This year, 48 million Americans will contract a foodborne illness, many from restaurants. Just last week, dozens were sickened by an E. coli outbreak at Chipotle restaurants in Washington and Oregon. However, municipal governments are using open data to fight back against this problem. On October 27, 2015, the Center for Data Innovation convened a group of leading civic technologists and policy experts to discuss how publishing restaurant inspection data can help consumers make more informed choices, improve public health, and to improve food safety. Though many municipalities have a long way to come, the panel concluded that the potential of open inspection data could transform how society approaches food safety.
Health inspections are consumers’ main line of defense against foodborne illness when eating at a restaurant, as inspectors can punish restaurants for any health code violations that could result in unsafe food. The data that these health inspectors collect can provide valuable insights into food safety in restaurants, but many municipal governments do not share or use this data effectively, which can put consumers at risk. Sarah Schacht, public health data advisor at Socrata, noted that while most jurisdictions do make restaurant inspection data publicly available, it can often be very difficult to access. In Cambridge, Massachusetts, for example, consumers can only obtain restaurant inspection data if they physically go to a government office and examine a paper file. More commonly, this data is available online but in a format difficult for the average consumer to understand. Schacht estimates that between 50 to 60 percent of jurisdictions do not post restaurant inspection scores, in many cases because there are laws prohibiting it. For example, inspectors in Maryland and New York are not legally allowed to post inspection scores, preventing consumers, unable or unwilling to analyze inspection data themselves, from making more informed choices about where to eat based on information about restaurant safety.
To overcome these challenges, Yelp developed the Local Inspector Value-Entry Specification (LIVES) data standard in 2012 in partnership with the cities of San Francisco and New York. The goal of the LIVES standard, said Luther Lowe, vice president of public policy and government affairs at Yelp, is to facilitate the use of open inspection data by third parties, such as businesses like Yelp and civic technologists. While some jurisdictions may be unable to post inspection scores, by publishing the underlying inspection data in a usable format, third parties can easily derive these insights themselves by processing this data. Now, for restaurants in municipalities that have adopted the LIVES standard, Yelp displays health scores on its site to allow consumers to evaluate a restaurant’s safety alongside customer reviews. By making inspection data more transparent, the LIVES standard can help create a feedback loop for restaurant safety—as restaurants realize that consumers have increased access to and make decisions with this data, they are more likely to self police and reduce dangerous practices.
In addition to helping consumers make better choices, all panelists agreed that publishing inspection data in formats that users can easily interpret have substantial public health implications. While there is not yet peer-reviewed research on the subject, Schacht expects wide adoption of LIVES to substantially impact the prevalence of foodborne diseases. Schacht cited the city of Toronto’s practice of posting health inspection placards at restaurants indicating their compliance with food safety standards. After using this system for just two years, Toronto experienced a 30 percent decline in foodborne illness rates—an improvement that Schacht said would be equivalent to eliminating 118,000 foodborne illnesses and preventing two to three deaths every year in the United States. As Socrata launches LIVES data in 500 jurisdictions across the United States and the United Kingdom, Schacht expects that the increased accessibility of inspection data will have an analogous effect.
Just as the LIVES standard makes inspection data easier to use by third parties, it can also make this data more useful for governments. Non-standardized inspection data, data that exists only on paper, or data that is difficult to interpret restricts the ability of governments to access, compare, and analyze this data to identify new opportunities and address shortcomings in their inspection practices. But with easy to interpret open inspection data, the city of Chicago Chicago was able to develop a method to forecast where food safety violations are likely to occur by incorporating inspection data with weather data and data from 311 complaints, the city’s public non-emergency service line. This method, said Carey Anne Nadeau, chief executive officer of Open Data Nation, can identify 25 percent more violations, seven days sooner, on average, than the city’s traditional approach. Both Jack Madans, product growth manager at Code for America, and Schacht noted that governments are already using LIVES data to improve their approach to inspections, such as by prioritizing where to send inspectors based on analysis of inspection data. Madans also explained how public officials respond very strongly to lists that compare cities, such as a hypothetical “top 10 most unhealthy places to eat in the country” news article. No city wants to be on that list, Madans said, but these kinds of comparisons are only possible if inspection data is easily accessible and standardized. With standards like LIVES, cities can better assess their relative performance and are more likely to take corrective action.
Deploying the LIVES standard and improving open inspection data is a crucial first step towards transforming how governments and businesses alike work to protect public health. Lowe announced that Yelp is on the verge of deploying an algorithm developed by civic hackers in Boston that can predict the risk of an outbreak of foodborne illness by analyzing historical inspection data. When a government adopts the LIVES standard, Lowe said, Yelp will provide them with a real-time feed of restaurants likely to have a high risk of health violations based on insights from this algorithm and consumer reviews. Additionally, Schacht said she expects that making large amounts of open inspection data more usable will empower academics and public health researchers to develop powerful new insights into food safety and better map outcomes.
Overall, the panel explored a wide variety of applications of open inspection data that can reduce the likelihood of foodborne illness and save lives. As governments continue to improve the utility of open inspection data, the panel made it clear that the public and private sectors alike have a key role to play in capitalizing on the opportunities that this data will create.