252 Days of Resistance: What We’ve Learned About Calling Behaviors Over Time

By Brett Major

The resistance movement has brought with it an abundance of new tech tools and resources designed to help everyday citizens use their voices to influence the political system. For instance, organizations like Flippable and Swing Left have created tools to help fundraise, volunteer, and connect with other progressive citizens in key election districts. Groups like Daily Action and 5 Calls have built websites, apps, and tech platforms to help citizens take action by providing topics, scripts, and phone numbers for calling elected officials about pressing political issues. These tools have become vital to the resistance movement for mobilizing citizens to take action. At CitizenBe, we’re using behavioral science to help progressive groups understand how to make these tools more effective.

Recently, we’ve been working with our friends at Daily Action to uncover how to sustain calling behaviors over time. For the last year, Daily Action has been sending its users daily text messages asking them to make a call about an important issue and tracking the response rates. We explored a data set of 259 call requests sent to over 200,000 Daily Action users over a period of time from December 2016 until August 2017. We looked at whether the percentage of people who responded to the call request changed over time and whether this relationship varied by the topic of the call. In particular, we were interested in topics that emerged repeatedly in call requests, like the investigation of Russian meddling in the US election and Republican attempts to repeal or replace Obamacare. Here’s what we found:

1. Over time, calls have decreased. The graph below depicts a high peak in response rates around Inauguration Day. Since then, response rates have decreased, but appear to have leveled off somewhat. Though this decrease in activity may be somewhat discouraging to progressive activists, the pattern follows a cycle that is typical of civil movements.

2. Across all topics, people were significantly less likely to call if they had already been asked to call about that topic before. On average, users were almost twice as likely to take action when they were asked to make a call about a topic for the first time compared to when they were asked for a second or third (or more) time.

3. This was especially true for calls about Russia: The more people were asked to call about Russia, the smaller the response rate. During the span of call requests we analyzed, users were asked to make 20 calls about Russia and there was a significant downward trend in response rates over time.

4. For calls about health care, the opposite pattern emerges: The more people were asked to call about health care, the greater the response rate. We probed two periods of time in which the topic of the calls to action was predominantly health care related. The first was the two week period leading up to June 29th, around the time it became clear Mitch McConnell would delay the vote to repeal ObamaCare until after the Fourth of July recess. During this 13 day period, 100% of call requests were about health care. Unlike repeated call requests about Russia, we found that over time, the more people were asked to make calls about healthcare, the more likely they were to take action and make the call.

The second time span we explored was the three week period leading up to July 28th when a scaled-down version of an ObamaCare repeal bill was dramatically voted down in the early morning hours. During these 21 days, 70% of all call requests were about health care. Consistent with the healthcare-related finding above, we found that over time, the more people were asked to make calls about health care, the more likely they were to take action.

So why do requests about healthcare lead to more calls, while requests about Russia lead to fewer calls?

One possibility to explain why response rates increased with repeated call requests about healthcare is that callers may have felt like this was an issue they were winning. During each of these time spans, there were many positive signs that Republicans would not reach number of votes needed to pass a bill to repeal or replace existing healthcare policies. As a result, callers may have felt like their calls were making a difference that might ultimately end in success, thus sustaining their engagement in calling. Contrast this with the investigation of Russian interference in the US election, where it seems less like activist efforts have resulted in tangible victories on the issue (at least as of August). In the case of Russia, callers may not have felt like their efforts were having a meaningful impact, thus resulting in burnout or habituation.

Behavioral science and psychological theory support this idea that people are more likely to sustain engagement in an activity if they feel like they are on the winning side. For instance, research on bandwagon effects indicates that news of winning makes people more likely to show their support — like when a successful sports team gains new fans. In addition, field experiments on email fundraising have found that messages indicating an incumbent candidate was narrowly leading in the polls were more effective in garnering donations from uncommitted voters than messages indicating a candidate was narrowly losing in the polls.

Another possibility to explain these findings involves differences in construal level between the topics of the Russian investigation and healthcare. Construal level refers to the psychological distance and the extent to which a person’s thoughts about a topic are abstract or concrete. Drawing from this perspective, it’s possible that the topic of healthcare hit close to home and felt personally relevant to callers, whereas the topic of the Russian investigation may have felt psychologically distant and unlikely to have many personal ramifications for callers. Ultimately, the greater psychological distance associated with the Russian investigation may have resulted in lower response rates because people felt like their immediate actions were less relevant and less likely to influence outcomes.

Though the above findings are promising, a more rigorous study would be needed to better understand whether construal level or perceptions about the “winnability” of an issue are mechanisms that drive increased engagement in calling. Going forward, it will also be useful to conduct experimental research with the goal of understanding how to help callers feel like every call they make is winnable and personally relevant. At CitizenBe, we believe that these are behavioral science questions we are uniquely qualified to address. If you are a progressive organization interested in incorporating these ideas into your work, we would love to hear from you. Please reach out to us at info@citizenbe.org.

CitizenBe is a non-profit group of social and behavioral scientists that works to create and sustain impactful political participation. We help groups become more effective through evidence-based research and testing. If you’d like more info, please reach out to us at info@citizenbe.org.