The Voting Behaviour Of Women Is Different From The Voting Behaviour Of Men: Comparative Politics Research
To introduce this paper, I will first introduce my research question which can be used in comparative politics research. As Giger (2009) concluded, in Europe, the voting behaviour of women is different from the voting behaviour of men. Her findings suggest that women tend to vote more to the ‘left’ than men.
She speaks of an absence of structural factors that can explain gender differences in voting behaviour. To the contrary, she speaks of generational factors. So the factors that can explain differences in voting behaviour between the two sexes in the 1970s do not apply to the current situation. Earlier research, done by Jelen, Thomas and Wilcox (1994), also discussed this ‘gender gap’. Instead of explaining left-right differences between men and women, they got the idea that this ‘gender gap’ is all about interpretation of the left-right dimension. They concluded that the left-right continuum appears to have a different meaning for men and women. This idea of an interpretation of left-right dimensions by both men and women is particularly interesting regarding the ‘traditional’ gender gap which showed women voting more to the right than men before the 1970s (Inglehart & Norris, 2000, pp.458-459). Seen the current situation, completely the opposite is true, as we previously discussed the phenomenon of women voting more to the left than men.
Then it made me wonder whether there are maybe structural factors that can explain differences in voting behaviour, without being bound to differences due to interpretation. Women tend to vote more to the left and what is interesting given this fact; the left is more dominated by women in parliament than the right. For example, in the Netherlands, there is a left-wing party called GL, which is divided equally by gender (7 men, 7 women). The VVD, a right-wing party, is doing worse with only ten women against twenty-two men (Tweede Kamer der Staten-Generaal, n.d.). Moreover, research about gender differences on support for radical right-wing parties shows that traditionally oriented women are less attracted by these parties than men (Lodders & Weldon, 2019, p.470). Maybe, this is because of radical right-wing parties not having many women seated in their parties. So this could be an explanation for the gender differences in voting behaviour. For me, it sounds very interesting to study this further. Therefore, I have chosen the following research question: ‘Do more women vote when the percentage of women in parliament is higher?’
According to Campbell and Heath (2017), it is not crystal-clear that women vote for women candidates. They conclude “that exceptional circumstances where issues of gender equality become highly politicized and extremely electorally salient, such as the 1992 elections in the United States, may yield a situation where more women are motivated to vote by feminist concerns but our research suggests that in a typical election this group will be small” (Campbell & Heath, 2017, pp. 19-20). Campbell and Heath (2017) tend to highlight the 1992 elections as an exceptional case. These findings are similar to earlier research, done by Dolan (1998), she concluded that women voting for women is a complex phenomenon and therefore difficult to understand. It is very different from the phenomenon of voting for democrats or republicans, which is a lot easier to understand (Dolan, 1998, pp.288-289). This ‘women voting for women’ phenomenon becomes highly visible in the research done by Campbell, Childs and Lovenduski (2006). According to them, conservatives have lost the support of women, because of the ‘scandalous’ underrepresentation of women in parliament. This loss namely is a result of the women votes shifting to Labour and liberal democrats (Campbell, Childs & Lovenduski, pp.20-21).
According to Paxton and Hughes (2016), the fact that the phenomenon of ‘women voting for women’ is very complex can be explained partly through the actions of women politicians. Women politicians could state that they feel a special responsibility to women. They could defend women preferences, such as the support for equal treatment for men and women. They could prioritize women’s issues, but this all does not have to be the case. First, women politicians may not have the desire to act ‘for women’. Second, they may not have the ability to act ‘for women’. Finally, women politicians who feel that they belong to a specific group based on ethnicity or class may not desire to act for all women. A good example of a women politician who did not have the desire to act ‘for women’ was Margaret Thatcher. Many feminists were not happy with her politics, as these were for the most part conservative and anti-feminist (Fukuyama, 1998, p.32). There are of course more women politicians like Margaret Thatcher, which, in my opinion, sounds very logical. However, it looks like scholars sometimes forget this. It is important to know that not all women politicians are feminists because this implies that not all women are acting in favour of other women. Some even say that women change through politics, that they copy the methods and behaviour of men (Paxton & Hughes, 2016, pp.13-15).
It is very doubtful whether the latter is true; other research highlights the fact that women politicians used to do mostly women politics, as men used to do men politics. Women politics include more domestic politics as social welfare and education, while men politics include finance, foreign policy and national security (Conway, 2001, p.233). If we take a look at German politics, for the most part, we see this political distinction between men and women (Deutsche Bundesregierung, n.d.). Other powerful players in Europe such as France and the UK show roughly the same distinction. However, both surprise somewhat with having a man politician on education (Gouvernment de France, n.d.; Government of the UK, n.d.). As Scandinavian countries lead women politics, it comes as no surprise that the distinction between men and women politics does not apply in countries as Sweden and Norway (Konungariket Sveriges regering, n.d.; Norge Regjeringen, n.d.).
Another aspect must be added to the complex phenomenon of ‘women voting for women’. Research, done by Deshpande (2009), namely suggests that not only women politicians do not always act in favour of women, but also that women not always vote as ‘women’. To explain this vague sounding sentence; gender is not always the most important factor related to voting. Sometimes caste and class may override gender. Gender is not an overall decisive factor of voting behaviour, it is among others one element that explains partly voting behaviour (Deshpande, 2009, p.83). The influence of gender and the other factors on voting behaviour, of course, differs per country. It is understandable that especially for the Indian case, caste plays such a huge role.
One aspect of ‘women voting for women’ is particularly interesting, namely the question if women back women candidates if they run against man (Darcy & Schramm, 1977, p.1). As many research concerning this subject, this is an American example. One should note that I want to have European states to be investigated and that the electoral system in most European countries does not share the features of the American system. However, in the American example, women did not have a tendency to vote for women candidates. But as earlier hypothesized by me in the introduction, the results of this research show that gender indeed tends to cause differences in party choice.
To make the phenomenon ‘women voting for women’ even more complex than it seems to be, other American research shows results that suggest that women candidates are elected easier when they behave as women during their campaign. Behaving as women include stressing issues that are favoured by the stereotype women and gaining explicit attention to the female voters (Hernsson, Lay & Stokes, 2003, p. 244). There is a remarkable trend going on in which women vote more than men, while they voted less than men until the 1980s. According to Harder & Krosnick (2008), this phenomenon can be explained through the more political interest women show and because women have more efficacy (Harder & Krosnick, 2008, p.532). That is an interesting assumption, but instead of explaining the factors behind this relation, they just state that it is as it is. I suspect that the increasing interest in politics among (American) women is not an autonomous process, but caused by the rise of the number of women in parliament (or congress).
Although these theories are helpful for comparative politics research, just theory is not enough, especially as I want to have investigated a possible causal relationship between two variables. Research, done by Broockman (2014), studied the question whether women politicians empower other women to vote or run for office. Except for ‘the run for office’ aspect and the fact that this research only includes the American case, this research is very similar to what I suggested. Broockman (2014) found that women politicians, (against the main idea that is centred in the literature and other research done in India), do not empower other women to vote or run for office. There is one but; Broockman studied the American case, and he believes that the American case is in some aspects very different from other cases (Broockman, 2014, pp.13-14). So it might be that in Europe there actually is a relationship between these variables. To conclude this theoretical frame, it is not easy to guess what the outcome is of my proposed research. The ‘women voting for women’ is not a one-way explanation for possible results. Also, the scholars who studied this topic are by no way single-minded. The relation between the share of women in politics and the percentage of women voting with national elections relates to the research findings that women in politics do not always act in favour of women and thereby not always encourage women to vote.
Case selection strategy
Following the research question and theoretical framework, one can say that the selected topic is extremely complex, as it covers many aspects. Therefore the case selection should be restricted to reasonable proportions. As we look for a possible causal relation between two variables, namely the share of women in national parliaments and the percentage of women voting with national elections, not many variables should be included to control the other variables. However, the number of cases should be high. If it is possible, I advise taking all European states into account. To take all European states implies 43 cases, which is, in my opinion, a reasonable number. I have searched on the web for understanding to which degree the European states can be taken into account in this research. According to the website of the International Institute for Democracy and Electoral Assistance, all states have data on the voter turnout with national elections. But there is one thing I do not know by now, which is if there is data on voter turnout levels which take into account differences in gender. Besides this, data on the number of female parliamentarians must be obtained also. It came to me as a surprise, but the number of female parliamentarians for all states can also be viewed on the web.
The Inter-Parliamentary Union has data on the number of female parliamentarians, which also include historical data. The historical data were not all accessible to the same extent for all states. Romania, for example, shows only the numbers of female parliamentarians, but not as a percentage of the whole parliament. Serbia and Montenegro provide not much data, as their data only covers the period of 2006-2019. The same can be said about Albania, of which the data only covers five years (2014-2019). Although Monaco’s data covers a longer period than Albania, in 55 years there were only ten moments of measurement (Inter-Parliamentary Union, n.d.).
To take all European states, the number of cases is relatively high. This way, a possible causal relationship can be investigated properly, increasing the level of probability. There must be a few intervals, as differences must be measured in time. I would say that per state the previous five elections must be taken into account. Maybe less, but at least two elections per state must be included to be able to measure a possible causal relationship. Because of what is concluded previously, I would suggest selecting the closed universe strategy (Keman & Pennings, 2016, p. 54).
Before doing the actual research, cases already can be placed in different groups that correspond to the different electoral systems. It is widely recognized that the electoral system is a strong explanatory factor for differences in voting behaviour (Kitschelt & Rehm, 2016, pp.315-316). I would suggest taking into account the following groups: multiparty systems, proportional representation systems and mixed systems. Maybe mixed systems can be divided also into mixed systems which tend to be more multi-party or more proportional.
To conclude this paper, I will briefly summarize the content of the paper as I will refer to the topic, research question, theoretical framework and case selection strategy. To start with the topic, I am very interested in the participation of women in politics. Political participation, especially with national elections, has always intrigued me. In this paper, I defined my research question, after looking into both Dutch and international findings. The research question I want to be answered is: ‘Do women vote more when the percentage of women in parliament is higher?’ In the theoretical framework, I tried to come with all kind of explanations which possibly cause or do not cause more participation in politics among women. There was one phenomenon which I paid extra attention to, namely ‘women voting for women’. For the reader, it became soon clear that the theories concerning women participation, combined with some personal critics on these, eventually led to a very complex theoretical framework.
Ending with the case selection strategy in which I defined some criteria for researchers. I have chosen the closed universe approach, which in my opinion is a logical result of the preceding theoretical framework and research question. To have a clear view of the possibilities of my proposed research I did some scanning research myself via the internet. Luckily, I found considerable data that could be used in my proposed research. Finally, this paper demonstrates that women participating in politics is widely researched in the US and context, but that there are relatively few data on the European context. I hope that this final comment will stimulate comparative politics researchers in investigating this remarkably understudied topic.