Better Government Through Crowdsourcing
Can Americans' collective wisdom solve our collective problems?
Approximately 220 million Americans are eligible to vote, but only 130 million or so are likely to cast a ballot next November. While the 90 million who fail to make a trip to the polls will inspire hot takes and handwringing about the public's apathy and its deleterious impact on democracy, what about the apathy on the other side of the transaction?
The United States is a network of 330 million people, all with different interests, aptitudes, skills, and experience. Yet the primary way the government attempts to tap their collective brainpower on a regular basis is by asking them all the same simple questions at election time. Candidate A or Candidate B? Yes or No on Proposition X?
Over the last two decades, great efforts have been made to use technology to democratize campaign fundraising and help people become more engaged and informed voters. In contrast, little attention has been paid to a far more ambitious and potentially transformative quest: using technology to help people become more engaged and productive citizens, in ways that truly harness the full range of their skills and expertise.
One effort to do exactly that is Challenge.gov. Created in 2010 and run by the General Services Administration (GSA), it offers a common platform where federal agencies list prize competitions open to the public. In April, for example, the Department of Defense (DOD) announced that it's seeking a turbine engine with two or more times the fuel efficiency of current small turbines and three or more times the power-to-weight ratio of a standard aviation piston engine. And it will pay $2 million to the first person or team who successfully fulfills these requirements.
In its first five years of existence, Challenge.gov hasn't generated a huge amount of attention or activity. To date, approximately 70 federal agencies have posted around 400 competitions, eliciting responses from 50,000 or so "solvers" and awarding $90 million in prizes.
But even with this small data set, the versatility and potential usefulness of Challenge.gov is clear. Some agencies are simply using it as a kind of outreach or publicity tool. In one recent contest, for example, the Department of Consumer Safety challenged schoolkids to create posters that convey the dangers of carbon monoxide poisoning.
Others, however, are essentially using the website as a new form of procurement. In some instances, the goods or services they seek are fairly general. The National Institute on Drug Abuse is now offering $100,000 for "bold new ideas" on how to manage and improve the clinical quality of addiction treatment. Others are far more specific. NASA recently solicited designs for a 3D printable handrail clamp assembly for the International Space Station.
Traditionally, federal agencies engaged in procurement write very detailed specifications about what they're looking for, circulate requests for proposals in established channels like FBO.gov, then evaluate the proposals that come in.
This approach essentially encourages the agencies themselves to frame and shape potential solutions, and to determine which of these proposed solutions will provide the best result. Shifting the frameworks of procurement to a prize format, however, alters the process. Instead of trying to define and thus solve problems in advance, agencies define a more general outcome they're seeking to achieve without telling potential vendors exactly how to pursue it.
Jim Adams, NASA's acting chief technologist, addressed this dynamic in a May 2014 report on prizes published by the White House Office of Science and Technology Policy. "NASA is one of many organizations that follow a strict procedural approach to solutions," Adams exclaimed. "This is valuable, but has the potential to blind problem solvers to alternative solutions."
Like most crowdsourcing platforms, Challenge.gov shifts much of the risk and resource investment to potential creators. Federal agencies can describe what they're looking for in more general terms, then evaluate actual solutions, not just proposals. And ultimately, they only have to pay for outcomes, not development efforts. While various parties will attempt to deliver the more efficient turbine engine the DOD is hoping to find, the government will only pay the $2 million prize if someone does in fact build an engine that meets its specifications.
And it's not just that these kinds of platforms only reward positive outcomes. Typically, they also end up leveraging the incentive money they offer several times over. In the case of the NASA handrail clamp assembly challenge, the total prize money offered was just $2,000. But it attracted 474 entries. Had NASA been paying market rates to even just the top 10 percent of these entrants for the time they spent designing their submissions, its costs would have been far higher.
In general, crowdsourcing platforms inspire innovation by putting problems in front of more eyes. And Challenge.gov is already working in this fashion. Aaron Foss, who won a 2012 Federal Trade Commission challenge that sought new methods of helping consumers block telemarketing robocalls, told Forbes that he "never would have worked on the robocall problem if not for the challenge." Similarly, a NASA survey of approximately 3,000 challenge participants found that 81 percent had never previously responded to government requests for proposals.
In addition to more eyes, crowdsourcing platforms like Challenge.gov tend to attract fresh eyes—i.e., people with skills and knowledge that may be useful in a given domain, but who aren't specialists in that area. Because they aren't familiar with which types of approaches should yield potential solutions, they often pursue unconventional methods that turn out to also be effective.
In essence, then, crowdsourcing is a supremely democratic mechanism: It starts with the assumption that everyone has something potentially valuable to contribute. Voting, jury service, and other forms of civic engagement, like commenting on proposed government rules, make this assumption as well, but they're all far more proscriptive in the ways they permit individuals to perform those actions. When you vote, you must follow the same basic set of processes as everyone else.
What makes participating in a Challenge.gov contest so unique in terms of civic engagement is how much latitude it gives to participants. Like voting and jury service, it's an explicitly collective activity that can help instill a sense of being a part of something greater than oneself. But it's also explicitly geared to individuals. It asks people to identify areas of interest to them where they think their unique skills and attributes may help them provide a solution that will ultimately be of value to all. In this regard, it both plays to our growing preference for highly autonomous and narrowly tailored experiences and promotes comity as well. We're all in this together, America, trying to invent the wearable alcohol biosensor!
And of course there's the lure of cash prizes. On the one hand, market-based civic engagement may not seem like civic engagement at all. But the federal government already spends hundreds of billions of dollars a year on procurement. Shifting some of that activity to a platform that is more open, transparent, and competitive than traditional channels will likely cost less, not more. And a greater number of participants will have a shot at the money that does get spent.
In the end, the democratic promise of Challenge.gov is so great it almost seems like a mistake to leave it in the hands of the government. What it needs is a Jeff Bezos or a Pierre Omidyar at the helm, someone with the expertise and the capital to take it from 100 challenges a year to 100 challenges a day. Or perhaps we should make it a self-directed contest: The first person who figures out how to triple Challenge.gov's traffic wins a million dollars.
Related: Should the Government Regulate Paid Dinner Parties?
This article originally appeared in print under the headline "Better Government Through Crowdsourcing."
