People's Action Deep Canvass Political Persuasion Experiment SUMMARY OF FINDINGS SEPTEMBER 2020 KEY FINDINGS ➔ 4.9% impact on Trump’s vote margin with women​ and ​8.5% independent women​. ➔ 3.1% overall impact on Trump’s vote margin. ➔ Impact is larger than the 2016 margin of victory in nine key battleground states including Michigan, Pennsylvania, Wisconsin and Florida which would translate to ​108 electoral votes​. ➔ Estimated 102 times more effective​ per person than the average Presidential persuasion program, as documented by academic research1 and is one of the only proven field strategies to shift presidential vote choice. EXECUTIVE SUMMARY People’s Action has completed the first-ever deep canvassing political persuasion experiment this month and the results are groundbreaking. We partnered with the New Conversation Initiative to help design the deep canvass persuasion strategy and researchers David Broockman Josh Kalla to conduct an evaluation of the program. DEEP CANVASSING & SCRIPT DEVELOPMENT Deep canvassing is one of the most proven and durable forms of persuasion2 and is far more effective than other forms of persuasion employed by traditional political campaigns.3 ​Deep canvasses are candid, two-way conversations where canvassers ask voters to share their relevant, emotionally significant experiences and reflect on them aloud. ​ ​The script developed for the deep canvass program incorporated many of the key insights from the Race Class Narrative Action. Kalla and Broockman conducted a recent study that demonstrates that pro-Biden messages are nearly twice as effective at moving voter choice.4 In addition, they found that specific pro-Biden messages are more effective than vague ones and anti-Trump messages do not have any substantive effect on shifting voter choice because we have reached a saturation point with voters on President Trump.5 The deep canvass phones experiment built a script that focused on story exchanges around care and personal stake in the election and also told a compelling story about VP Biden that speaks to voters' lived experiences and helps map their emotional journey in this moment. EXPERIMENT DESIGN & RESULTS People’s Action​ worked closely with researchers David Broockman and Josh Kalla to prepare a version of the program that persuades over the phone and measures effects using a “difference in differences” design. This involved identifying a list of 460,000 voters who received a letter inviting them to take an online public opinion survey and respondents took two different baseline surveys in May and June 2020. In July 2020, People’s Action canvassers attempted to call everyone who responded to both baseline surveys and completed 695 conversations. The findings are based on a post-survey that was conducted an average of six days after the deep canvass conversations were completed. ​Broockman and Kalla find that the deep canvass calling program increased Biden vote margin by a statistically significant 3.1 points (p = 0.03). That is, for every 100 completed calls, People’s Action was able to add 3.1 votes to Biden's vote margin. The program was particularly effective among women (4.9 points), self-reported Independents (6.7 points), and Independent women (8.5 points). In addition to increasing Biden’s vote margin, the program also increased the intensity of Biden support and self-reported intent to vote this November. 1 ​KALLA, J., & BROOCKMAN, D. (2018). The Minimal Persuasive Effects of Campaign Contact in General Elections: Evidence from 49 Field Experiments. American Political Science Review,​ ​112​(1), 148-166. doi:10.1017/S0003055417000363 2 ​https://www.vox.com/2020/1/29/21065620/broockman-kalla-deep-canvassing 3 ​https://www.vox.com/policy-and-politics/2017/9/28/16367580/campaigning-doesnt-work-general-election-study-kalla-broockman 4 5 ​https://osf.io/m7326/ ​https://www.vox.com/2020/6/9/21284758/broockman-kalla-biden-ads-trump-ads   PEOPLE’S ACTION   a powerful force for democracy and economic fairness    Chicago DC (312) 243-3035 peoplesaction.org 2