Looking for Civic Tech’s Important Problems

"Computers are useless, they can give you only answers." - Pablo Picasso


We are pleased to welcome Tom Steinberg as Civicist’s new senior contributing editor. Drawing on more than a decade of unique experience as the founding director of the United Kingdom’s mySociety.org and adviser to several British governments, he will be writing regularly for Civicist on current issues in the civic tech arena.

In my early 20s I worked in a government office that had a variety of thought-provoking statements printed on the walls. They were of a rather higher standard than the usual motivational platitudes, and a couple have remained knocking around my head into my 30s.

One was a quote from Pablo Picasso. It read:

Computers are useless, they can give you only answers.

Being much more techy than most of my civil servant colleagues I initially took this as typical prejudice by non-geeks against their technically-minded cousins. But as I grew older and I read much more about the history of science and innovation, I came to understand what Picasso was getting at: that the most impressive people are often successful because they ask the right questions at the right time.

Recently this idea popped back into my mind as I was reading and then watching a famous speech by the mathematician and computer science pioneer Richard Hamming—‘You and Your Research.’ In his speech he dwells extensively on the importance of working on ‘important problems,’ which he explains in this wonderfully lucid way:

Let me warn you, ‘important problem’ must be phrased carefully. The three outstanding problems in physics, in a certain sense, were never worked on while I was at Bell Labs. By important I mean guaranteed a Nobel Prize and any sum of money you want to mention. We didn’t work on (1) time travel, (2) teleportation, and (3) antigravity. They are not important problems because we do not have an attack. It’s not the consequence that makes a problem important, it is that you have a reasonable attack. That is what makes a problem important.

Inspired by this I took a bit of time over the holidays to think about what some contenders for the important problems in civic technology might be, at the moment. Caveat, Reader: these are really stabs in the dark.

The next time that a nation does a major post-crisis constitutional overhaul (think Ukraine or Cuba or Syria in the not-too-distant future) are there any digital ideas so significant that they should actually feature in the new constitution itself?

If, as has been frequently claimed, the affordances of digital technologies really do alter the possible nature of government and politics, we should expect to see important new ideas embedded in 21st century constitutions. But what are the digital ideas we might expect to see debated by constitutional drafting committees? What issues and principles are sufficiently-new, but also long-term, that they belong in these hard-to-change and critical documents? Do any new ideas actually merit inclusion, or has the world simply not changed that much after all?

Can the civic tech movement have any detectable impact on the level of openness and accountability of governments?

One of the things that excited me about civic tech from the very start was the possibility of smuggling public interest values into public service provision, disguised under the cloak of code and design choices. I believed that the moment where an offline digital service (or a rubbishy first generation online service) was replaced by a better one could be a moment where new levels of openness and accountability could be embedded in the day to day business of government.

You can see this in some of the design decisions that I drove in early mySociety products. FixMyStreet makes reporting problems much easier, but it also makes the status of those problems public. The ease of use drove traffic, which automatically led to more transparency over fault reporting, which was supposed to shift expectations about what a normal level of transparency was.

What preoccupies me now is the question of whether this embedding of transparency and accountability values hard into the code of digital public services is going to be widespread or not. It is entirely possible to make a service more usable without making it any more transparent. Was it naive to assume we could smuggle in a bit of better governance through the back-door of software supply? Or is this actually happening quite often?

What will it take for governments to create new public digital institutions that are about issues other than surveillance and security?

In Germany the post-war government created the BPB, a federal agency designed to give rigorously non-partisan information to prospective voters. It runs the truly huge Wahl-o-mat, the ‘Who should I vote for?’ website that trumps all others internationally in terms of reach and significance. The BPB is highly respected and widely trusted as being non-partisan: it has no parallel in the U.S. or the U.K.

This institution exists to keep Germany working—to keep it successful, peaceful, and prosperous. It is an example of an innovative public institution that doesn’t do anything as traditional as running schools or hospitals.

I’ve written repeatedly that I think we’ll see new public institutions emerge that have digital ideas at the core. The question is, where are they going to come from, and what confluence of events will precede them?

How do governments get good at providing digital services on a widespread basis?

The already-huge gap between the quality of digital services delivered by the private sector and the public sector is, with a few notable exceptions, widening every day. This ever-growing gulf threatens the very legitimacy of government as an idea.

Despite this threat nibbling away at the base of the social contract, most governments clearly see their digital service provision as ‘good enough’ and not a significant enough problem to justify any significant reorganisations of staff, power, budgets, or priorities.

This problem is particularly acute at the level of local government, where there’s a basic discrepancy between the cost of building good quality digital services, and the small size of the tax base.

So the important problem here is working out ways that governments can catch up in a scaleable, repeatable way, before it’s too late.

What kinds of civic features can the internet giants offer, and what will induce them to deliver?

There was no concept of the emergency telephone service 999 (or 911 for American readers) before there were telephone networks. But these days the provision of this civic feature is so key that it is a critical determinant of whether a service even officially counts as telephony or not. Television and radio brought with them public service concepts of adult-content watersheds, public service announcements, nuclear attack warnings, and a range of other civic features that were largely incidental to the bottom lines of the companies.

What are the equivalents going to be for the biggest, most mainstream digital apps? And through what processes will they arrive?

Your turn, Reader

I’m not really sure that any of the above definitely counts as an important problem. But I hope by writing this that readers will perhaps suggest their own ideas, or changes to what I’ve written that would get us closer to something significant. I look forward to seeing your ideas.

  • Forest Gregg

    Problems are another name for our desires. But, when we speak about problems, we speak as if it weren’t also talking about ourselves, but only describing something about the world that anyone can see.

    Sometimes, within a community that shares the same way of seeing the world, problems can be a very useful shorthand. If you are working at Bell Labs, you don’t need to preface every memo with an argument about why faster, cheaper, and less error-prone communication would be a good thing to have.

    I’m not sure civic technology is such a field. Within the civic tech community in Chicago, there have been two very different, competing ideas about the opportunities of our moment.

    The first idea is that information technology can 1. enhance the ability of governments to make good policy, by helping them get a more accurate understanding of reality and 2. improve their ability to deliver policy through digital services or better planning and monitoring of operations. This idea tends to assumes that policy should be made by professionals and that government is mainly full of smart people who want to serve the public.

    For subscribers to this idea, the main problems are

    * procurement reform
    * hiring practices
    * organizational transformation

    The second idea is that information technology can enhance the ability of the public to monitor and influence the institutions that are supposed to serve public interests. By first reducing the costs of accessing, managing, and analysing information, and by second using the rhetorical facilities of the web, more organizations, serving a wider set of interests, will be able to make more effective arguments about what Chicago should be doing. This idea tends to assume that policy should emerge from debate among interested parties and that governments are not always incented to serve the public.

    For these people, the main problems are

    * business models (since the organs of the press ar critical here)
    * access to information
    * effective communication technique
    * understanding the variety of publics and their problems (sometimes called user centered design)

    It’s a very live question which approach will be more useful. In Chicago, we basically running experiments in parallel. Some groups, like Tom Schenk team, DSSG, or Impact Lab are working on projects to improve policy making from the inside out; at DataMade we trying to enhance public debate. We’ll see.

    Across these two approaches, there are a few needs that we really do share.

    1. A need to deeply understand how policy works in Chicago
    2. Deep familiarity with available data on what is happening in the City
    3. Common data processing tasks (geocoding, record linking, normalizing and denormalizing data structures… etc)

    I apologize if this all seem very parochial, but I try to only write about things that I feel I somewhat understand. I do think there are real parallels to this basic split in other cities and levels of civic life.

  • loreleikelly

    thanks for this! it is a good overview of the different civic tech analytical themes ongoing re: Congress as well.

  • loreleikelly

    thanks for this article. I work on civic tech issues around the US Congress–re: your question about the private sector, what would be great is a link on the major search engines to do a “civic search” that shows information supply chain and conflicts of interest or allows for voluntary disclosure. In a post Citizens United world, our legislature is going to face increasing trends toward transactional decision making if we can’t create an alternative that wields new political currencies created by disruptive technology (i.e. location, timing and disclosure) The anonymous and unending money in the system is a huge distortion across the board for policymakers and staff. Money is now an access rule, but not necessarily a decision rule. We need to fully block this from happe
    ning, though I don’t know exactly how at this point. 2. the biggest and most important single step forward for Congress would be if the next Librarian of Congress is an innovator and an actual librarian. The Library of Congress should be a data fusion center for the most modern technologies and data analytics available for the practice of democracy. Fully and unabashedly public facing and public serving. No outside entity could do for Congress what it can do for itself.

  • jpmckinney

    There’s a lot more to explore, but I’ll share some thoughts on each question.

    To your first question on constitutions, the work of http://www.datasociety.net/ may suggest some possibilities – including possibilities of new public digital institutions that would address the issues Data & Society researches.

    On the second question about civic tech’s impact, I think it’s less about smuggling software into government and more about smuggling people. I think the UK and US governments provide good examples of that. This is also part of the answer to the fourth question on how governments get good at digital services.

    Regarding internet giants, there’s Facebook’s safety check and Google’s elections cards. The homepages of some giants are also used in a PSA capacity. It’d be useful to document existing efforts.

    As for new questions, I’m mainly preoccupied with technology’s (civic tech’s) influence on governance (your second question), and more specifically in what contexts is it a better idea to:

    1. deliver a service from outside government (e.g. operate a PMO website)
    2. demonstrate/validate a service for the government to implement/adopt (e.g. demonstrate the advantages of a specific consultation process)
    3. lobby the government to provide a service/implement a policy
    4. join the bureaucracy to provide a service/implement a policy
    5. become a minister or staffer to dictate policy

    I’m concerned about the sustainability of (1). I want to better understand when it’s better to pursue (2) and (3) instead of (4) or (5). This may be an important question related to yours.

  • Patrick Atwater

    Schools dude schools… kinda obvious that ubiquitous access to information affects learning paths though the structure of public schools been the same for decades and decades. Easy to say and lots of smart people working on it though seems like a good candidate for a “digitally native public institution” with 911-eque fundamental public services tied in. Many teachers might argue that donorschoose.org is a pretty foundational.

  • kristin wolff

    Wow @forestgregg:disqus. This is a super useful framing. I’m borrowing it for thinking about day-job work and fun with @hackoregon.