Open Data Makes Big Social Impact
The availability of Open Data has the potential to solve a wide range of problems, but questions remain about what role the government should play, a panel of experts said recently.
At a July 23 event, leaders from the public and private sector hailed the recent use of data to assist in everything from fixing potholes to sending kids to school. But how can Open Data's potential be maximized? And will government ultimately inhibit or enable data-driven innovation?
The Information Technology and Innovation Foundation’s Center for Data Innovation co-hosted the heavily attended forum with the Sunlight Foundation on the Social Impact of Open Data. The event featured remarks by a Commissioner of the Federal Trade Commission (FTC) followed by a panel of three speakers. The director of the Center, Daniel Castro, served as the moderator.
In her remarks, FTC Commissioner Maureen Ohlhausen stressed government’s role as a major producer of data, much of which could also be used by the private sector to create value. She focused on the policies the FTC should pursue to promote access to government data. First, it should promote a realistic view of big data, aware of both its promise and limitations. Big data sources are not perfect. Correlation does not prove causation and data collection may be partial and/or biased. The FTC should critically evaluate data’s potential and focus on actual, as opposed to speculative, harms. Second it can continue to provide guidance on how to protect privacy while using data. Lastly, it can embrace a culture of open data and innovative data analysis and promote transparency within government.
During her remarks, Commissioner Ohlhausen mentioned the well-known example of Street bump, a mobile app that identifies pot holes when users of the app drive over them. Some have implied that Street bump’s value is tainted by the fact that drivers in wealthier neighborhoods are more likely to use the app. At the end of her remarks, the moderator asked the Commissioner a question about how to approach regulating the app, which she successfully avoid answering. That is a shame, because I suspect that one’s approach to this app says a lot about whether government will be an enabler or inhibitor of big data. One approach is to view Street bump as an unmitigated good because it dramatically reduces the cost of doing something that the city had to do anyway; identifying potholes in rich neighborhoods. This approach would not hold the app responsible for either the quality of how the city performs its traditional job of spotting potholes in poor neighborhoods or the city’s decision of how to allocate repair services between rich and poor areas. Another approach might make Street bump responsible for any discrepancies in service, whether or not they preceded the use of the app or, worse, prohibit use of the app until its developers could prove that it would not lead to a discriminatory result. So far, government officials have glided over this question.
Emily Shaw of the Sunlight Foundation talked about how open data creates social impact. Open data is a lot like a classic public good because many people can use it without diminishing its value and, once it is made public, it is difficult to exclude people from using it. This justifies having the government devote resources to building and maintaining important data sets. But maximizing the value of this resource requires communication. Users need to communicate about successes in promoting goals such as government transparency, accountability, and efficiency. They also need to be vocal about shared goals and the data sets they want access to, as well as about problems in the data set.
Sandra Moscoso of the World Bank discussed a large number of cases in which open data is currently having a positive social impact. In general, these applications share the general effects of promoting government transparency and strengthening the autonomy of the individual. For example, several apps in different countries help families chose the right school to send their children to. Other apps help provide educational content to classrooms. In transportation, a number of programs help commuters use public transportation efficiently. Like the previous speaker, she emphasized that data’s value depends on its ability to solve real-world problems.
The final speaker was Brian Rayburn of Symcat, a new company that uses data from hundreds of thousands of patient records to match a user’s symptoms with possible diagnoses. While the company does not provide a firm diagnosis, it does give users information on the conditions that might be causing their symptoms and suggests professionals who might be able to provide the right therapies. Like other uses of big data Symcat allows users to ask questions that they could not ask before. More than most companies, however, Symcat has to worry about the accuracy of its models, because the consequences of both a correct and an incorrect diagnosis can be great. Mr. Rayburn also stressed the frequent need to merge data streams. Unfortunately, many app companies do not have an incentive to share the raw data that they get from customers. This sharply limits the ability to use the data for other purposes.