You are not logged in. LOG IN NOW >

Civic Tech and Engagement: In Search of a Common Language

BY Micah L. Sifry | Friday, September 5 2014

Marten van Valkenborsch, Construction of the Tower of Babel (c. 1600)

This past Wednesday, we hosted a fascinating conversation with three innovative thinker-doers: Ben Berkowitz, the co-founder of SeeClickFix; Marci Harris, the CEO of PopVox; and Erhardt Graeff, an MIT Civic Media grad student who is working on a project called ActionPath.

The conversation was the result of my hearing Graeff's short presentation on his research at last June's MIT-Knight Civic Media conference. He's focused on how we can improve the design of projects that aim to increase civic engagement, and envisions ActionPath as a mobile tool that would connect people with relevant civic information that literally crosses their path as they go through their daily lives.

Imagine an app on your phone that knows your location and sends you push notifications as you walk to your bus-stop like "there's a proposal before the city planning commission to put up a skyscraper here; click here to register your opinion." Then, after giving you a short survey to fill out, the app invites you to subscribe to updates about the status of that proposal. In short, Graeff's idea seeks to reduce the friction associated with getting informed and involved in local governmental processes, using open government data to jump-start the conversation and giving government agencies (or other civic actors) a new way to listen to the public's concerns.

As I heard Graeff describe ActionPath, I immediately thought of similar projects that he could learn from or even collaborate with, and at the top of my list was SeeClickFix, because it too creates a two-way channel for communication about local issues, and PopVox, because it too aims to channel relevant information to voters and from voters to their representatives.

Usually, when that happens (these days, fairly often), my next move is to make a bunch of email introductions, but this time I thought--why not see if all three folks would be willing to talk together about how to design for online engagement, and to do that as a Google Hangout on Air that others could watch? Luckily, everyone was up for it and the result was yesterday's more-or-less impromptu panel discussion (or, as Matt Stempeck, Microsoft's newly-installed NY director of civic tech, put it, "radio hour"), which you can view here.

Happily, not only was it a lively discussion, but some great folks joined in the comments, like Steven Clift of E-Democracy, whose knowledge of local online-enhanced neighborhood forums is extensive; Noel Hidalgo of BetaNYC, whose civic hacking experience in NY is second to none; the ubiquitous Alex Howard, aka @digiphile; and Lucas Cioffi, who has toiled for years in the deliberative democracy trenches.

One really cool outcome of the hangout was how quickly Graeff and Berkowitz saw a way to potentially work together: the SeeClickFix API should make it possible to run an "If This Then That" query against someone's location to allow ActionPath to surface relevant local issues. We'll keep an eye on that! (Also, don't miss Berkowitz's semi-humorous proposal for a "Look up, asshole" app that would alert cell phone users to stop texting while walking through busy intersections; we could use something like that here in New York City!)

But I have two additional takeaways from the conversation. The first one is small: we should keep doing these kinds of informal chats. While Google's Hangout on Air is a bit confusing to navigate, next time we can iron out the issues that kept me from seeing the chat comments, and we'll also come up with a hashtag to better tap the Twitter-stream. Where possible, we'll tie this in with pieces that we're working on as part of our "Civic Tech and Engagement" series. Stay tuned for details; my thinking is we'll go for something like this once or twice a month and keep the format fluid.

Towards a Civic Tech Taxonomy?

The second takeaway is big, and is actually the main subject of this post: We need much clearer language to describe civic tech. During yesterday's conversation, I noticed all of us struggling to put into words what it was that we were striving for. How would we know if something like ActionPath is succeeding? Is the purpose simply to increase the amount of citizen feedback decision-makers get? Is it to convert a little bit of the spare time that ordinary people "waste" on their daily commute into a civic good? Is it to draw such people into more consistent civic engagement? Is it to increase the voice and capacity of ordinary people to affect the decisions that affect their lives? Is it to change the balance of power that runs a city?

It's not enough to assume that, like the Supreme Court and obscenity, we know good civic tech when we see it. And if we can't say why something is good (or even great), how can we know what to design for? Indeed, how do we even know if we're after the same design goals?

I'm hardly the first person to raise these questions. Indeed, others with years of experience making and studying civic tech have themselves voiced frustration with our field's ill-defined parameters. At the same time, some very useful forays at definition are worth reviewing and building upon.

Back in April 2013, Tom Steinberg, the longtime director of mySociety in the UK, wrote a post called "What should we do about the naming deficit/surplus?" He noted that it was fairly easy to characterize organizations like Greenpeace, Oxfam, Human Rights Watch, and Facebook, but asked,

"What primary movement or sector is mySociety a part of? Or Avaaz? Or Kiva? Or Wikileaks? When I ask myself these questions, no obvious words or names race quickly or clearly to mind. There is a gap – or at best quite a bit of fuzziness – where the labels should go."

As he cogently argued, "…this worries me because consistent names help causes to persist over time….Flighty, narrowly used language confuses supporters, prevents focus and is generally the enemy of long term success."

In June, participation and democracy researcher Tiago Peixoto chimed in with support for Tom's overarching question, and then Nathaniel Heller of Global Integrity offered the beginning of a response, with a "quick and dirty taxonomy/tagging system" that uses letters to summarize key attributes. Heller's outline looked like this:

A: Involves citizens in the policy process
B: Involves citizens/beneficiaries in monitoring service delivery
C: Relies on structured information to inform decisions (whether digitized data or other information)
D: Leverages technology (can be broken down into additional sub-categories, software, hardware, web, mobile, etc.)
E: Makes previously hidden, inaccessible, or opaque information more public
F: Empowers citizens/beneficiaries to better hold service providers to account
G: Democratizes previously elite processes (e.g. aid, policy activism)
X: Traditional public sector service/transaction or other traditional service provider

While this list of attributes is obviously relevant to the field, Heller's tags don't quite help us figure out if something is good civic tech. At best, they just unpack the definitional problem for somewhat easier analysis. Thus, to Heller, "e-government = D + X" and "Gov 2.0 = C + X (and sometimes D)" and "e-participation = A + D." And so on. How and why these kinds of combinations might be desirable or improved upon isn't really addressed.

Steinberg re-entered the conversation in August with a post suggesting that the main thing uniting the entire sector together was "serving people's need to obtain and deploy power"--hence, he suggested "civic power sector" as the phrase to define the field. This isn't bad, if you ask me, though the word "power" is actually rarely uttered because so many people are afraid of it. Even though it's a neutral term! So, while I'm 100% with Tom, I'm partial to calling this whole sector the "civic technology" sector, since in its most generic sense"technology" just means tool or process; today, most of us understand it to mean "new networking tools and processes."

Tom went on to helpfully segment the whole sector into four parts, which I think is where his intervention starts getting really helpful:

  • 1. Decision influencing organizationsthat try to directly shape or change particular decisions made by powerful individuals or organizations.

  • 2. Regime changing organizations that try to replace decision makers, not persuade them.

  • 3. Citizen empowering organizations that try to give people the resources and the confidence required to exert power for whatever purpose those people see fit, both now and in the future.

  • 4. Digital government organizations that try to improve the ways in which governments acquire and use computers and networks. Strictly speaking this is just a sub-category of ‘decision influencing organization’, on a par with an environmental group or a union, but more geeky.

Tom also included a list of organizations sorted by sub-sector (the old, "I know what it is when I see it" strategy).

Now we're beginning to get somewhere. Groups that seek to change how government works are different than groups that aim to exert influence on decisions by government and other powerful actors. And groups that aim to empower citizens generally are different than those that seek to win more power for one group of citizens.

But what makes mySociety's approach to citizen empowerment different and perhaps preferable to Nextdoor's approach? Why do people keep coming Netroots Nation conventions while Organizing for America seems dead in the water? What makes Anonymous so different from Avaaz? Tom's civic power chart doesn't yet help us make those distinctions clearer.

Another MIT Civic Media grad student, Nate Matias, took a slightly different approach to the naming problem with a long post in July 2013 on the MIT Civic Media site. For him, the framing question was how to improve the conversation "among those who are trying to use the Internet to organize people for good." His post, which he titled "How to choose the best way to create change," was more of a catalog of pointers to people searching for answers than an answer itself. (Very meta.) But bookmark it nonetheless; Nate's list is pretty comprehensive. And by adding those words "for good," Nate signifies that in our discussion of civic tech, we're not neutral. We're in this for reasons.

Defining civic tech

To me, civic tech ought to be seen broadly as any tool or process that people as individuals or groups may use to affect the public arena, be it to gain power, influence power, disrupt power or change the processes by which power is used. In the broadest sense, civic tech includes centuries-old methods like voting, petitioning, leafleting, lobbying, donating, writing letters to the editor, attending public meetings, publishing news/opinion, demonstrating and civil disobedience. It also includes all the new forms of public participation, like blogging, commenting, sharing media, etc. But for today's purposes, I think there's a general agreement that when we refer to "civic tech" or "the civic power sector" or "using the Internet to organize people for good" we're talking about the new environment of radically cheaper networked communications, sensors and collaboration. In general, most people don't think of those older forms as "tech," they're just like wallpaper, part of the background (indeed, a new technology has truly become commonplace when people stop making a special effort to refer to it). Voting, writing letters to the editor, going to public meetings--that's all just considered civics. So I think if we want to define this field, then calling it "civic tech" puts us on solid ground.

But in order to helpfully draw distinctions about different kinds of civic tech uses, now we have to talk about what it's for. The sometimes unspoken and naive assumption is that because new networking technologies are so much cheaper to use and the barriers to entry are lower, greater empowerment is the natural outcome. Tools and processes that seem locked in stone (the Constitution hasn't been amended in decades) are, seemingly, much more open for reinvention when all that it takes is changing a few lines of software. And thus, in the earliest days of new tech's application to civic engagement, we got some very starry-eyed notions of where this was going. (In truth, going back to the earliest days of the Internet, people like Lawrence Lessig and Mitch Kapor warned that, respectively, "code is law" and "architecture is politics." In other words, tech in and of itself doesn't change power in a particular direction--it's the choices designers make that invest technologies with politics.)

For example, at the December 2004 conference on “Votes, Bits and Bytes” that the Berkman Center for Internet & Society held at the Harvard Law School,’s founder Scott Heiferman fantasized openly about the “Napsterization of organization.” He imagined that in the same way Napster had turned every kid with a computer into a music sharer, the Internet was about to generate a flowering of distributed democracy: “We’re in the emergent organization business,” he said. “From flash mobs and brick associations to flash, emergent, people-powered long-lasting open, influential, agile, chapter-based institutions and organizations that have card-carrying members,” he declared, “Net-based collective action and group power has hardly begun.” Here's a grab from my notebook, where I quickly copied a slide that Scott had flashed on screen during his talk:

But while Heiferman may have been carried away a bit (or just a tad ahead of the curve) during that talk, let's stop and zero in on that two-dimensional grid. Because here's where we can begin to usefully draw some distinctions about kinds of civic tech, and the design choices that come with them. The horizontal axis is "form of membership" (ad-hoc, fleeting vs formal, long-standing). The vertical axis is "degree of familiarity" (in-person, face-to-face vs virtual or by proxy). Using this framework, we can begin to separate an Avaaz from an Anonymous, or a Pirate Party chapter from an Obama for America group, and begin to sort out what kinds of organizational forms are enabled by different forms of civic tech.

Wrestling with this same set of questions, last summer Ethan Zuckerman, the director of MIT's Center for Civic Media, proposed thinking of civic engagement in terms of this two-dimensional matrix.

Here, instead of focusing on the type of organization, we're looking at the different sorts of engagement. Thin kinds of engagement require little thought on the part of the doer; thick means the participant has to really consider and deliberate. Symbolic engagement is centered on voice, while impactful changes measurable outcomes. (In more recent work, Zuckerman has shaded this distinction, since certain kinds of symbolic campaigns aimed at raising the visibility of some issue and educating the public can be deeply impactful, over time, if they change cultural assumptions and norms of behavior.) Helpfully, he illustrates the four resulting quadrants: "clicktivism" would go in the top left, as heavily symbolic and thin; voting would go in the top right, as impactful but also thin; the Occupy movement's on-the-ground occupations of public squares would go in the bottom left quadrant, as examples of symbolic but thick kinds of engagement; and something like Occupy Sandy, that mobilized a huge number of volunteers to deliver tons of tangible relief to victim of Hurricane Sandy, would go in the bottom right quadrant.

Using Zuckerman's matrix, I think we can similarly start to sort different kinds of civic tech. The e-petition world of and would go in the top left quadrant of symbolic and thin civic engagement; whereas a story-sharing campaign like "We Are the 99 Percent" or "Planned Parenthood Saved Me" would go in the bottom left quadrant, of symbolic but thick. ActionPath, with its theoretical emphasis on quickly capturing people's feedback as they go about their day, might go in the top right quadrant, thin but impactful. While SeeClickFix, with its combined emphasis on persistent public attention and getting actual local problems addressed, would go in the bottom right, thick and impactful.

Now, try to abstract outward from those four concrete examples to more generic categories. I've taken a stab at an answer below. The four kinds of civic tech tools we have, depending on whether participation is thin or thick and whether action is symbolic or impactful, seem to be clicktivism, story-sharing, feedback filters or public squares.

So far we've considered the kinds of organization civic tech can support (virtual, face-to-face, ad-hoc, formal) and the kinds of action it can foster (thin, thick, symbolic, impactful). I'd like to offer one more matrix for consideration, that aims to tease out the degree of empowerment being produced. My premise is that, as with these other schema, there's a spectrum of possibilities ranging from tech that maximizes the user's power to tech that maximizes the aggregate's power, and forms of power that are short-term or small-scale and forms that are long-term and larger-scale.

A few years back I wrote a post that tried to get at the first issue, arguing that we needed a way to differentiate between campaigns that were "grassroots vs grassrootsy." Just because a campaign included a lot of people didn't mean it was grassroots. If it was led from the top, then it was just grassrootsy, I argued. With this matrix, I'm trying to make a more general argument distinguishing between tools and processes that vest more power in the hands of their users vs those that enable someone to form a larger aggregate that itself has power, but is not directed by its users as much as by its managers.

So, a presidential campaign (say, Obama for America or Ready for Hillary) that uses tech and big data to organize a base of volunteers and donors is aggregating a lot of power, but doing so for a short-term goal (the candidate's election) and not really giving individual users much agency. Whereas a hashtag campaign (like the recent #IfTheyGunnedMeDown response to Ferguson) vests more power in its users' hands--they decide if an issue catches fire--but it too only has a short-term life-span. Conversely, we can think of organizations that uses tech and data to create an ongoing community that, in the aggregate, has the power to influence public events, like the Netroots Nation community, which annual convenes several thousand progressive Democratic activists who found each other initially via the mega-blog, or the New York Tech Meetup, which has grown to some 40,000 members that meet frequently to share tech demos and occasionally flex some political muscle. The final quadrant, movement networks that are both long-term in scale and user-centric in design are scarce in the political arena (perhaps the LGBTQ and marriage equality movements fit here?), but examples might be the "Maker" movement in tech or the Burning Man festival, which now has many local self-organizing spin-offs.

The questions I want to end this post with: what kind of design choices in civic tech maximize particular characteristics of civic action? What sorts of tools and processes empower users vs aggregators? What choices tend to produce thicker or more long-term kinds of engagement? What kinds of civic tech do we need? What do we want?

Next: Wrestling with the calculus of civic engagement, mapping the field, pondering where civic duty comes from, and asking again: what do we need and want.

This post is part of our ongoing series on civic engagement and tech--what makes it thick, supported by a grant from the Rita Allen Foundation.