In 2008 there were only four countries in which at least 40% parliamentarians were female (Inter-Parliamentary Union, 2008).i Given the low number of female legislators worldwide, it is not surprising that political parties and governments in many countries are aiming efforts at redressing the situation. While some parties are self-imposing internal gender quotas, governments are increasingly passing legislation mandating quotas in candidate lists (European Commission 2009).
It has been argued that the lack of female legislators may reflect voters’ preferences for male politicians (Frechette et al 2008). Other authors point out that the low number of women legislators may not necessarily be due to lower popularity among voters but rather to a political failure (Persico 2009). According to this view, political parties’ constitute filters between voters’ preferences and the politicians who get finally elected, and such filters may favour male incumbents. In a recent study, we examine parties’ nominations to the Spanish Senate during its last four elections (1996, 2000, 2004 and 2008) (Esteve-Volart and Bagues 2009). In line with the latter view, we find that in constituencies where the political arena is not competitive, the quality and the quantity of female politicians are relatively lower.
The Spanish case
Spain is nowadays one of the most advanced countries in terms of gender equality policy. The Spanish Equality Law, passed in 2007, requires political parties to choose roughly equal numbers of men and women in candidate lists. However, it has not led to significant changes in the number of women elected; the percentage of women in the House of Representatives increased from 36% in the 2004 election to 36.3% in the 2008 election; figures were 25.1% and 28.2% for the Senate respectively (Instituto de la Mujer).
The existence of individual voting in Senate elections allows us to disentangle the sources of the low female representation. In Senate elections, each party presents three candidates per province, who are listed in the ballot by alphabetical order. Each voter can vote for up to three candidates, and the four most voted candidates in each province become senators. Generally one of the two main parties (Socialist Party and People’s Party) gets all three candidates in the list elected, while the other party only gets one candidate elected.
Political parties can affect the number of female candidates who get elected in two ways. First, in the absence of gender quotas, parties decide how many women they nominate in each province. We find that parties nominate relatively fewer female candidates in constituencies where, based on previous electoral results, the likelihood of being elected is relatively higher. In provinces where all three party candidates are expected to be elected, 26% of candidates are female, on average. In provinces where only one out of the party’s three candidates is expected to be elected, this figure is 38%. That is, female candidates are especially underrepresented in winning constituencies.
Second, parties can also affect candidates’ chances of being elected through their order in the ballot. In over 97% of cases, the candidate who gets more votes is the candidate at the top of the party list. While order in the ballot is alphabetical, parties can potentially manipulate the order by strategically nominating candidates on the basis of last name. In this sense, if male and female candidates were not selected based on their last name, we would expect to find approximately a third of them in each position in the ballot. In contrast, we find that male and female candidates are allocated to different places in the ballot depending on parties' electoral expectations. Whenever the party expects to obtain only one seat, a meagre 6% of female candidates are placed at the top of the ballot (and so 94% are in either the second or third position, where they will have almost no chances of getting elected). Whenever the party expects to get all three seats, the pattern reverses – even though order in the ballot does not matter there, 59% of female candidates in these lists are placed at the top. Unsettled races are somewhat in between: women tend to be more towards the bottom of the ballot, but much less so than in expected losing lists (about 19% of women are at the top of the list). Symmetrically, male candidates exhibit the opposite pattern.
Comparing candidates’ last name order with last name order in the overall Spanish population, it is possible to directly test whether certain candidates are chosen on the basis of last name. For this we construct a measure of last name order between zero and one using the distribution of last names in the Spanish population. For instance, an individual with last name Vázquez would be in the 96th percentile, that is, 96% of the Spanish population has a last name alphabetically prior to Vázquez. Since our measure is by construction on average equal to 0.5, any systematic deviation from 0.5 in the distribution of male and female candidates' last names is consistent with surname-based nomination.
In Figure 1, each column represents the average candidate’s last name order as measured with respect to the distribution of last names in the Spanish population.
Figure 1. Senate candidates’ last name order, by type of list
Notes: Authors’ calculations using data from 936 candidates who were running for the main two political parties in every peninsular province in the Senate elections in years 1996, 2000, 2004 and 2008. Party lists have been classified in three groups according to the previous electoral outcomes. The vertical axis denotes the percentile for a particular last name order. Vertical lines denote confidence intervals at the 95% level.
In provinces where the party expects to win all three seats, women tend to be drawn from the beginning of the alphabet (0.29 on average, corresponding to the last name "Fernández"). In provinces where the party expects to win only one seat, women's last names tend to be drawn from the end of the alphabet (0.70 on average, corresponding to the last name "Pardo"). In contrast, in unsettled races, the distribution of women's last names is much closer to the population distribution (0.55 on average). In all three cases, the distribution of male candidates' last names does not differ significantly from the distribution of last names in the Spanish population.
What is the rationale for such nomination strategies? In provinces where the party expects only one seat, the party nominates women whose last names are from the end of the alphabet. Female candidates tend to be placed towards the bottom of the ballot so they have almost no chances of being elected. These nomination strategies cannot be justified in women’s lower popularity among voters; in fact, we find that women attract slightly more votes. But the order effect is so strong that women’s greater popularity seldom manages to overcome the order disadvantage. In provinces where the party knows that they will win all three seats, parties nominate women whose last names are from the beginning of the alphabet and hence are placed at the top of the list. This might seem puzzling; there is no clear benefit in the strategic selection of women in this case, as all candidates in the list will get elected, however this strategy is costly because these women are selected from a constrained pool, and thus their quality cannot be the highest. One possible explanation is that parties are explicitly trying to have a number of women heading lists, perhaps for positive media treatment.ii In contrast, in provinces where there is uncertainty about the outcome of the election, the strategic nomination of female candidates is weaker.
In sum, when there is no uncertainty about the number of seats that parties will obtain, parties use women as pawns, in that they are chosen according to how their presence in the list would affect male candidates’ possibilities of success and gender statistics. That is, political parties base relatively more their nomination decisions on quality when it is more costly not to do so. Not only does the absence of political competition hinder the quality of female legislators, but it also reduces the number of female candidates who become senators. We find that in settled races the number of women elected is 20% lower than in unsettled races.
What can be done?
The current policy in Spain involves gender quotas. Gender quotas can help to increase the number of female candidates that are nominated in different provinces. However, and as we show, as long as parties do not have adequate incentives, quotas do not erode the surname-based nomination of female candidates. As a result, in the absence of political competition, quotas only have a limited impact on the number and quality of female senators.
One possibility to improve the selection of female candidates involves fostering internal competition within parties by changing the ballot design. The existence of an order effect in Senate elections currently allows parties to favour certain candidates. In order to enhance political competition within parties, we propose ballot-ordering rotation; ballot positions should be rotated across candidates within the same list, and three copies printed, so that every candidate tops the ballot as often as every other candidate. Rotation would therefore neutralise the existing order effect and hence erode surname-based nomination. Given our results, such rotation would increase both the number and the quality of female senators.
ii For instance, in the 2004 election, the Socialist Party’s central committee gave explicit instructions that at least a few provinces should have a woman heading the province party list. http://www.diariodeleon.es/noticias/noticia.asp?pkid=116604, retrieved 22 September 2009.
Esteve-Volart, Berta and Manuel Bagues (2009), “Are Women Pawns in the Political Game? Evidence from Elections to the Spanish Senate”, FEDEA Working Paper 2009-30.
European Commission (2009), “Women in European Politics - A Time for Action”, Report by the Directorate-General for Employment, Social Affairs and Equal Opportunities Unit G1.
Frechette, Guillaume R., Francois Maniquet and Massimo Morelli (2008), Incumbents' Interests and Gender Quotas, American Journal of Political Science, Vol. 52(4), pp. 891-909.
Inter-Parliamentary Union (2008) "Women in Parliaments: World Classification", (retrieved 26 January 2009), compiling information provided by National Parliaments by 30 November 2008.
Persico, Nicola (2009), Quote Rosa: Un Fallimento della Politica, LaVoce, 26 September.