Creating Technology for Social Change

America’s Interested Bystander: New Research from Google on Civic Duty

This is a liveblog of the talk “Understanding America’s Interested Bystander: A Complicated Relationship with Civic Duty,” by Kate Krontiris, John Webb, Charlotte Krontiris, and Chris Chapman. Blogged by @natematias and @erhardt, with illustrations by @willowbl00.

What motivates everyday people in America to do things that are civic, and how do we engage the unengaged? Kate Krontiris and John Webb shared the results of a major study carried out by Google’s Civic Innovation team today at the Berkman Center for Internet and Society.

[note: we were asked not to include any photos of the event, which was not recorded, but we were allowed to publish these notes]

Kate Krontiris is a researcher, strategist, and facilitator working to transform civic life in America. In pursuit of a society where more people assert greater ownership over the decisions that govern their lives, she uses ethnographic tools to design products, policies, and services that enable a more equitable democratic future. Charlotte Krontiris is a principal at KN research, and who has conducted research at the Harvard Graduate School of Education, the Harvard Business School, and Google. John Webb is a senior user experience researcher at Google who conducts tactical and investigative research to inform design and product direction for Google’s Social Impact team with a particular focus on developing Civic Engagement experiences.

Kate begins by outlining the social impact and civic innovation group at Google. They include the civic innovation team, which organizes election data and making it universally accessible and to broaden collective decision-making.

  • leverage Google’s technology for the common good
  • Organize election data and make it universally accessible and useful
  • Broaden engagement in collective decision-making

What motivates ordinary Americans to do things that are civic? Kate and her colleagues concede they are not the only ones researching this subject. A lot of this falls under the question of “How do we engage the unengaged?” and then to support details of platform design at Google including their Google Now cards, as well as support a broader civic technology ecosystem. They hoped that by conducting and sharing this research they could contribute to informing the broader set of tools being developed. We conducted quantitative and qualitative research to try answer these questions. 

What do we mean by civic engagement? Kate refers to a literature review on civic engagement (pdf) published by the Engagement Game Lab:

  • Election related behavior
  • volunteering
  • signing petitions
  • reporting public problems
  • organizing service events

Most often, civic engagement is assessed in modern democracies in terms of participation in voting.

Kate and her colleagues carried out 101 in-person interviews and 2058 digital surveys, gathering information from 6 different parts of the US (San Francisco, Chicago, Atlanta, Phoenix, Chicago, Boston). Their qualitative research allowed them to get rich qualitative data about what people are doing and why, which they then followed up with quantitative surveys influenced by those conversations.

Out of the 101 participants in their research, they found 65 fell into a category they are calling “interested bystanders,” which was the target of most of their analysis. They conducted 90 minute interviews with them in their homes and places of work. The researchers used choice-based conjoint quantitative surveying technique, where participants identified profiles that were more like them. Participants also responded to a series of demographic questions that the researchers used to correlate that with the deeper findings they observed (profiles).

In interviews, when participants said something like “I don’t have time for civic or related activities,” the researchers tried to pull out factors that would influence their proclaimed behaviors, put them into the personas in the survey, and then conducted latent class analysis to link survey responses to those personas. In each survey, attitudinal areas like career involvement, civic engagement, family involvement were mixed with a set of randomly associated statements to create the personas that survey takers would say best represented them.

Motivating Hypotheses
For this study, the team started with the problem that too many Americans feel disconnected from public policy and legislative decision-making in the US.

Hypothesis 1: While a portion of Americans are engaged in community and social context as volunteers, most people are not participating in a politically defined notion of civic life in a broad-scale way.”

Hypothesis 2: There is something meaningful and particular about the characteristics of these “interested bystanders” as a group—including their civic behaviors motivations and barriers to action—that can be underst to customize civic interventions.”

Emy Tseng asked about the sampling technique and how representative the sample was. Kate replies that their goal was to get as broad as possible sample of people to talk to. In the cities they went to, they tried to get a broad set of perspectives. The researchers were trying to source all these perspective in order to describe what was present in the population rather than trying to be perfectly representative.

Another audience member asks about the demographic or cultural makeup the sample in terms of faith, gender, etc.? Kate notes (referring to Wasow’s work on “Race as a Bundle of Sticks“) that it’s challenging to rely on these factors, especially in such a small sample. She replies that the team tried to be as broad as possible in their sampling but it’s hard to rely on data and responses on certain demographic indicators. Because the sample size is only 65 in the qualitative study, they don’t expect to make meaningful conclusions across demographic groups.

Interested Bystanders
The researchers focused especially on “interested bystanders,” people who are already civically aware but may not be doing as much as they want. “If we learn what motivates these people to do things that are “civic”—and what holds them back when they do not act—we can better engage this “silent majority” to be more active participants in civic life.”

Qualitative Findings
Each interviewee was asked about a time they had recently done something civic. John Webb tells us about what they did and what their motivations were. The most common activities were signing petitions (very thin), volunteering in a continuous series of engagements (much thicker), and then volunteering at a one-off event. Notably absent were activities that we would typically think of as civic: voting, volunteering in a political campaign, or participating in a political event.

Bystanders were doing things that were easy, but they were also doing things that are very complicated to do. What made the difference? John argues that most interviewees had three main motivations. Firstly, people were motivated because they had prior personal experience or expertise. Secondly, bystanders had clear interests at stake that they were trying to protect. Finally, participants were seeking emotional fulfillment—the task or action helped them fulfill that.

Common barriers to action were being too busy or having no time, concerns about what they could actually do, or that they thought that their actions wouldn’t have any impact.

In the literature of civic engagement, there is the concept of the “civic ladder”: sharing an opinion (thinnest) up to organizing (thickest). This idea of a hierarchy of participation has been internalized by many people. Many interviewees gauged themselves on this hierarchy of action. Taking action in-person included voting, protesting/campaigning, or organizing, versus actions taken online like opinion sharing. Many participants put things associated with opinion sharing at the bottom of the ladder, and they put themselves at the bottom of that ladder. They voted regularly, kept informed, signed petitions, and shared information with their peers. And yet they felt embarrassed—like they weren’t engaged.

Across most of the participants, people often talked about “being involved in the community,” even if they didn’t always see it as an important part of “civic” life. For example:

“My sister is making me and my sister volunteer at the Latin extravaganza…”

“With church, we went to a women’s homeless shelter and cooked food…”

People’s civic life is much richer than the things associated with the “civic ladder.” Webb describes two main areas of this broader civic life: the traditional political realm, that includes everything from voting and politics, and a broader civic life, a community / social engagement side. In traditional politics, people saw political engagement as advocacy. They imagined civic life as more associated with community and service. Webb notes that civic life is sometimes contentious but often collaborative.

In contrast with the ladder everyone imagines, civic life taps into their existing experiences and expertise, is more often local, and often gives individuals more power. Civic life is more emotionally fulfilling and helps people get over the “what’s the point” hump of inaction.

“There’s a lot of power in the city if we could band together as the neighborhood association, rather than John Smith.”

When interested bystanders do engage in politics, they do so via civic activities that they perceive as less useful and where they feel least powerful.

When people were asked about petitions they had signed, they often couldn’t remember what the petition was about. They remembered doing the action but couldn’t remember the actual issue or cause was. It’s possible that interested bystanders sign in order to carry out a duty, or feel disconnected from others who care about the issue. Signing petitions can be done individually, on your computer, without having to confront the anger of opposing views.

Many interested bystanders said that power comes from having a voice, but they’re very disinclined to share their opinions unless they’re extremely knowledgeable, or they don’t care about what other people say. For example: “You can go on Facebook and tell people you don’t like surveillance by the NSA, but it makes you sound like a jackass. No one likes those people and no one respects those people.” Across the board, people were unwilling to enter the debate: “I won’t voice them unless it’s absolutely necessary.” People view politics as a source of contention and they don’t want to engage in those debates with friends or even others online.

Participants tended to believe that they had the most power at a local level, either because they felt like they could more easily persuade others in their neighborhood, or because they felt proportionally more influential. One participant said, “I would say I don’t have power, but there are so many smaller scales I do have power, at the community-level.” But when you look at where they’re voting, they’re actually voting more in national elections.

When the researchers asked why they voted, participants wondered, “Does my vote really count?” but did so out of duty or hope, “I don’t know, but I feel like I should do my part.” People still vote out of routine or civic piety. And most often that voting happens at the national level, rather than at the local level where they will have more power.

Bystander Archetypes
John presents a set of eight user archetypes created by the team to support persona design at Google:

Neighborhood Stewards: feel they have the permission to fix something in their neighborhoods and are willing to address the problem, engaging with adversarial actors and authorities. They do good deeds locally and vote in national and sometimes local elections. They have an ideology about what the community could look like and are willing to act. They need information that they can use to be better connected and informed.

Kate offers two examples from Phoenix that fit this category. One gentleman is very concerned with security in his community. While he is ineligible to vote because of his past, every evening at 11pm, he goes around his area, checks that everything is okay, and investigates problems in the community: “I can handle myself, but … there’s plenty of senior citizen women in this complex that are naive enough to open the door and allow [scammers] in… I believe in looking out, especially for this community that I live in.”

Another gentleman was very involved in his homeowners association. There are guidelines about where people put their trash. He gets angry when people put their trash in the wrong place and has organized to inform people about the rules. He has even put their trash in front of their garage door so that they would run into it.

Nesters: focus on managing family life and property ownership. Their civic lives are already full with community activity. They are most likely to sign a petition or volunteer once, especially around issues that affect their kids. They’re motivated because they have personal and family interests at stake, or have prior interests with an issue. Being a parent or property owner requires a lot of work and often involves local government around permits and schools. Given all this work, the nesters may be receptive to more seamless information about how government works and what services meet their needs.

Careerists with a Cause: pursue professional ambitions; they look for crossover between career and civic life.

Meaning Makers: are driven by a defining personal experience; they perform service for others. An illustrative quote is, “I’d probably be somewhere else in life if I got involved in some program. I just had to take my own life experiences and talk to kids”

Transitioners: are dislocated in their careers or their residence, they seek community and fulfillment in civic life. An illustrative quote is, “I’m trying to get more involved in non-profit organizations.”

Young Ideologists: have a clear worldview motivating their actions.

The Committers: once they decide how to help, commit to a cause for emotional connection.

Careerists: first and foremost drive for mastery and success at work or school, which means little time for much else.

David Sengeh asks if these are roles that people stay in, or if they shift from role to role at different times in their lives. John replies: I don’t think it’s a hard and fast rule that these segments persist. When someone starts a family, they might transition into the interested bystander segment.

An attendee asks, “did you see any difference between signing petitions physically versus online ones?” Kate replies: When we asked about what action they took that was civic, they usually said it was a petition but they didn’t distinguish between online or paper.

An attendee asks about the exclusion of donations from the ladder of engagement. Did people without time give money? Kate replies: The fourth most common thing people mentioned doing was giving money; so it is definitely happening. In terms of how people responded to the question of civic engagement, they asked if we could define “civic.” It did not mean something specific to them. And donating money to political causes did not come out as a specific thing interested bystanders were doing.

Were barriers to participation specific to archetypes? Kate replies: These archetypes are useful for groking some research. They are not statistically representative, which is a disadvantage of using them. For disinterested bystanders, time was more of a factor. But because we were not looking at the statistical correlations, we can’t make firm statements about that.

Quantitative Profiles
Kate presents the results from the conjoint survey, where they attempted to get a more representative perspective on the qualitative findings. The team used the qualitative findings to construct surveys in hopes of linking these personas to the behavioral and attitudinal findings. The survey was meant to help reveal what was most important in people’s lives, not assuming that civics would be a high priority. Across roughly 2,000 responses, they used latent class analysis to find six groups that fell along a spectrum of most to least engaged:

  • 20.7% Community Active
  • 14.7% Neighborhood Advocates
  • 11.6% Vocal Opinionators
  • 22.6% Issues Aware
  • 15.3% The Absentees
  • 15.8% Civically Disconnected

Most people fell in the middle of this distribution—roughly aware but not active in civic life. Two groups scored high on engagement: community active and neighborhood advocates could be considered “civic participants” rather than interested bystanders.

Community Active people were likely to know their neighbors, interact with them frequently, engage in neighborhood activities, follow local issues, vote in elections, and keep tabs on the news.

Neighborhood Advocates are more focused on the local level. They don’t feel that career education is a priority. This could be someone retired. They felt that career or education were not a priority, were strongly family oriented and settled in their neighborhood. And they got involved in local things occasionally.

Vocal Opinionators had a very strong negative beliefs about government, have large news diets and a strong belief system, and make sure to vote in national elections. They’re not civically engaged locally, and they’re settled in their neighborhood but do not engage with neighbors or neighborhood activities.

Issues Aware have a strong disengagement with family and neighborhood. They feel that career or school are not a priority, but other obligations are more important than civic pursuits. More than all groups, they feel strongly that government provides useful services, pay attention to the news, and vote in national elections. This group does not know their neighbors and don’t feel settled.

The Absentees are the most family oriented of all groups. They are most likely to not be working or studying, and feel they do not have time for civic or community activities and avoid neighborhood interactions. They pay attention to the news but are not likely to do anything as a result. And they are unlikely to vote, feeling it won’t make a difference.

Connecting the Quantitative and Qualitative Research
Do these quantitative and qualitative findings overlap? Career is a defining driver of identification and behavior. Neighborhood and local stewardship is important to many people, but many are neighbors in name only and not settled in their neighborhoods. Many see sharing of opinions as a contentious act and worry about the resulting controversy.

Instead of trying to find patterns in the quantitative data, the researchers took an exploratory approach to show what was interesting, rather than trying an approach to confirm the qualitative work. Methodologically, they can only use their quantitative and qualitative approaches for explanatory purposes. They can’t make any claims about how their profiles apply across the population. The research takes a point in time snapshot of people’s lives—it’s not longitudinal research, and it doesn’t identify how or why people move from one type to another.

In spite of those constraints, Kate argues that this research identifies a portrait of American civic life that is valuable. And they want to encourage future research to connect the dots.

How Google Uses This Research Internally
After several years of focusing on election information, the Civic Information Team carried out this research to take another look at how might they help people be even more civically aware? The team knew that the least engaged people are by definition not even searching for information. Might Google Now be an important, relevant channel for providing people with civic information that they might not otherwise search for or find?

Google Now is a mobile platform on Android, also on iOS though less well-integrated, which serves up personalized information to users at a the right time based on information that users give Google permission to use.

Election Now Cards

This research has informed the overall goals and strategy of the Election Now Cards product. It helped select the interested bystanders as a target segment for the product. And the research gave the team the confidence to develop the cards.

Question and Answer
Eric Gordon asks: You don’t talk positively about the internet or connective technologies in your definition of civic activities. Is there room for other modes online in which people are already engaging, and how does that inform how this research will be useful to others in the future?

Kate responds: We wanted to look into an aspect of civic behavior that many others don’t have the resources to study. We didn’t want to go in with an assumption that people are already engaging in meaningful ways online and simply look at digital traces. For instance, the question about doing something they recently did which was civic was meant to be open and then they would dive into whatever they mentioned. Our focus was on people who were aware but not particularly involved. That said, Google is obviously a tech company. We asked, “What does this mean in terms about how we or others might design technology products?” The challenge posed by this research is that we need to use design to help people feel like their own interests are aligned with the public interest.

Charlotte responds: There are two major barriers we heard about. The intractable barrier is that the civic ladder is a deep historical memory/expectation about how citizenship is conducted. People sometimes overtly or more implicitly kept referring to things like the Civil Rights movement or the Arab Spring as the ways that change come about. The Internet is new and doesn’t yet feel like that great tradition. The shame that people feel about not doing enough is linked with their sense that Internet things aren’t enough.

The more tractable barrier is that people really want to do things that they know matter. It’s easier to see the effect on change you create with your neighbors in a local context.

Saul Tannenbaum asks: In the range of civic activity both you and Facebook are looking at interventions in elections. I have converted from interested bystander to deeply civically engaged, and I see voting as one of the least important things I do. So why would Google Now intervene on elections when the range of activities you have listed is wider and arguably more valuable? Is it because that’s where the data are?

John responds: Google has been working on elections for a long time and has an elections team. But as you can see, the participants were talking about a broader notion of civic life—it goes way beyond elections and voting. We’re thinking about what this means and what we can do to help people better engage on the community or neighborhood level.

Alex Howard asks: One of the times Google civically engaged the most people in your history is when you put a link on the Google homepage for people to call their congressperson about the Stop Online Piracy Act. Do you talk to the rest of Google about that kind of thing? Are you thinking about putting links in people’s pages telling them about things happening in their area?

John responds: Whether it’s the home page space or Google Now, which might be the better place for what you are talking about with local meetings, this is something we are currently exploring but we haven’t launched yet.

Alex responds: Google Now is relatively small but the Google homepage is one of the top 10 most visited pages in the world.

John responds: That’s a question that is well beyond my pay grade.

Kate responds: Google is not the only actor working in this civic engagement and technology ecosystem. Google is a key place that people go for information, so there is a scale effect at Google. But I think it takes a whole community of actors to do something meaningful.