Canvassing campaigns aimed at increasing women's political participation in developing countries with patriarchal gender norms are more likely to succeed when they target men as well as women, according to a new study co-authored by Yale political scientist Sarah Khan.
The study, published in the journal American Political Science Review, is based on a field experiment conducted in Lahore, Pakistan in which researchers tested the efficacy of a nonpartisan canvassing campaign conducted by Pakistani civil society organizations, Aurat Foundation and South Asia Partnership-Pakistan, to increase women's turnout in the July 2018 national elections. The experiment showed that canvassing only women had no effect on their turnout in the election, while households where both men and women were canvassed had the largest increases in women voting and in men's willingness to help them cast ballots.
"Efforts to close the gender voting gap tend to focus on women, but this approach may not be effective in patriarchal settings where men act as their households' gatekeepers, exerting varying levels of control over women's participation in the public sphere," said Khan, an assistant professor of political science in Yale's Faculty of Arts and Sciences. "Our study suggests that interventions in such settings aimed at increasing women's participation in the political process should take these household dynamics seriously and engage with men as well as women."
Co-authors of the study, in addition to Khan, were Ali Cheema of the Lahore University of Management Sciences and the Institute of Development and Economic Alternatives, Shandana Khan Mohmand of the Institute of Development Studies in the United Kingdom, and Asad Liaqat, research scientist at Meta, formerly known as Facebook Inc.
In their experiment, they randomly assigned 2,500 Lahore households to one of four experimental conditions: a canvassing visit by a female targeting women in the household; a visit by a male canvasser targeting men; separate visits by female and male canvassers targeting women and men, respectively; or no canvassing visit at all.
Canvassers visited households unannounced for 20 minutes, during which they would discuss the importance of women's participation in the upcoming national election and show participants a five-minute video on a handheld tablet device about a woman inspired to vote out of concern over the delivery of services in her neighborhood. In the video, the woman's brother supports her decision and gives her a ride to the polling place. Afterward, the canvasser would share practical information on the election and voting process.
"The decision to tie the encouragement to vote to accountability for local service delivery issues was based on focus group discussions conducted with women in Lahore which revealed the centrality of such concerns in women's everyday lives," said Khan Mohmand, who specializes in mixed methods research.
While studies of get-out-the-vote efforts in the United States can rely on publicly accessible voter records to verify turnout, the survey team in this study had to visit the households in the sample and identify whether the targeted individuals' thumbs were marked with ink, which was the only way to know whether they had voted. This verification had to occur quickly because the ink tends to wear off within a couple of days, particularly on women who do most of the household washing and cleaning, Khan explained.
The researchers discovered that targeting only women had no effect on turnout. However, women's turnout increased by 5.4 percentage points in household in which canvassers targeted men. It increased by 8 percentage points in households where both men and women were canvassed. The gender gap in the 2018 national election was 9.1 percentage points, with 11 million fewer women voting than men.
"That 8-percentage-point increase is substantial when you benchmark it against the average gender gap in Pakistan's 2018 national election," Cheema said. "We are hopeful that these results will lead the Election Commission of Pakistan, political parties, and civil society organizations to introduce campaigns that convince men to act as enablers of women's political participation in Pakistan's upcoming general elections."
Two months after the election, the researchers offered the men in their sample group the opportunity to post a publicly visible sticker on the entrances of their homes that either bore a statement supporting women's role in democracy or a generic statement in support of democracy. They found that households where both men and women were canvassed were significantly more likely to post the sticker announcing support for women's participation in democracy than households that received no intervention.
"Posting a sticker is a costly action for men in this context, and this measure provides a glimpse into real-world behavior that surveys cannot," Liaqat said. "The result indicates that our intervention had lasting effects on men's behavior in these households."
The researchers also found that in households where both men and women were canvassed, respondents were significantly more likely to report discussing politics with each other. Those households were also more likely to report that men had enabled women's participation in the election by organizing their transportation to the polling station and waiting for them as they cast their ballots.
"Our study demonstrates that canvassing both the men and women in a household can achieve significant short-term change in places where social norms designate men as gatekeepers of women's actions," Khan said. "However, it also reveals a fundamentally unequal status-quo under which women's participation in politics remains contingent on male gatekeepers, and this status-quo requires longer-term efforts for transformative change."
