John M. Kamensky in Governing: “In a recent report for the IBM Center for the Business of Government, Brabham says that an important distinction between crowdsourcing and other forms of online participation is that crowdsourcing “entails a mix of top-down, traditional, hierarchical process and a bottom-up, open process involving an online community.”
Crowdsourcing in the public sector can be done within government, among employees as a way to surface ideas — such as the New York City government’s “Simplicity” initiative — or it can be done by nonprofit groups in ways that influence government operations. For example, a transportation advocacy group in New York City has created a site where citizens can report “near miss” accidents, which are then mapped to determine patterns. The idea is that, while the city government already maps accidents that have happened, hazardous traffic zones can be detected and resolved faster by mapping near-misses without waiting for a large number of actual accidents.
Brabham offers a strategic view of crowdsourcing and when it is useful to address public problems. His report also identifies four specific approaches, describing which is most useful for a given category of problem:
• Knowledge discovery and management. This approach is best for information-gathering and cataloguing problems through an online community, such as the reporting of earth tremors or potholes to a central source. This approach could also be used to report conditions of parks or hiking trails or for cataloging public art projects as have been done in several cities across the country.
• Distributed human-intelligence tasking: This approach is most useful when human intelligence is more effective than computer analysis. It involves distributing “micro-tasks” that require human intelligence to solve, such as transcribing handwritten historical documents into electronic files. For example, when the handwritten 1940 census records were publicly released in 2012, the National Archives catalyzed the electronic tagging of more than 130 million records so they could be searchable online. More than 150,000 people volunteered.
• Broadcast search: This approach is most useful when an agency is attempting to find creative solutions to problems. It involves broadcasting a problem-solving challenge widely on the Internet and offering an award for the best solution. NASA, for example, offered a prize for an algorithm to predict solar flares. The federal government sponsors a contest and awards Web platform, Challenge.gov, that various federal agencies can use to post their challenges. To date, hundreds of diverse challenges have been posted, with thousands of people proposing solutions.
• Peer-vetted creative production: This approach is most useful when an agency is looking for innovative ideas that must meet a test of taste or market support. It involves an online community that both proposes possible solutions and is empowered to collectively choose among them. For example, the Utah Transit Authority sponsored the Next Stop Design project, allowing citizens to design and vote on an ideal bus-stop shelter. Nearly 3,200 people participated, submitting 260 high-quality architectural renderings, and there were more than 10,000 votes leading to a final selection….”