parliamentary democracy in 1993 (no figure available from that election). Azerbaijan held its first independent election in 1995. No figure available from the first election in Serbia & Montenegro. The figure for Latvia is from 1993; the figures for TFYR of Macedonia and Republic of Moldova are from 1994. Hence, the figures in the column to the right hence show the change between 1995 and 2005 for Andorra, Azerbaijan and Serbia 5
& Montenegro. Regarding the sources, it is unclear whether the figures come from elections or from other dates. Sources: 1990: Bosnia & Herzegovina, Croatia, Estonia, Georgia, Germany, Latvia, Lithuania, Republic of Moldova, Russian Federation, Serbia and Montenegro, Slovakia, Slovenia, TFYR of Macedonia, Ukraine: Montgomery (2003). The remaining countries: United Nation Statistics Division (2005b). 1995: United Nations Statistics Division (2005a). The figures represent 1 July 1995. 2000: United Nation Statistics Division (2005b). 2005: Council of Europe (2005), official websites of the member states and IPU (2005). One general indication from the average figures in Table 1 is that the process of increasing women s legislative representation in the past 15 years has been rather slow. The average percentages in 1990 and in 2005 are quite similar, and less than 20 percent of the seats are held by women today. But, as Table 1 also shows, there are a few remarkable changes in women s representation in some of the member states that are not apparent in the average figures. Since 1990 the largest increases have occurred in Western European states, such as Andorra, Austria, Belgium, Liechtenstein and Spain, where women s share of the seats in the parliaments has risen about 20 percentage points in 15 years. On the other hand, in some Eastern European states, such as Albania, Armenia and Romania, women s share of the legislative seats has declined by more than 20 percentage points during the same period of time. Previous research on the critical turn from communist to democratic regimes in Eastern Europe has shown this to be a clear pattern; the decline in women s representation is massive in most of these countries (Matland and Montgomery 2003). Kathleen A. Montgomery labels the transitionary period from communist to democratic regimes democratic consolidation (Montgomery 2003:2) and refuses to accept some researchers definition of the decline in women s representation as a post-communist phenomenon. Montgomery s argument is that the low share of seats for women is not isolated to post-communist areas, since very few of the Western countries in the world reach the level of 30 percent in relation to women s representation in parliament (2003:5). For some of the countries in Table 1, the major changes in women s representation occurred before 1990. Denmark, Finland, Norway and Sweden had already reached the 30 percent level and have continued to increase these figures during the following 15 years (Bergqvist 1999). Iceland and Netherlands had reached the 20 percent level in 1990. In Western Europe, the political institutions have historically been more open to women in the protestant North than in the Catholic South. While women gained 6
suffrage around the First World War in the North, women did not become enfranchised until after the Second World War in the South, Liechtenstein being the last country in Europe to give women the vote (1984). It is possible to demonstrate a close link between the level of women s representation and the number of years women have been enfranchised. However, the introduction of new active measures, such as quota provisions, are challenging our traditional theoretical explanations of what factors lead to a high representation of women. Due to a recent move, Spain with 36 percent women in parliament is about to surpass the countries in the North (Threlfall 2005). The remarkably high level of women s representation in the Nordic countries is usually explained by a successful combination of structural factors (extended welfare states, social democracy, high level of education and labour market participation for women, secularism) and strong interventions by women s organisations within the political parties and by the autonomous feminist movements (Freidenvall et al 2006). However, the previous constant increase in women s parliamentary representation in Norway, Denmark, Finland and Iceland seems to have come to a halt at the one third level. In the last elections in Denmark and Iceland, women s representation even decreased. Perhaps the prevailing time lag discourse explains the lack of public outcry following the actual drop, contrary to the extended debate in Sweden in 1991 after the first decrease in women s representation in Sweden since the 1920s, a debate that resulted in Sweden surpassing the 40 percent threshold in the following 1994 election (Freidenvall 2003). The time factor While Denmark, Finland and Norway seem to have hit a high level glass ceiling at about 30 percent, other countries are stuck at a much lower level, a low level glass ceiling, defined as constantly below 15 percent: Cyprus, Greece, France, Georgia, Ireland, Malta, Slovenia, Turkey and Ukraine. Are these differences related to the length of time that women have been enfranchised in the various countries? 7
Table 2. Women s enfranchisement and parliamentary representation. Country The year women received the right to vote The year women received the right to stand for election The first woman elected or appointed Women in parliament, 1 Sept 2005, % 1900-1919 Finland 1906 1906 1907 E 38.0 Norway 1907 1907, 1913 1911 E 37.0 Denmark 1915 1915 1918 E 36.9 Iceland 1915, 1920 1915, 1920 1922 E 33.3 Austria 1918 1918 1919 E 33.0 Estonia 1918 1918 1919 E 19.8 Georgia 1918, 1921 1918, 1921 1992 E (current) 9.5 Germany 1918 1918 1919 E 32.8 Hungary 1918 1918 1920 E 9.1 Ireland 1918, 1928 1918, 1928 1918 E 13.9 Latvia 1918 1918 18.0 Poland 1918 1918 1919 E 20.4 Russian Fed. 1918 1918 1993 E (current) 9.8 United Kingdom 1918, 1928 1918, 1928 1918 E 19.7 Belgium 1919, 1948 1921, 1948 1921 A 34.7 Luxembourg 1919 1919 1919 E 20.0 Netherlands 1919 1917 1918 E 34.7 Sweden 1919, 1921 1919, 1921 1921 E 46.4 Ukraine 1919 1919 1990 E (current) 4.7 Average: 24.8 1920-1939 Albania 1920 1920 1945 E 5.7 Czech Republic 1920 1920 1992 E (current) 16.0 Slovakia 1920 1920 1992 E (current) 14.7 Armenia 1921 1921 1990 E (current) 5.3 Azerbaijan 1921 1921 1990 E (current) 10.4 Lithuania 1921 1921 1920 A 20.6 Romania 1929, 1946 1929, 1946 1946 E 11.2 Turkey 1930, 1934 1930, 1934 1935 A 4.4 Portugal 1931, 1976 1931, 1976 1934 E 25.2 Spain 1931 1931 1931 E 36.0 Average: 15.0 1940-1959 Bulgaria 1944 1944 1945 E 21.3 France 1944 1944 1945 E 12.3 Croatia 1945 1945 1992 E (current) 21.3 Italy 1945 1945 1946 E 6.8 Slovenia 1945 1945 1992 E (current) 13.3 Bosnia & Herz. 1946 1946 1990 E (current) 14.3 Macedonia 1946 1946 1990 E (current) 20.0 Malta 1947 1947 1966 E 9.2 Greece 1949, 1952 1949, 1952 1952 E 13.0 San Marino 1959 16.7 Average: 14.8 1960- Cyprus 1960 1960 1963 e 10.0 Andorra 1970 1973 28.6 Switzerland 1971 1971 1971 e 26.5 Moldova 1978 1978 1990 e 21.8 Liechtenstein 1984 24.0 Average: 22.2 Notes: Current means the first year a woman was elected in the current parliamentary system. When two years are presented, the first included some restrictions in voting rights. 8
Sources regarding women s rights to vote, women s rights to stand for election and the first woman elected/appointed: San Marino, Andorra and Liechtenstein: IPU (2006). The rest of the countries: UNDP (2005). Source for Women in parliament 1 sept 2005: Council of Europe (2005). Table 2 provides an overview of women s enfranchisement in Europe. Comparing women s present representation in countries of the first wave of enfranchisement, 1990-1919, to the second, 1920-1939, and the third, 1940-1959, shows that women on the average are best represented in countries with early suffrage. The countries that deviate from this pattern are mostly, though not entirely, countries that have gone through major historical changes in the political system. When reading Table 2, it should be noted that it is not possible to isolate the effect of the length of women s franchise from other important factors that influence women s representation. This also becomes obvious when looking at the last countries to give women the right to vote and stand for election, 1960-. These countries have a relatively high average representation of women, 22.2 percent, and here women started in a totally different gender regime than those around the First World War. Electoral systems and women s representation Research has shown that a proportional electoral system has a better chance of securing women s representation than other systems (Matland 2002). International IDEA:s figures regarding women s representation on the basis of majoritarian and proportional systems shows an increasing gap in PR:s favour since the 1970:s (Matland 2002). 3 This general finding is confirmed for Europe by the figures in Chart 1, which shows that women s representation is by far the highest in Council of Europe member states which use a PR electoral system. Chart 1. Women in the national parliaments, s/l house, according to electoral system. 25% 20% 15% 10% 5% 0% PR P/M Mixed 3 Norris (2004:188) has, however, shown that this correlation holds true mainly for industrialized countries, not for developing countries. 9
Note: Figures as of 1 September 2005. Key to electoral systems: PR = Proportional Representation systems; P/M = Plurality/Majority systems. Mixed = Mixed PR and P/M systems. Some of the systems in the P/M family do not demand that the winner obtain a majority of the votes, only more votes than any other candidate (e.g. the FPTP system). Source: Council of Europe (2005), official websites, Reynolds et al (2005). Almost 70 percent of the Council of Europe s member states use PR electoral systems, see Chart 2. Chart 2. Electoral systems in the Council of Europe s member states. P/M Mixed4% 26% PR 70% Sources: Council of Europe (2005), official websites, and Reynolds et al (2005). The advantage of PR is confirmed when comparing the ranking of the member states regarding women s representation in lower / single houses (see Chart 1). Of the top 10 countries in the ranking, 9 have PR systems, and these are the first 9 on the list. Of the 10 member states on the bottom of the ranking, 7 have mixed systems. Thus PR systems seem to represent a relative advantage for women. One advantage of PR systems is high district magnitude, i.e. the number of seats in the districts. This leads to high party magnitude, which is the anticipated number of elected per party in the districts. In plurality-majoritarian systems the district magnitude is low, often just one seat, and the party has to choose just one candidate to represent the party. (Matland 2002:6ff) This means that women and men compete against each other for the post, and since women are considered an electoral risk they are not likely to be chosen (Norris and Lovenduski 1995:195). When the district magnitude is higher and, hence, also party magnitude, the parties can balance their lists, with less risk of losing votes. A third advantage of PR systems is contagion. Parties that compete over votes tend to adopt policies from other parties, e.g. policies concerning the number of women in top positions on party lists. (Matland 2002:6ff) Some PR systems have further advantages, such as closed lists: no preferential voting is available that can stop women in the top positions from being elected. (Krook 2005:7, Matland 2002:8) But in some Scandinavian elections voters do vote for women, and here the voters can be more progressive than the parties. 10
Sainsbury (1993) reformulate the effects of PR systems in relation to increased representation for women. It is not the PR systems as such that increase women s representation, but they are more open to change: Because of its multiple numbers of candidates, the proportional representation system is less resistant to change than the plurality-majority system, where each party only has one candidate (Freidenvall et al 2006:59). The choice of PR as the electoral system is, however, just one factor, though different electoral systems do present different opportunities for interventions in order to enhance women s political representation. The difficulties in applying any gender quota system in a plurality/majority electoral system (see below) may in the future lead to even bigger differences between women s representation in the two systems. In the future, the dividing line in Europe in terms of the level of women s representation will probably not follow any North/South or East/West line but will rather follow from variations in active measures to identify and overcome the barriers that prevent women s equal representation with men. Table 3. Women in government, 1980-2000. Member state Women in gov. (%) 1990 Women in gov. (%) 1995 Women in gov. (%) 2000 Change 2000-1990 Armenia --- 0 0 0 Austria 12 15 36 +24 Croatia 0 10 11 +11 Cyprus 0 9 0 0 Czech Republic --- --- 0 --- Denmark 21 35 45 +24 Estonia --- 13 13 --- Finland 28 39 39 +11 Georgia --- --- 11 --- Iceland 9 10 33 +24 Ireland 7 13 20 +13 Latvia 0 7 7 +7 Lithuania 0 5 8 +8 Netherlands 14 29 29 +15 Norway 47 42 42-5 Republic of Moldova --- --- 7 --- Romania --- --- 24 --- Slovenia 7 11 20 +13 Spain 11 19 12 +1 Sweden 36 50 50 +14 Turkey 3 6 0-3 United Kingdom 8 10 29 +21 Average 10 15 20 +10 Note: --- No figures available. The figure for Romania is from 2001. Source: United Nations Economic Commission for Europe (2005). The figures for Denmark and Sweden in 1980, 1990 and 1995: Bergqvist (1999). The figure for Denmark in 2000: Council of Europe (2002). Data not available for all countries. 11
Women s share of the executive power in the form of government posts in the 20 member states, see Table 3, has seen a small rise in the years 1980-2000. The average figure for women s representation in minister posts has increased during the years mentioned, but only by 10 percentage points in total. For some of the countries, the 20 and 30 percent levels had already been reached in 1990. One big change, not visible in this table, occurred in Norway between 1980 and 1990, where women s share of the minister posts increased from 12 percent to 47 percent in ten years (United Nations Economic Commission for Europe 2005). Major changes also occurred in the rest of the Nordic countries, as well as in Austria, Netherlands and United Kingdom. Austria, Iceland and UK all saw increased representation for women of about 20 percentage points between the years of 1995 and 2000. Some other countries have seen both a rise and a decline of women in executive power, for example Cyprus and Spain. The present situation will be dealt with in Section 4. Table 4. Women judges 1980-2000. Percent. Member state Women judges (%) 1980 Women judges (%) 1990 Women judges (%) 1995 Women judges (%) 2000 Armenia 19 19 13 21 Austria --- --- 19 23 Croatia 31 45 53 61 Cyprus 0 5 15 21 Czech Republic 47 56 62 63 Estonia --- --- --- 62 Finland 28 46 51 --- France --- --- --- 43 Georgia --- --- 15 34 Germany 14 19 26 28 Hungary --- --- 70 68 Iceland --- --- 27 27 Ireland --- --- 13 16 Lithuania --- 26 45 51 Netherlands --- --- 34 41 Norway --- --- --- 27 Republic of Moldova 16 13 13 30 Romania --- --- 63 69 Slovenia 45 55 62 68 Spain --- 27 31 36 United Kingdom --- --- --- 16 Average 25 31 36 40 Note: --- No figures available. The figures for Norway and United Kingdom are from 2001. Source: United Nations Economic Commission for Europe (2005) For some of the member states, figures for the previous distribution of judicial power are available. Table 4 shows the average distribution among judges between women and men. In some of these 21 member states, women s share has been stable at around 12
20 percent. But in Croatia, Czech Republic, Estonia, Hungary, Lithuania, Romania and Slovenia, women s share has been lifted to between 51 and 69 percent during these years. The average figures show a modest increase between each given year, but they also demonstrate that a 40-60 distribution in these occupations is possible. To sum up: Women are still strongly under-represented in political and public life in Europe. The last decades have, however, witnessed some progress, even if the speed of the increase has been very modest and some backlashes have occurred. 13
Analyses of the current situation 2. Legislative power: women in parliaments 2.1 National parliaments The share of seats in the lower/single houses of national parliaments today is, as evident from Chart 3, still unbalanced between women and men. In the Council of Europe s member states, women s average share of the legislative seats is 19.8 percent. In other words, of the total seats in lower/single houses in all member states, men hold more than 80 percent. Women reach 30 percent in only ten of these 46 states, and the Council of Europe s target of 40 percent women has only been attained in one state. 14
Chart 3. Women and men in national parliaments, single / lower house. Sweden Finland Norway Denmark Spain Netherlands Belgium Iceland Austria Germany Andorra Switzerland Portugal Liechtenstein Rep of Moldova Bulgaria Croatia Monaco Lithuania Poland Luxembourg TFYR of Macedonia Estonia United Kingdom Latvia San Marino Czech Republic Slovakia Bosnia & Ireland Slovenia Greece France Romania Azerbaijan Serbia & Montenegro Cyprus Russian Federation Georgia Malta Hungary Italy Albania Armenia Ukraine Turkey 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Women Men Sources: The figures for Albania, Andorra, Bulgaria, Georgia, Malta, Poland, Republic of Moldavia, Romania, Russian Federation and San Marino: IPU (2005) and official websites of the member states. The remaining figures: Council of Europe (2005). 16 of the Council of Europe s member states have bicameral parliaments. In relation to the lower houses presented previously, there is an even more distorted balance 15
between women and men in the upper houses. The average share of seats for women is 17.3 percent in upper houses. Chart 4. Women and men in national parliament, upper house. Belgium Netherlands Austria Spain Switzerland Germany United Kingdom France Ireland Poland Czech Republic Romania Slovenia Italy Bosnia & Herzegovina Russian Federation 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Women Men Sources: The figures for Germany, United Kingdom, Poland, Romania, Slovenia and Russian Federation: IPU (2005) and official websites of the member states. The rest of the figures: Council of Europe (2005). The 40 percent target, set by Council of Europe, has not been reached by any of the member states, and in only two of these states do women constitute 30 percent of the members. Some of the members in upper houses are appointed, like in the Belgian case. These nominations could have been used as a counterbalance to the elections in relation to women s representation, but the nominations in Belgium s upper house still do not provide women with an equal share of the seats. In the most recent composition of the upper house, 34 of the 74 members were appointed; 10 of these were women and 24 were men. 16
2.2 Regional parliaments The number of regional parliaments in each state ranges from 1 to 26, with a total of 103. In these parliaments women hold 2162 of the 7814 seats, which is equivalent to 27.7 percent. In the top three countries, women have obtained 30 percent of the seats, but the 40 percent goal set by Council of Europe is not met by any of the states (see Chart 5). Azerbaijan and Portugal have the lowest number of women in their regional assemblies. Chart 5. Women and men in regional parliament. Spain Germany Belgium Austria Switzerland Bosnia & Herzegovina Serbia and Montenegro Italy Portugal Azerbaijan 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Women Men Source: Council of Europe (2005). 2.3 Local Councils The average figure for women in local government is 24.7 percent, but the situation in the member states varies greatly, as seen in Chart 6. On the one hand, nearly 80 percent of the Hungarian municipality councillors are women. On the other hand, in 11 of the 27 countries, more than 80 percent of the municipal councillors are men. Except for the extraordinary situation in Hungary, only two other member states reach the Council of Europe s target of 40 percent women (see Chart 6). A comparison of women s representation in national and local assemblies (see Chart 3 and 6) shows that women in general do better at the local level, even if there are exceptions to this, as in the cases of Denmark, Belgium and Croatia. 17
Chart 6. Women and men in local councils. Hungary Sweden Latvia Finland Norway Monaco Iceland Estonia Liechtenstein Denmark Belgium Spain Germany Serbia & Lithuania Cyprus Ireland Luxembourg Italy Bosnia & Netherlands Portugal Slovenia Czech Croatia Armenia Azerbaijan 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Women Men Source: Council of Europe (2005). Data not available for all countries. 2.4 International Assemblies On the supra-national level, the European Parliament has held elections since 1979. The MEP:s are elected for a five-year period, and six elections have been held during the 26 years. Before the first election in 1979, the non-elected parliament consisted of 6 percent women and 94 percent men (European Parliament 2005). Over the 6 elections between 1979 and 2004, women s representation has increased from 16 percent to 28 percent (European Parliament 2004, 2005). Over the last three elections, women s representation in the European Parliament seems to have stabilized at approximately one third of the seats. The top result of 30 percent in 1999 was followed by a small decrease in women s representation in the election of 2004 (European Parliament 1999, 2004). It is, however, difficult to make comparisons between these election results, since the number of member states has changed 18
between each election. Because of the inclusion of 10 new member states, serious concern was expressed prior to the 2004 election. There was a general fear that women s representation would drop. That the result was only this minor decline may be due to the campaigns to prevent this from happening, as well as the fact that women s representation in general is not lower in the Eastern than in the Western part of Europe. In this election to the European parliament, Slovenia and Italy for the first time introduced gender quotas by law. The Council of Europe Parliamentary Assembly consists of representatives from each of the current member states, with a total of 630 Assembly members. The size of the country determines its number of representatives and, hence, its number of votes. The delegations are appointed by the parliament in each member state and they consist of representatives of the national political parties. The number of representatives in the national delegations currently ranges from two to eighteen. In 2003 the Council of Europe Parliamentary Assembly adopted the Resolution 1348 (2003) Gender-balanced representation in the Parliamentary Assembly establishing that [n]ational delegations should include the under-represented sex at least in the same percentage as is present in their parliaments and in any case one representative of each sex. As Ireland and Malta did not comply with this rule in 2004, Resolution 1360 (2004) Contested Credentials of the parliamentary delegations of Ireland and Malta suspended their voting rights in the Assembly. The Parliamentary Assembly can, therefore, be said to use a form of active measure for the national delegations, including implementing sanctions for non-compliance. 19
3. Measures and practices for promoting women s political representation in legislative bodies Today, many international bodies recommend that active measures be applied in order to reach gender balance in politics. The target of such recommendations is the member states and their political parties as well as the international organizations themselves. Such recommendations have rendered legitimately to the claims for more women in public life made by women s organizations and other NGOs. As the focus is increasing the quantitative representation of women in political institutions, electoral gender quotas are increasingly becoming an option. 3.1 Electoral gender quotas: A global trend and maybe a future European trend Electoral gender quotas are being introduced all over the world today with amazing speed. Although highly controversial, the use of gender quotas in public elections is a new global trend. During just the last decade and a half, slightly more than 40 countries in the world introduced legal quotas, e.g. quota rules inscribed in constitutional or electoral law. In more than 50 other countries major political parties have introduced gender quotas for their own list for public elections, e.g. voluntary party quotas. This is an amazing new development, which challenges our previous theories that improvement in women s education and women s access to the labour market is a prerequisite for an increase in women s political representation. In this section we will look at the extent to which electoral gender quotas have been introduced in Europe, and what quota type is the preferred in a European context. Quotas represent a challenge to the time lag discourse, so well known in Europe. According to the time lag thesis, women s under-representation will correct itself as time goes by, in connection with the social and economic development of society and, consequently, no specific or only very modest measures are needed in this regard. But, as Mona Lena Krook argues, women s representation in some rich countries remains well below that achieved in many poor countries, suggesting that political measures to increase women s representation appear even in absence of previously assumed social and economic prerequisites (Krook 2005:3). Today, we are witnessing historical leaps in women s representation, most often in post-conflict societies and in countries in transition to democracy. The 20
Scandinavian countries previous world record in women s representation is being challenged by South Africa, Costa Rica, Mozambique, Argentina not to speak of Rwanda, which now has the highest share of women in parliament in the entire world, 48.8 percent. Gender quotas are part of the explanation behind the exceptional historical leaps in women s representation in the countries mentioned. Where does Europe stand in this development? Defining electoral gender quotas Electoral gender quotas are here defined as written regulations, in law or in party statutes, which establish a fixed percentage or number for the representation of women for public election. Soft quotas may be used about less strict rules, such as targets or recommendations. In this section we focus on formally decided and written fixed quota provisions in law or party statutes. Quotas for internal party structures are also important but are not dealt with in the following. There is, however, some confusion as to what constitutes different quota regimes. In the book Women, Quotas and Politics we distinguish between two separate dimensions in the definition of quota systems: The first dimension covers the question of who has mandated the quota system, while the second dimension indicates what part of the selection and nomination process is targeted by the quota. As for the mandating, legal gender quotas and voluntary party quotas are the two main types. Legal quotas are mandated either by the constitution (as in France, Burkina Faso, Nepal, the Philippines and Uganda) or by the electoral law (as in many parts of Latin America as well as, for example, in Belgium, Bosnia Herzegovina, Serbia and Sudan). But quotas may also be decided on voluntarily by political parties themselves, voluntary party quotas. In some countries, including Sweden, Norway, Italy, Germany and outside Europe, Argentina, Bolivia, Ecuador, a number of political parties have introduced quotas. In many others, though, only one or two parties have opted to use quotas such as Spain. However, if the leading party in a country uses a quota, such as Spain or the ANC in South Africa, this may have a significant impact on the overall rate of female representation. Yet most of the world s political parties do not employ any kind of quota at all. Concerning the second dimension, quotas may target the first stage of the selection process, the stage of finding aspirants, e.g. those willing to be considered for nomination, either through a primary or by a nominating committee or other parts of 21
the party organization. Gender quotas at this stage are rules that require a certain number or percentage of women or of either sex be represented in the pool of candidates that are up for discussion, such as the controversial women s short lists in Great Britain. The second stage is the actual nomination of candidates to be placed on the ballot by the party. This frequently used quota system involves a rule (legal or voluntary) being established according to which, for instance, 20, 30, 40 or even 50 percent of the candidates must be women. This may also be formulated in a genderneutral way, stating that no sex should have less than 40 percent and no more than 60. At the third stage, those elected, we find quotas as reserved seats. Here it is decided that a certain percentage or number of the members of a parliament or local council must be women. Increasingly, gender quotas are being introduced using reserved seat systems, although not in Europe. The difference between legal quotas and voluntary party quotas should not be exaggerated. Within both systems we find effective as well as purely symbolic cases. However, legal quota systems may include some legal sanctions for noncompliance, which are absent when quotas are voluntary. In their investigation of quotas in the Balkan region, Milica Antic and Sonja Lokar conclude that women s organizations had to start demanding quotas by law, since during the first period of voluntary party quotas, initiated by the parties on the left, the parties were not willing to respect the quota rules of their own statutes (Antic and Lokar 2006) The preferred gender quota types in Europe By combining the two above mentioned dimensions, firstly, the question of mandating and, secondly, the question of where in the nomination process quotas are placed, one can identify different types of gender quotas in use in various regions of the world. That different regions in the world prefer different gender quota systems is no doubt a consequence of variations in electoral systems and political systems in general (Dahlerup 2006, figure 14.1, p. 294, see also IDEA and Stockholm University 2006). Whereas legal reserved seat systems are in use in countries with gender quotas in the Arab world, in Africa and in South Asia, this quota type is today not in use in Europe. Latin America is the leading region when it comes to using legal candidate quotas. Where does Europe stand? 22
3.2 Legal gender quotas in Europe As Table 5 shows, legal candidate quotas are used by only six European countries. Only six of the world s more than 40 countries with legal quotas are found in Europe. The post-conflict region of the Balkans represents three of these six European cases. It could be added that even the United Nations Mission in Kosovo (UNMIK) has operated a gender quota system (33%) since 2001 for its regional assembly. Table 5. Countries in Europe with constitutional quota and/or election law quota regulation for candidates to national parliaments, lower/single house. Country Elect. System Year of Intro Percent Quota % Women in Parl. (1 Sept 2005) Armenia Parallel 1999 5% 5.3 Belgium List PR 1994, 2002 50% 34.7 Bosnia-Herz. List PR 1999 33% 14.3 France Two Rounds 1999, 2000 50% 12.3 Macedonia List PR 2002 30% 20,0 S&M: Serbia List PR 2004 30% 11.3 Note: Year of introduction is the year the law or constitution was passed, not the year of the first election after the passage of the quota regulations. The figure for Serbia is from the last election (2003). For the 2004 election to the European Parliament, Slovenia and Italy introduced legal quotas. Greece has legal quotas only for the local level (Leyenaar 2004:227). Italy operated a quota law for national, regional and local levels from 1993 to 1995. Source: Council of Europe (2005); Dahlerup and Freidenvall (2005, updated); Antic and Lokar (2006). The quotas regulations range from 5 to 50 percent of the electoral candidates. Even so, the only one of these member states that actually reaches the legally prescribed quota is Armenia, the country with the lowest limit (5 percent). Women s share of the seats in the Armenian national parliament is 5.3 percent, just above the legal quota prescription. Even if such a low quota requirement might pave the way for the first representation of women in a country, there is a risk that this becomes not just a minimum but also a maximum limit, a low level glass ceiling for women s representation in the future. Table 5 reveals that a quota provision of, say, 30 percent does not automatically result in one third women being elected. Women s representation in these member states varies from 5.3 to 34.7 percent, and the average share of parliamentary seats is 16.3 percent. It is thus important to look at some of the critical factors for a successful quota system. 23
Additional rules: Sanctions for non-compliance and rank order Research on legal gender quotas has pointed to the importance of having sanctions in case of non-compliance in the law. Otherwise, quotas may be just a symbolic gesture. A crucial issue has also proven to be whether there are any rules concerning the rank order on the party lists. A requirement of, say, 40 percent may not result in any women being elected, if all women candidates are placed at the bottom of the list. The crucial question is whether the women nominated are placed in a position with a real chance of election. Five of the six countries mentioned (all but Armenia) have legally prescribed sanctions for non-compliance, and there are also rank-order rules present in four of these six states. Yet, it is apparent from Table 5 that something is missing or going wrong in the implementation processes. One possible explanation is that the sanctions are not implemented or that they are too timid. The financial penalty in the case of France at the national level was evidently not a problem for the parties, and women s representation only increased from 10.9 to 12.3 percent. In contrast, a big change occurred at the municipal level in France, where women s representation rose as much as 22 percentage points (from 25 percent to 47 percent), almost reaching the parity target. The reason for this was that the electoral commission in local elections has the authority, and used it, to reject party lists that did not comply with the law. Consequently, the political parties searched seriously for women candidates, which is in fact the raison d être of the quota system! (Sineau 2002; Norris 2004:196, Krook et al 2006). In the case of the Republic of Macedonia, a PR system was introduced before the national elections in 2002 and was one of the factors behind the increased share of seats for women, along with stern lobbying action (Antic and Lokar 2006:156). The fact that the 30 percent quota law has still not been met, despite clear regulations regarding sanctions for non-compliance, is explained by the lack of rank order rules, according to Antic and Lokar (2006:156) Women s share of the seats in the current Macedonian parliament is 20 percent, in contrast to the 30 percent quota law. Compared to France, Belgium s quota regulation has been successful. As a consequence of a fruitful interplay between the voluntary party quotas adopted during the last decade and ever more radical legislation, Belgium has now move up among the top of the world rank order in terms of women s parliamentary representation (Meier 2004). Today Belgium matches countries like Denmark, Norway and Finland 24