A Robustness Principle for Government

President Obama’s memorandum calling for transparent, participatory, collaborative government is also just a statement of philosophy (see the Appendix A). But it’s a statement of philosophy that’s fundamentally actionable in the same way that the TCP robustness principle was, or the design rules that are the heart of Unix. And even though none of these things is a formal specification, it is a set of design principles that guide the design of the platform we are collectively trying to build.

It’s important to think deeply about what the three design principles of transparency, participation, and collaboration mean in the context of technology.

For example, the word “transparency” can lead us astray as we think about the opportunity for Government 2.0. Yes, it’s a good thing when government data is available so that journalists and watchdog groups like the Sunlight Foundation can disclose cost overruns in government projects or highlight the influence of lobbyists (see Chapter 17). But that’s just the beginning. The magic of open data is that the same openness that enables transparency also enables innovation, as developers build applications that reuse government data in unexpected ways. Fortunately, Vivek Kundra and others in the administration understand this distinction, and are providing data for both purposes.

Likewise, we can be misled by the notion of participation to think that it’s limited to having government decision-makers “get input” from citizens. This would be like thinking that enabling comments on a website is the beginning and end of social media! It’s a trap for outsiders to think that Government 2.0 is a way to use new technology to amplify the voices of citizens to influence those in power, and by insiders as a way to harness and channel those voices to advance their causes.

Participation means true engagement with citizens in the business of government, and actual collaboration with citizens in the design of government programs. For example, the Open Government Brainstorming conducted by the White House is an attempt to truly engage citizens in the making of policy, not just to hear their opinions after the fact.[27]

Open government APIs enable a different kind of participation. When anyone can write a citizen-facing application using government data, software developers have an opportunity to create new interfaces to government.

Perhaps most interesting are applications and APIs that allow citizens to actually replace functions of government, in a self-service analogue to Craigslist. For example, FixMyStreet, a project developed by UK nonprofit mySociety, made it possible for citizens to report potholes, broken streetlights, graffiti, and other problems that would otherwise have had to wait on an overworked government inspector. This concept has now been taken up widely by forward-thinking cities as well as entrepreneurial companies like SeeClickFix, and there is even a standard—Open311—for creating APIs to city services of this kind, so that third-party developers can create applications that will work not just for one city, but for every city.

Taking the idea of citizen self-service even further, you can imagine government using a platform like Meetup to support citizens in self-organizing to take on major projects that the government would otherwise leave undone. Today, there are thousands of civic-minded meetups around issues like beach, road, and waterway cleanups. How many more might there be if local governments themselves embraced the idea of harnessing and supporting citizen concerns as expressed by self-organized meetups?

Do It Ourselves: An Example from Hawaii

One of the most dramatic contemporary examples is a story reported by CNN, “Island DIY: Kauai residents don’t wait for state to repair road”:[28] “Their livelihood was being threatened, and they were tired of waiting for government help, so business owners and residents on Hawaii’s Kauai island pulled together and completed a $4 million repair job to a state park—for free.”

Especially striking in the story are the cost and time savings:

“It would not have been open this summer, and it probably wouldn’t be open next summer,” said Bruce Pleas, a local surfer who helped organize the volunteers. “They said it would probably take two years. And with the way they are cutting funds, we felt like they’d never get the money to fix it.”

And if the repairs weren’t made, some business owners faced the possibility of having to shut down….

So Slack [owner of a kayak tour business in the park], other business owners and residents made the decision not to sit on their hands and wait for state money that many expected would never come. Instead, they pulled together machinery and manpower and hit the ground running March 23.

And after only eight days, all of the repairs were done, Pleas said. It was a shockingly quick fix to a problem that may have taken much longer if they waited for state money to funnel in….

“We can wait around for the state or federal government to make this move, or we can go out and do our part,” Slack said. “Just like everyone’s sitting around waiting for a stimulus check, we were waiting for this but decided we couldn’t wait anymore.”

Now is the time for a renewal of our commitment to make our own institutions, our own communities, and our own difference. There’s a kind of passivity even to most activism: collective action has come to mean collective complaint. Or at most, a collective effort to raise money. What the rebuilding of the washed-out road in Polihale State Park teaches us is that we can do more than that. We can rediscover the spirit of public service, and apply the DIY spirit on a civic scale. Scott Heiferman, the founder of Meetup.com, suggests going beyond the term DIY (Do It Yourself) to embrace a new spirit of DIO: Do It Ourselves!

Citizen self-organization is a powerful concept. It’s worth remembering that early in our nation’s history, many functions now handled by government were self-organized by citizens: militias, fire brigades, lending libraries, not to mention roads, harbors and bridges. And even today, volunteer fire departments play a major role in protecting many of our communities. Traditional communities still perform barn raisings. Those of us who spend our time on the Internet celebrate Wikipedia, but most of us have forgotten how to do crowdsourcing in the physical world.

Everyone Has Something to Offer

The reflex exerted by government to gather new information, whether in pursuit of spreading around money for housing or planning its next steps in Afghanistan, is to convene an advisory committee of experts. A whole set of laws and regulations, such as the Federal Advisory Committee Act (FACA), controls this process. Such panels are typically drawn from a limited group of academics and industry experts. A list of these advisors would no doubt show a familiar pattern of high-ranking universities.

Recent popular research on crowdsourcing and the wisdom of crowds suggests a totally different approach. Asking everybody for input generates better results than just asking the experts. Certainly, a single recognized expert will tend to offer better facts, predictions, or advice than a random individual. But put a few dozen random individuals together—on the right kind of task—and the facts, predictions, or advice that shake out are better than what the experts alone produce.

The reasons behind the success of crowdsourcing are still being investigated, but the key seems to be this: in a mix of right and wrong answers, the wrong ones tend to cancel each other out, leaving the right ones. This is the secret behind the famous appeals to the audience in the game show Who Wants to Be a Millionaire, as well as the success of prediction markets such as the University of Iowa’s Electronic Market.[29]

Wikipedia, which invariably makes a central appearance in every reference to crowdsourcing, plays the different opinions of the crowd against each other in more explicit ways. On relatively uncontroversial articles, contributors are expected to discuss their differences and reach consensus. This process is aided by a rarely cited technical trait of web pages: because they present no artificial space limitations, there can always be room for another point of view. On controversial topics, Wikipedia has over the years developed more formal mechanisms, but the impetus for change still wells up from the grassroots.

It’s also worth mentioning, in regard to crowdsourcing, the use of low-paid or volunteer labor to carry out simple tasks such as identifying the subjects of photographs. These are called Mechanical Turk projects, in reference to a crowdsourcing technology platform provided by Amazon.com, which is itself named after an eighteenth-century hoax[30] in which a person pretended to be an intelligent machine; in the modern incarnation, thousands of people are serving as functions invoked by a computer application.

Crowdsourcing has already slipped into government procedures in low-key ways. Governments already use input from self-appointed members of the public on all kinds of things, ranging from reports of potholes to anonymous tips that put criminals behind bars.

One of the key skills required of both technologists and government officials is how best to aggregate public opinion or data produced by public actions to reveal new information or patterns. For example, cities learn a lot about neighborhoods by aggregating crime reports from residents. They could understand their needs for broadband network access much more accurately if they took resident reports into account and didn’t depend just on what the broadband vendors told them (because geographic anomalies often cause dead zones in areas that the vendors claim to serve).

In general, people can provide input on several levels:

Some of those applications may operate on existing government data, but they can also be designed to collect new data from ordinary people, in a virtuous circle by which private sector applications (like SeeClickFix) increase the intelligence and responsiveness of government.

Governments are more likely to use some form of filtering than to rely on public consensus, as Wikipedia does. The combination of free debate among the public and some adult supervision from a government official makes a powerful combination, already seen in the open government brainstorming session mentioned in Lesson 3.

Finally, crowds can produce data without even realizing it—implicit data that smart programmers can collect and use to uncover whole worlds of information. In fact, smart programmers in the private sector have been doing that for years. Lesson 5 covers this trend.

—Andy Oram


[27] http://www.whitehouse.gov/blog/wrap-up-of-the-open-government-brainstormingparticipation/

[28] http://www.cnn.com/2009/US/04/09/hawaii.volunteers.repair/index.html

[29] http://www.biz.uiowa.edu/iem/index.cfm

[30] http://en.wikipedia.org/wiki/Amazon_Mechanical_Turk