The exercise of ‘her’ and ‘his’ right to vote is an important requirement of political accountability in democracies. The gender gap in electoral participation and in party engagement with voters - and the lack of autonomy exercised by women voters (when they do vote) - are hard challenges facing emerging democracies. These challenges impose severe contraints on whether women’s voices make it through to parties, and whether parties are responsive to women’s preferences. But we know little about the nature and extent of these challenges in the mega-cities of emerging democracies. Our research – Politics, voice and responsiveness in urban Pakistan - has garnered evidence on these questions for the Pakistani mega-city of Lahore, home to 10 million people.
The main focus of the research is on understanding how parties organise to aggregate voice in an urban context, and what mechanisms ensure that this voice is reflected in policy formulation. The evidence discussed here is based on data from a political attitudes survey that was conducted in early 2017 with a representative and gender stratified sample of 2,127 voters in three out of Lahore’s 13 national assembly constituencies; for one of these we also analysed data for the 2013 election and a subsequent by-election. Our sample included constituencies with high, moderate and low levels of electoral competition.
The gender gap in electoral participation and party engagement
Our analysis of the 2013 electoral returns data shows an 8% gender gap in voter turnout; by 2015, this had increased to over 12 percentage points. This gap is only slightly lower than the 11% gender-gap in turnout in Punjab overall, which suggests that the gender gap in electoral participation is a big challenge in both rural areas and large urban centres.
Data from our survey shows that political parties engage with women at much lower rates than with men. In the almost four years since the 2013 general election in 2013, only 3% of female voters in our sample had contact with any political party worker – as opposed to 17% of male voters. Women’s voices are simply not making it to political parties.
The puzzle is, why don’t political parties have high-powered incentives to mobilise female voters and be responsive to their preferences, when these voters can give them a definite edge in closely contested electoral races?
Our field research with party leaders and workers in the most closely-contested constituency in our sample suggests that this is partly a consequence of deeply ingrained beliefs about women voters among leaders, workers and political pundits – which we find aren’t supported by hard evidence.
Do women voters lack autonomy?
A commonly held belief that we encountered is that women voters lack autonomy. The implication drawn from this belief by party workers is that they need to focus their effort on male family members, as the vote of women members, if cast, is expected to follow the lead of male members. So, do women voters vote the same way as men?
We tested this in the most closely-contested constituency by comparing the direction of ‘swing’ in male and female polling stations between the 2013 election and the 2015 by-election. We found that while men’s polling stations witnessed a 6 percentage point swing against the ruling party in 2015, women’s polling stations in the same neighbourhoods displayed a tiny swing of 0.5 in favor of the ruling party. This isn’t the behavior of voters who lack autonomy.
Evidence from our 2017 political attitudes survey reinforces this. We interviewed a randomly selected man and woman from each sample household in the same closely-contested constituency, finding that women were 15% more likely to be undecided voters than men in the same household. We found that 60% of ‘undecided voters’ were women. So - contrary to conventional belief - this evidence suggests that not only do women’s votes matter, but that they can be pivotal for electoral outcomes. When parties fail to engage with women, and do not take their voices into account, they are failing to capitalise on the majority of undecided voters.
Related to the belief about women voters’ lack of autonomy is the belief that they support the same issues as male family members. The implication drawn from this is that there is no point in running election campaigns that specifically target women, and that being responsive to men is the same as being responsive to women.
Again, the evidence from our political attitudes survey suggests the opposite. We asked respondents to name the three important issues that will matter in the 2018 general elections. We found that women voters were much more likely to name service provision issues such as electricity (8% higher), sanitation (6% higher) and gas (12% higher) in the top three issues, compared to male family members. They were less likely to name corruption (30% lower) or purchasing power (8% lower) as key issues compared to male members.
So campaigns that are designed to appeal to men aren’t likely to appeal to women voters. On the other hand, engaging directly with women voters and taking their voices into account when designing campaigns and policies has the potential not only to close this big gap in Pakistan’s democratic engagement, but also to deliver electoral success to the party that adopts such stratetegies.
Why don’t women voters matter?
So why don’t parties make an effort to mobilise women voters in constituencies that offer clear electoral returns to such an effort? The social conservatism of Pakistani society is a popular explanation, but it does not explain our sample neighbourhoods in the heart of Lahore city. There are other factors at play here, and we believe that an important one is the weakness of party structures in emerging democracies.
In this context of weak party organizations political strategies are driven by the prior beliefs of party leaders and workers and are not informed by evidence. The result is a deep disconnect between women voters and political parties. It is unsurprising that our survey finds that ‘undecided’ women voters are 21% more likely than their male counterparts to agree or strongly agree with the statement that “political parties are only interested in people’s vote and not in their needs.”
Going forward: engaging women effectively
Closing the gender gap in voter turnout has potentially high returns for the parties that are able to successfully mobilise women’s votes. Moreover, equalising participation across different groups in society is an important indicator of democratic progress.
Although civil society organisations (CSOs) in Pakistan have worked to encourage women to turnout and organised female-centric voter education campaigns in past elections, these efforts have not been undertaken with the purposive involvement of political parties. It is an open question as to whether efforts by non-partisan CSOs or partisan political actors are most effective at mobilising women to turnout, and whether there are additional gains to be had from coordinating these efforts.
We tackle this question in an upcoming joint project with Sarah Khan (Columbia University) through the Action for Empowerment and Accountability Research Programme. In this project, Exercising Her Right to Vote: Civic and Political Action as Pathways to Women’s Empowerment in Pakistan, we are first working to identify the context-specific factors contributing to low political engagement among women in a select number of localities through citizen surveys with men and women, and interviews with key actors such as political party members and workers, election officials and polling officers and informal community leaders. In the second phase of this project we will work with political parties and CSOs to test the efficacy of women’s turnout campaigns in the next 2018 elections in Pakistan using a field experimental approach.
Our work in Lahore shows that men and women find different issues to be salient for the upcoming election; we plan to use this insight, along with evidence on gendered barriers to participation, to motivate the design of issue-specific mobilisation efforts targeted at women. Treating women as autonomous voters with shared collective interests is not the norm, but it may be an effective strategy for closing the gender gap in political participation.
About the authorAli Cheema (Institute of Development and Economic Alternatives and Lahore University of Management Sciences), Asad Liaqat (Harvard Kennedy School) and Shandana Khan Mohmand (Institute of Development Studies, Sussex)
PROJECT | August 24, 2016
Politics, voice and responsiveness in urban Pakistan
BLOG | March 28, 2017
Whose voice really counts? Machine politics and citizen voice aggregation…
BLOG | May 23, 2016
Politics, persistence and publicity - transforming education in Pakistan
BLOG | August 25, 2016
Context and accountability in Pakistan: a new generation of champions?
More on these themes
PUBLICATION | October 16, 2017
Doing digital development differently: lessons in adaptive management from technology…
EVENT | October 11, 2017
Making All Voices Count Learning Event in Indonesia
EVENT | October 11, 2017
Working with local government - Making All Voices Count Event…
EVENT | October 11, 2017
Know your context - Making All Voices Count Learning Event…
EVENT | October 11, 2017
Making All Voices Count Kenya close-out event
NEWS | October 10, 2017
Research, Evidence and Learning Newsletter - October 2017