I just spent a fascinating few days at the Code for America summit where we had the opportunity to learn about many projects occurring at the intersection of technopole and civitas: the conceptual space where a tech-enabled government and its people meet.

Civic tech aims to solve the problems facing the civitas, the people, who are still the source of the legitimacy for all the moral claims that we’re making on volunteers, on philanthropic funders and on governments. But while the moral argument about public benefit continues to inspire many people’s participation, there was an odd lack of specificity about a few important concepts that are really central to the endeavor. As a member of the civic tech community, and after three intensive days of listening and talking, I still don’t feel I have answers to a few important questions:

• What are the main problems we’re solving?
• Who is this public we’re benefiting?
• What exactly is “the civic tech approach” to this public’s problems?

These are critical questions for our community, and we should demand and create more conceptual clarity around them. Here are a few preliminary thoughts to put us in that direction.

What’s a civic problem?

You can’t talk about “problems” without evoking “solutions,” and “solution” is a word with an important double meaning in the context of civic tech. While a “solution” is something that resolves a problem, “solution” is also popular marketing euphemism for computer software. IBM may sell business “solutions”; perhaps our work is to sell civic “solutions.” Happily, I heard an admonition from the Code for America summit stage that we should not push apps ahead of the need for them: that we should avoid being “a solution in search of a problem.” And yet this only tells us what to avoid doing — and not what, affirmatively, to do. How do we identify a civic problem? And, more to the point, how do we know it’s a real and meaningful problem that’s worthy of attention and investment?

We could consciously explore the wisdom of other fields. We could draw on philosophy and take a utilitarian approach to identifying real civic problems: ranking potential problems by severity, determining how many were affected and calculating the total amount of harm caused by one problem.

A less structured approach could also address the same thing by asking: Of the problems we’re looking to address, which of these fulfill civic wants, and which of these fulfill civic needs?

While there is no absolute definition of “need” versus “want,” this utilitarian shorthand reminds us to connect to what is essential. Psychologist Abraham Maslow’s description of a “hierarchy of needs” provides another way to think about categories that are fundamental and prior — the physiological needs of food, water, shelter and physical safety — and those that are additive and secondary, like self-esteem and self-actualization.

If a civic tech community is focused on fulfilling “wants” while it ignores “needs,” then we have to worry about whether the work is too trivial and that the community is not focused on addressing serious civic problems. For this reason, Code for America’s expressed interest in focusing more deeply on health, on public safety and justice, and on economic development (to the extent this means helping more people obtain decent employment) is a welcome development.

However, this focus will have to happen in the context of a particular constituency and approach in order for this targeted focus to achieve the promised solutions for problems.

Who is “the public” and how do we relate to them?

This is the question of defining our identified constituency. It’s a complicated concept and you can really define “the public” in an endless variety of ways. It all becomes even more complicated when we’re also trying to identify what kind of problems “the public” has, and then trying to figure out how the public and its problems relate to civic technologists and their problems.

On the summit stage it was observed that the audience of civic technologists is also “the public,” that we represent some of the citizens we’re seeking to serve. This is indeed true, although it’s clearly an incomplete answer. But why is it incomplete? It’s not really because the civic tech community fails to match the demographics of the national community, although that does point at the issue.

Rather, it’s incomplete because it leaves out a consideration of power.

Power can be defined as the ability to make others do what they wouldn’t otherwise do. Because the kind of power we’re talking about here is governmental power — which is amazingly substantial, since government is our official repository for the legitimate use of physical force — there is a really important difference between those who directly control that power and those who don’t. We can use “lack of government power” as a measure of the basic meaning of “the public”: those who are not in government.

“The public” are those who are the on the receiving end of public policy. Having delegated their primary power to elected representatives, they have no direct ability to change it. This definition of “the public” connects fully to the “public” in “public schools,” “public services” or “public transportation” — a public that is a recipient of what a government has decided to provide.

Here’s a corollary to this definition: The further people are from being able to affect the decisions of those representatives, the less connected to government they are and the more fully they become “the public.” We already sense that the meaningful “public” is located at some distance from government. For example, some summit presenters identified a need for civic technologists to go out into “the community,” with “the community” representing a close cognate of “public” in the way I mean it, in the sense that one “goes out” from the policy-implementing place into it. It’s a community that is distant from and does not itself come to government.

And how does this “public,” in its power-distant state, relate to us? The distribution of power – and the benefits that flow from it – is the essence of the practice of politics. According to political scientist Harold Lasswell, the central function of politics is to distribute power: to determine who gets what, how and when. Government is then the mechanism by which these politically-determined benefits are doled out. Realistically, despite our interest in “reinventing government,” there is no possibility at all that we will be disrupting this basic governmental function.

At the same time, it is critical that we realize that joining with the work of government means assuming a form of power ourselves — and leaving “the public” in an important way. By implementing tools to smooth or channel the work of public policy, we are joining in the work of allocating benefits.

Civic tech and its tools serve either to reinforce or redistribute existing lines of power. It’s incumbent on us to contemplate just how we interact with that.

To explore what I mean by this, let’s talk about “efficiency.” Though it might sound like an uncontroversial general good, governmental “efficiency” is not neutral. Its effects depend on which policy is being made more efficient. The efficient implementation of an unjust or punitive policy means an increased amount of injustice or punishment.

This insight is especially important when we recognize that when we describe becoming “efficient,” we inevitably aren’t tightening down all the screws at the same time. It is a well-known observation that laws are selectively implemented and enforced — and in fact, many of us depend on that. Are we actively seeking the perfect enforcement of traffic speed limits? How about penalties for marijuana possession?

So if not all laws are fully implemented, then which are the first screws to be tightened when we become efficient? Often, the ones on those farthest from power, who are least able to complain about it effectively.

When we improve the speed or completeness with which we process the end of a public benefits program, that is substantively less benefit that a particular person will receive. One summit presenter expressed a quiet concern that if Detroit improved the integration of their data about which people had no water because of an inability to pay water bills, that the city would be quicker to take away those people’s children. When data erases the spaces and strategies that people have developed to cope with surviving under conditions of scarce resources — and does not at the same time create additional support for those people — governmental efficiency can materially harm “the public.”

Thinking about how we become closer to government and further from public through the power of our work raises important questions for civic technologists. Whose work are we smoothing or making more effective first? What is the order of attention to problems, and who is advantaged by that order? How is our work political, in the sense of affecting the relative distribution of government goods?

If we seek to benefit the public, we need to think about whether our work reinforces existing power relationships or redistributes power. Does our work help to get people who are distant from power an obviously better claim on it? Does it improve their ability to get their real needs met by giving them additional control over government? The more the beneficiaries of our work are currently distant from government power, and gain more power over government through our work, the more our work can genuinely be considered to benefit the public.

What is the “civic tech” approach to solving civic problems?

A third question for us to ask is whether there are consistent elements describing a “civic tech” approach to civic problems. What distinguishes the use of civic tech for social benefit, as opposed to all other methods of achieving social benefit? Including volunteer (“civic”) labor in software development? Centering the civic solution around a new piece of software? The simple quality of using a new piece of software in the civic solution, while simultaneously including some organizing element?

Though achieving agreement on our problems and constituencies is essential for aligning our community’s goals, identifying our common methods would help increase our joint impact. Allow me to assume that if you’ve read this far you agree with my definition of problems and constituency. If that’s the case, you probably agree with the following as well: If we are focused on addressing the high-priority needs of power-distant people, then we have just set ourselves an enormous challenge. We therefore really need to get more thoughtful about the ways civic technologists go about solving civic problems.

Here is an example of what I mean. The series of conditions and events that collectively describe what happened in Ferguson, Mo. concern high-priority personal safety issues and a number of power-distant people, and thus represent an appropriate focus for the civic tech community. How do civic technologists increase the power that the public has to affect and improve their outcomes in this situation?

At the summit, the solution that was discussed in connection with the Ferguson case was a messaging app that reminded you to attend your court date and to encourage you to pay your fine online.

This is the kind of approach that aims to create small change at the margin rather than addressing deeper, causal issues. It does nothing to address the problem of whether it was appropriate for those citations to be issued in the first place. It does nothing to address the racial tension that leads police to spiral so quickly into escalation. It doesn’t inquire into the nature of civilian oversight over police behavior, it fails to involve municipal funding conditions that lead police to rely on fines as a revenue stream, and it does not increase avenues for appeal and feedback from the public.

This is certainly not to dispute the fact that people should be able to pay their fines and citations online. They absolutely should; making it easier for them to do so is a positive social contribution and a genuine improvement. It’s just to say that this is in no way enough of an improvement given the scale of the problem, and it should be widely understood that this is not the place where civic tech stops.

Rather, civic tech solutions must come to be characterized by marginal changes that are arrayed and coordinated in such a way that they help effect systemic change.

Marginal change contributes to broad, systemic change when the problem and the constituency — as well as the need for further improvements — are all clearly identified. It requires outreach to the affected members of the public in order to determine what marginal changes would be both useful and practical. It requires an awareness of the multiple dimensions of the problem, and that the technology itself can affect some but not all of those dimensions. Change at the margin contributes to systemic change when the apps we develop form part of a comprehensive approach.

To move more deeply into a systemic change approach, civic technologists need to consciously value non-technical skills. Skills in cross-cultural communication, in facilitation, mediation, legal defense (and offense!), and public writing and argumentation are all necessary for achieving deep and systemic change. Coordinating software development with these other activities amplifies the effects of the software and also amplifies the effects of the organizing. Civic tech as a “software plus organizing” effort will bear the most fruit. One pointed — and excellent — question at the summit asked:do we need to teach our civic technologists how to organize or our organizers how to code? It may not matter so long as these skills are understood to be complementary, not substitutes for one another.

The desperate need for civic technologists to pair with people who have a wide variety of organizing-relevant skills becomes particularly acute when we recognize that our core constituency is the power-distant. Many people who are distant from power are also distant from technology. (This is a somewhat self-reinforcing loop: Technology is actually a pretty decent way to access power, if you haven’t noticed!) It was moving to watch a leader from Allied Media describe Detroit’s Disco Tech program, where residents teach each other immediately useful skills — such as teaching older people to text in order to communicate more easily with their grandchildren. It was useful to be reminded that immigrant communities might be more likely to respond to radio spots than text messages. In at least one panel, we celebrated the enduring power of a well-placed bulletin board. We saw at least two excellent examples of policy work, organizing and software development working together in Chicago: the Large Lots program that allowed neighborhood residents to gain legal ownership of the vacant lots near their homes, and the Expung.io program, which provides an easier interface for juvenile record expungement. In both Chicago cases, the app followed initial community and policy work, and the software now provides a persistent backbone that works to encourage further policy evolution. In addition, both of these projects redistribute economically meaningful resources to power-distant groups: Large Lots lets local residents of working-class neighborhoods obtain land for $1, and Expung.io frees people with juvenile records to seek a more lucrative employment future.

The true meaning of the word “technology” encompasses all useful tools, not only the digital. Since this isn’t an automatic insight for the civic tech community, it is especially vital that we remain conscious of the range of necessary skills when developing our strategies. To solve problems with the power-distant public, digital work must be put in service to information gathered through use of “soft skills” like training, like facilitating, outreach, and liaising. People who are not comfortable with technology often do not feel empowered by interacting with it — but everyone feels empowered when they are listened to or taught new skills.

But what if I don’t have a problem with the way power currently works?

Although I know many share my belief that the point of civic tech is to serve the real needs of a power-distant public, I also know this opinion isn’t universal. Based on an unscientific survey of the noises made by my seatmates during the summit plenaries, I know some people seem more excited about projects that seemed to provide an aesthetically-pleasing satisfaction of civic “wants.”

I can understand the logic of this. “Wants” can certainly seem more real than “needs” because government workers are more likely to hear them. “Wants” tend to be expressed by people who are both comfortable – and comfortable working with government. Projects that aim to further the seamlessness of the government experience for this community mean more pleasant interactions with people who have a demonstrated capacity to make themselves heard. Furthermore, these wants likely resonate deeply within the civic technology community as it’s currently constituted. As an educationally and economically privileged community, we are more likely to have experienced these wants ourselves.

Reinforcing this tendency, the problems that spur “wants” are comparatively so easy to solve! “Wants” describe lack of delight, lack of novelty, lack of smoothness. Someone with “wants” does not seek to disrupt the system – merely to make it run more smoothly and efficiently. These are problems for which a simple software solution may in fact be a quickly implemented, actual solution.

To these people I say: We can be so much more than that. We can be bigger than Band-Aids. The gap that currently exists between our awareness that governments are not serving all citizens and the marginal “smoothing” effects of many of our new civic tech products stems from our failure to think about how these products work with the distribution of power.

Where products help redistribute government attention by improving a government’s focus on the expressed desires of the less powerful members of the civitas, civic tech is disruptive in just the way we all hope it will be: It provides a new way to make government more responsive to the constituencies who are currently shut out.

However, civic tech runs the risk of failing to achieve its transformative potential if the community involved in creating it does not pay attention to how their products work with existing power structures. We need to acknowledge that saving an individual a few minutes of time may not truly benefit them if it reinforces their distance from decision-makers. Lumping together civic wants with civic needs as equivalent public goods diminishes the humanity of those with real needs while it reinforces existing privilege.

This is a problem that we can avoid. But in order to achieve our larger goal of social transformation, we’re going to need to show — and talk about, and plan for, and coordinate around — how we will achieve it, in a more conscious and specific sense.