The Effect of Electoral Systems on
Changing a country's electoral system often represents a far more realistic goal to work towards than dramatically changing the culture's view of women.
Political scholars and women strongly emphasize the effect that electoral systems have on women's representation for several reasons. First, the impact of electoral systems is quite dramatic. As can be seen in Table 5 and Figure 2, the differences in women's representation across electoral systems are not trivial; they are substantial. Just as important is the fact that electoral systems can be, and regularly are, changed. Compared to the cultural status of women in society or a country's development level, electoral rules are far more malleable. Changing the electoral system often represents a far more realistic goal to work towards than dramatically changing the culture's view of women.
TABLE 5. Percent of Women MPs Across 24 National Legislatures 1945 - 1998|
Majoritarian (SMD) versus Proportional Representation (MMD) Systems
Majoritarian or Single-Member District Systems (SMD):
Australia, Canada, France (1960 and beyond), Japan, New Zealand (19451990), United Kingdom, and United States.
* Israel did not exist, and West Germany did not hold elections
in 1945. They are therefore not included in the 1945 numbers. They are all included for all years following 1945.
** Greece, Portugal and Spain became democratic in the 1970s and are therefore only included in the 1980, 1990 and 1998 calculations.
Proportional Representation or Multimember
District Systems (MMD):
Austria, Belgium, Denmark, Finland, France
(1945 and 1950), Greece **, Iceland, Ireland, Israel *, Italy, Luxembourg, Netherlands, New Zealand (1998 only), Norway, Portugal **, Spain **, Sweden, Switzerland and Germany (West Germany * prior to 1990).
© INTERNATIONAL IDEA
Table 5 and Figure 2 present data for 24 established democracies over the post World War II period. They reveal that women have always had a slight advantage in proportional representation (PR) systems. Until 1970, this advantage was quite small: there is only a couple percent difference in women's representation in countries with majoritarian or single-member district systems, versus countries with proportional representation or multi-member district systems. In the 1970s, 1980s, and 1990s, however, there is a dramatic increase in women's representation in PR systems, while only modest gains in majoritarian systems. 13
Figure 2. Percentages of Women in Parliament
Majoritarian vs. PR Systems|
© INTERNATIONAL IDEA
Different electoral systems lead to different outcomes. Throughout the developed world in the 1960s and 1970s we saw a wave of what was called "second generation feminism" women demanding equal rights on a whole array of issues, among them greater representation in political bodies. In countries with PR systems, women were able to translate those demands into greater representation. In majoritarian systems, on the other hand, the same demands were made but they were largely unsuccessful or only very modestly successful.
Advantages of PR Systems
The obvious question is why? Why should countries with proportional representation electoral systems show such a strong increase in representation and majoritarian systems show such a modest effect? There are a number of explanations. First, proportional representation systems have consistently higher district magnitudes, which lead to higher party magnitudes. District magnitude is the number of seats per district; party magnitude is the number of seats a party wins in a district. Party and district magnitudes are important because they affect party strategy when choosing candidates. The party gatekeepers, who must consider which aspirants to choose as candidates, will have a different set of concerns and incentives depending upon the electoral system.
When district magnitude is one, as it is in almost all majoritarian systems, the party can win, at most, one seat in a district. By definition, the party has no chance to balance the party ticket. Because of the strictly zero sum nature of nominating decisions in single-member districts, female candidates must compete directly against men; and often when nominating a woman a party must explicitly deny the aspirations of a man in the same district. When district magnitude increases, the chances that a party will win several seats in the district increase. When a party expects to win several seats, parties are much more conscious of trying to balance their tickets. Gatekeepers will divide winning slots on the party list among various internal party interests.
There are several reasons for this balancing process.14 First, party gatekeepers see balance as a way of attracting voters. Rather than having to look for a single candidate who can appeal to a broad range of voters, party gatekeepers think in terms of different candidates appealing to specific subsectors of voters. Candidates with ties to different groups and different sectors of society may help attract voters to their party. A woman candidate can be seen as a benefit to the party by attracting voters, without requiring powerful intra-party interests represented by men to step aside, as would be required in a majoritarian system. Conversely, failing to provide some balance, i.e., nominating only men, could have the undesirable effect of driving voters away. A second reason for balancing is that within the party balancing the ticket is often seen as a matter of equity. Different factions in the party will argue that it is only fair that one of their representatives be among those candidates who have a genuine chance of winning. Especially when a woman's branch of the party has been established and is active in doing a significant amount of the party's work, women can be one of those groups demanding to be included on the list in winnable positions. A third reason for balancing the slate is that dividing safe seats among the various factions in the party is a way of maintaining party peace, and assuring the continued support of the various groups within the party.
Proportional representation systems help women because a process of contagion is more likely to occur in these systems than in majoritarian systems. Contagion is a process by which parties adopt policies initiated by other political parties. We set out to test whether major parties would more quickly move to promote women when challenged on this issue by another party in PR systems than in majoritarian systems. The assumption was this should happen both because the costs of responding would be lower in PR systems compared to majoritarian systems and the gains may be greater. The costs would be lower in a PR system because the party would have several slots from which it could find room to nominate a woman; in majoritarian systems, where the party has only one candidate, the party might have to deny renomination to an incumbent or deny a slot to the male candidate of an internal faction which has traditionally received the nomination, in order to nominate a woman. The gains may be greater because in PR systems even a small increase in votes, caused by adding women to the ticket, could result in the party winning more seats.
To study this question, we looked for contagion effects in Norway and Canada. Looking for contagion effects in elections prior to the dominant Labour Party adopting quotas, we found that contagion occurred within local districts in Norway. The Norwegian Labour Party increased the number of women in winnable positions in exactly those districts where they faced a serious challenge by the Socialist Left, the first party to adopt quotas in Norway. When we tested for a similar effect in Canada that is whether the Liberal Party was more likely to nominate women in those districts where the New Democratic Party had nominated women, we found no evidence of such an effect. In other words contagion occurred in the country with a PR electoral system and did not in the country with a majoritarian electoral system.15
More generally it is worth noting that gender quotas as a policy clearly have been contagious in Norway. In 1977, only two parties with less than four per cent of the parliamentary seats had quotas. Today, five of the seven parties represented in parliament, with approximately 75 per cent of the seats combined, have officially adopted gender quotas.16
Why Some PR Systems are Better than Others
Three factors that facilitate womens representation in PR Systems:
- Higher District Magnitude: Parties have the chance to compete for and win several seats, allowing them to go further down the party lists, where women are usually listed.
- High Electoral Thresholds: Discourage the creation of mini-parties which often let in only one or two representatives, usually male.
- Closed Party Lists: The party determines rank ordering of candidates and thus womens names cannot be struck off or demoted.
While proportional representation systems are superior for women, not all PR systems are equally preferred. There are a number of particulars that can help or hinder women's representation within the broader umbrella of PR systems. There are three specific issues that deserve mention: district magnitude, electoral thresholds, and the choice between "open list" and "closed list" forms of proportional representation.
As noted, the driving force behind women doing better in PR systems is the ticket balancing process which occurs when the party sets up their election list in each electoral district. What is crucial, if women are to win seats in parliament is that parties have to win several seats so that they go deep into the party list when selecting MPs. Previously party magnitude was defined as the number of seats a party wins in an electoral district. In designing electoral rules, women will be helped both by having high district magnitudes and by electoral thresholds, because of their effects on average party magnitude. Not surprisingly there is generally a strong positive correlation between average district magnitude and average party magnitude. As the number of seats per district increases, parties will go further down their lists (that is, win more seats) and more parties will have multi-member delegations. Both should increase women's representation. The limiting case, and the one that may be the most advantageous for women, is if the whole country is simply one electoral district. There are other considerations that may render this proposal unattractive. In many countries it is often seen as important to guarantee regional representation, in which case some geographic form of districting may be preferred.
Figure 3. Why PR Systems are Better for Women|
|HIGHER DISTRICT MAGNITUDES|
|Has higher number of seats per district (higher district magnitude)|
|Party can expect to win several seats in each district (higher party magnitude|
|Party more likely to balance ticket by including women (balancing)
|Party lists present greater opportunities to nominate women|
|Greater capacity to promote women when challenged by another party (contagion)|
|And party does not have to pay the cost of denying a slot to incumbent or male candidate in order to nominate a woman.
© INTERNATIONAL IDEA
This is a system similar to the one used in the Netherlands, which has a very high level of women's representation (31.3 per cent) and in Israel, which has a low level of women's representation (7.5 per cent). As the results for the Netherlands and Israel indicate, electoral systems cannot guarantee high representation levels. One lesson that can be learned from looking at Israel is that having a high electoral threshold, which is the minimum percent of the vote that a party must have before being eligible to win a seat, is important to help women's chances. In Israel the level of support needed to win a seat has been extremely low; it was recently raised to 1.5 per cent which continues to be quite low. The low threshold has encouraged the creation of many mini-parties, which often let in only one or two representatives. Overwhelmingly parties tend to have male leaders, and party leaders inevitably take the first few slots on the list. Women first tend to show up a little farther down the list when the party concerns turn to ensuring ticket balance. If the party only elects one or two representatives, however, even though many of their candidates in mid-list positions are women, women will not win any representation.
When designing electoral systems there is in effect a trade-off between representing the voters who choose small parties and increasing the descriptive representation of the legislature by having more women from the larger parties. To test this hypothesis, data from both Costa Rica and Sweden was evaluated. Both of these countries use electoral thresholds. Simulations show that electoral thresholds had precisely the predicted effect of increasing women's representation. Women may look favourably upon proposals to establish the whole country as one electoral district, but it would be an important strategic addendum to make sure that electoral thresholds are included in the proposal.
Another characteristic that distinguishes proportional representation systems from each other is whether they use closed party lists, where the party determines the rank ordering of candidates, or open party lists, where the voters are able to influence which of the party's candidates are elected via personal voting. There is relatively little empirical work as to whether these different forms of ballot structure help or hinder women gaining access to parliament.
The crucial question is whether it is easier to convince voters to actively vote for women candidates, or easier to convince party gatekeepers that including more women on the party lists in prominent positions is both fair, and more importantly, strategically wise. It would not be too surprising if the answer actually varied from country to country. It is possible, nevertheless, to make some cautious suggestions. While there is a temptation to recommend open party lists, because this would allow women voters to move women up through preferential voting, closed lists are likely to be superior for women.
First, the experience from preferential voting, that is, open lists, in local elections in Norway for the last 25 years has been unambiguous: it has hurt women. In every local election after 1971 there have been fewer women elected than would have been elected without a preferential vote. One must realize that while preferential voting provides the opportunity for some voters to promote women, this can easily be outweighed by the opportunity for other voters to demote women. In Norway, the negative effect has consistently outweighed the positive effect. It is perhaps important to note that if this effect has shown up in Norway, which has a deserved reputation for being highly progressive on issues of gender equality, it would hardly be surprising to find similar effects in countries with more traditional views on the proper role for women. It may be that in countries with more traditional views, or even within specific districts within a country, voters with traditional views of women's roles would go out of their way to strike or lower the women's names on the party list. So the first objection is that strategically the use of preferential voting may backfire for women.
The second objection to open lists is that it lets the parties "off the hook". That is, they are not responsible for the final outcome. The final outcome then rests with thousands of individual voters making individual decisions. If the sum of all those individual decisions is that women are voted down and out of parliament, the parties cannot be held responsible, as they cannot control how their supporters vote. With closed party lists, however, it is clear it is the party's responsibility to ensure there is balance in the party delegation. If women do poorly under these conditions it cannot be explained away as the responsibility of voters. By using closed lists, the party has the opportunity to look at the composition of the complete delegation rather than having the final outcome be the summation of number of individual decisions. Under these conditions parties could be held responsible for women's representation. If representation failed to grow, women could search out parties that were more willing to consider their demands for representation.