Published on August 25th, 2015 | by Joshua New0
Not Your Grandparents’ Community Service: Why Civic Hacking Should be Open Source
Forget coaching Little League or registering voters—for many people, hacking is the new community service. Government-sponsored hackathons around the country are challenging civic-minded data scientists, coders, and engineers to build apps and tools that use government datasets to improve everything from how Asheville, North Carolina informs its residents about its budget to how the National Aeronautic and Space Administration (NASA) monitors asteroid threats. But software development can pose tricky legal questions that planting trees does not. What if a civic hacker’s code infringes upon a company’s intellectual property? Will hackers get to use their code later? Who should be liable if the code produced during a hackathon unintentionally causes damage to a government server or user’s mobile device? To sidestep these issues, government agencies sponsoring civic hackathons should require all the code produced at their events to have open source software licenses. Doing so would limit the liability of participants and agencies, and encourage people with data skills to contribute their valuable talents to improving their communities.
Detroit Mayor Mike Duggan is one of many converts to the potential of open data, launching Detroit’s open data initiative just three months after the city emerged from bankruptcy in late 2014. And in June, the city hosted its “#hack4detroit” hackathon to challenge civic technologists to build apps using Detroit’s newly open data to benefit the city. Tucked into the participation agreement were well-intentioned provisions designed to protect the city from potential damages, but that turned off at least one volunteer from participating. The reason? To participate, civic hackers would have to agree to turn over all intellectual property rights for anything they produced during the hackathon, while still assuming liability for any damages that might result from their contributed code. Not only would this agreement not allow coders to later use what they have developed for other projects, but it would also expose them to potentially enormous risks should their code end up unintentionally causing harm. Such clauses disincentivize people from donating their time and talents to put open data to work for government.
If a major tech company developed a new app, it would take the time to ensure that it did not violate any patents or copyrights and that it did not contain damaging code before releasing it. But a city or agency may not have the tech savvy to sufficiently review hackathon submissions or the financial capacity to assume liability for these risks, and instead administrators may be tempted to place the burden of liability on volunteer hackers. But this approach is self-defeating. After all, the very reason governments open their data and sponsor hackathons is to improve government function, and any barriers to participation necessarily limits how much governments can gain. However, if government administrators specify that all code produced in hackathons is to be licensed as open source, they could simultaneously encourage civic hacking while shielding all parties from substantial amounts of such risks. Some civic hackathons already recognize the benefits of open source licensing. For example, the city of Houston’s hackathon encourages participants to make their projects open source and NASA’s Space Apps Challenge requires all submitted materials are open source or available without restrictions.
Using an open source software license would nullify concerns about intellectual property ownership rights for hackathon-produced software because agencies would be free to repurpose volunteered code for other applications, and developers could continue to use their code after the completion of a hackathon for any reason they choose, commercial or otherwise. Additionally, by allowing public scrutiny of this software, governments could supplement their capacity to police against potentially malicious or IP-infringing code. The burden of liability for using open source code rightly falls on whomever decides to use it.
Requiring the results of hackathon to be licensed as open source would also encourage its reuse. Tools and apps produced by a hackathon using San Francisco’s open data could be just as beneficial for Chicago. Without any stake in protecting such software, sharing and reusing this technology would only mean net gains for governments. There are already a handful of promising examples of this kind of sharing. For example, a city councilman in Whitewater, Wisconsin repurposed a budget data transparency tool originally built for Cook County, Illinois, and an app designed to help Boston residents locate and care for fire hydrants is now helping Honolulu residents test tsunami sirens. Local governments can achieve more by pooling their resources.
As the open data movement continues to grow, government administrators eager to kickstart its benefits with civic hackathons should recognize that this new kind of community service requires a new way of thinking about how to treat what people create.