Open data by default: Challenging but worth it
Open data is about more than releasing data for outside use. For Tempe, it’s about creating a data-driven culture by changing the ways that we think about and use data inside of government. We believe that working toward a culture that views data as a strategic asset will allow us to accomplish our strategic goals while meeting the needs of the community.
In February 2017, Tempe launched its open data program. Working with the Sunlight Foundation through the What Works Cities initiative, we started developing an open data policy that defines the city’s objectives and the key staff positions needed to run the program.
We are now asking for feedback on our draft policy from residents and subject matter experts. The draft open data policy is available online in a format that allows you to leave comments directly in the document text. The draft open data policy will be open for comment until Wednesday May 31, 2017.
Our policy’s central idea is that data should be released proactively, without waiting for someone to request it, without fees, and without restrictions on its use. This represents an exciting culture change for Tempe. Our policy and the open data program are major elements of Tempe’s commitment to using data to make better-informed, transparent decisions.
While the policy is an important step in launching a successful open data program, there are many challenges in creating a culture that values data as a public asset and presumes that data should be open and accessible by default.
- Documenting impact. The benefits of, and demand for, open data in the community are often uncertain or unclear to city staff or departments. It’s understandably hard to take on change when it’s not clear that your efforts will impact the community in a positive way. While the policy emphasizes the potential benefits and value that opening data to other departments, the community or businesses may bring, telling people something works isn’t a powerful motivator for change. We have worked to overcome this uncertainty by seeking out and using tangible examples of open data projects that have benefited residents, and shared them with our staff. As a city starting out with an open data program, we did not have a large reserve of examples for our City’s open data program. But as part of the What Works Cities initiative, we have been able to leverage the stories and challenges shared by other cities to illustrate the benefits of open data in different communities.
- Protecting privacy. Being good stewards of the city’s data means ensuring that there are protections in place to avoid releasing protected or sensitive data or putting the public at risk. A process for reviewing each dataset is a key element of meeting the security and privacy objectives of the open data policy. Potential risks to privacy and security is a complicated issue that begins — but doesn’t end — with protecting personally identifiable information. This includes not only ensuring that no personal information is released with a single dataset, but also that the combination of datasets does not enable the identification of individuals. Additionally, providing the specific location of certain types of incidents, even without specific information about an individual, may still have an impact on the area. Moreover, providing raw data does not ensure that meaningful information or conclusions will come from that data when it is used.
- Managing Risk. We are approaching these issues during the data inventory and publication process in several ways. We complete a Sensitive and Secure data worksheet for each dataset, including prompts when additional legal reviews are needed. We will maintain information on published datasets in a single catalog location. Departments can review related datasets and will begin working to ensure that they cannot be combined with new data to uncover personal information. For sensitive location information, locations may be randomized at the block level. For example, the city might generalize sexual assault locations to the 100-block level. If mapped, then the location could be randomized within that 100-block.
- Catalyzing Collaboration. City staff often have concerns about the potential for negative feedback or misinterpretation that may come from releasing data or performance analytics. Releasing data to the public or other city departments requires trust that departments will have support addressing challenges or questions that arise from releasing that data.Our draft open data policy makes the release of data a collaborative effort that engages people beyond the group responsible for maintaining the data, including city leadership, directors, supervisors and city staff. We hope to continuously address this challenge by learning from positive examples of other What Works Cities overcoming setbacks through collaboration between city staff.During the process of creating the open data policy and working on the portal, we experienced internal challenges that we faced (and continue to face) but addressed through collaboration. Working across departments has helped us gather information more efficiently, develop complete materials for the data coordinators, and even recreate the design of the data pages. This illustrates that it is possible to move past challenges through trust and collaboration.
These represent only some of the challenges that arise when you try to spark a change in data culture. They are not unique to Tempe. Initial resistance to change is understandable when people are asked to shift the ways that they think about and use data.
When we started this work, our goal was to create a policy that addresses the needs of our community, city officials and city staff. Meeting our challenges, incorporating the lessons into our policy, and collecting feedback from residents are all key elements of achieving that goal.
Stephanie Deitrick serves as the manager of Tempe’s Open Data Program and the Enterprise GIS, Data and Analytics group. She has been with Tempe since 2014, when she began working to establish Tempe’s Enterprise GIS Services Team. You can email her at stephanie_deitrick@tempe.gov