Mid-America Regional Council Pilot of the Data Business Plan for State and Local Departments of Transportation: Data Business Plan

Appendix I. Best Practices

This appendix highlights two organizations that have been successful in implementing data initiatives, namely the City of Chicago and the Delaware Valley Regional Planning Council (DVRPC).

City of Chicago

The City of Chicago has been nationally recognized for its efforts in making data accessible to the public. The following are some lessons learned and recommendations from Brett Goldstein, the City's first Chief Data Officer:⁶

• The first step should be to assess existing baseline and decide where to take vision and direction for the organization.
• Philanthropic support was an important component for Chicago in this initiative. The MacArthur foundation sponsored a competition to encourage businesses and software engineers to use Chicago's open data to create helpful apps for residents. This competition also helped create a framework to engage with the community.
• The city created a new senior-level post within the Mayor's office: The Chief Data Officer (CDO), tasked to make government data available to the public and use data analysis as a tool to inform policy and improve services. This ensured that data initiatives had a clear mandate.
• They discovered that "there is enormous benefit to a high-profile release of a high-interest dataset early on." City officials know that crime incident data was hard to obtain in disaggregate, raw form. There was also a strong interest from the public to obtain prompt and transparent crime data. The City prioritized this data to be the first one launched, and they created publicity and buzz around it.
• Rather than getting into the business of developing apps, the City of Chicago provided a standards-based data portal that enabled them to be a platform that supports innovation from researchers, civic developers, and for-profit use.
• Providing data in machine-readable formats is of utmost importance. This may require the "tedious, but critical, work" of an intern to convert an unusable file into one that can serve as a data source.

  For the data to be successful, they had to:

  • Reduce the data to block size and scatter spatial coordinates in order to protect privacy.
  • Capture updates and replicate them into the data set as the source system records were updated.
  • Have a system in place to handle uploads, updates, and queries of large datasets.
• Proprietary platforms are often much easier to use and are ready to go. However, they are an investment that requires ongoing funds to be sustained. An open-source platform may demand significantly more technical skills to set it up, but may be potentially much cheaper.
• Agencies need to find ways to extract data, understand it, and load it into the platform. Think about network, storage, and systems.
• Automation is a key component to work with large datasets. "An open data program that relies on a human to keep it updated is fundamentally flawed." The Chicago portal updates itself every day.
• Sometimes public agencies will get bad press coverage due to errors or oversights in releasing data. To help prevent that from happening, it is important to develop a strong relationship with stakeholders, including explaining to the press the importance and significance of the initiative.
• Top-Down and Bottom-Up: As this data initiative gained traction and maturity, to take to the next level, the mayor issued an Open Data Executive Order mandating that each department would designate an Open Data Coordinator and determine a system of annual accountability regarding the release of open data. In the case of Chicago, Goldstein claims it made more sense to let this initiative evolve and gain momentum before an executive or legislative action. Otherwise there is a risk that it might become too prescriptive.
• There are two key items that are crucial for the success of a data initiative: 1) clear and vocal support of the executive sponsor, and 2) financial support.

Delaware Valley Regional Planning Council

DVRPC was identified by MARC and the project team as leader from whom to learn about data management practices. Kimberly Korejko, Data Coordination Manager at DVRPC, shared through an interview the following lessons learned:

• It is helpful to have a clear sense of organization to coordinate data initiatives.
• In the case of DVRPC, they have set a series of coordinating levels, as shown in Figure 8.
• Data Resources and Coordination Team: This core group is comprised of staff whose daily tasks are strongly oriented toward data management. They are vital in helping complete the tasks identified through data coordination efforts.
• Advisory Teams are in charge of identifying and prioritizing data initiatives each year, as well as assisting in creating standards and policies. It is comprised of Planning, Technical, and Management staff.
• Innovation Teams: These teams are formed on an as-needed basis for specific needs or initiatives.
• Member Governments and Planning Partners: These are external stakeholders that provide data to DVRPC and may participate in data sharing initiatives.
• Other End Users may be organizations or individuals interested in information or data.
• Start with what you can, and build from there.
• DVRPC is rarely a producer of data. Instead, it uses other organization's data. As a two-State Metropolitan Planning Organization (MPO), the data it receives is often not compatible with one another. Although DVRPC has not been able to set standards, this has not prevented it from leading many data initiatives. For instance, DVRPC has an online, searchable GIS Data Catalog with data location, abstract, purpose, use limits and licensing, and data elements. The MPO is now working to create metadata for non‑geographic information system (GIS datasets and hopes to have a unified, searchable interface to make data available online.
• Make management aware of the importance of data initiatives.
• It is crucial to be an advocate for data initiatives and data governance. Although one may need to repeat oneself doing this, having buy‑in from upper management pays off well. In the case of DVRPC, they were able to formally establish that members of the Innovation Team should dedicate 5 percent of their time in data governance.
• Working with Information Technology (IT) Department is key.
• It is critical that IT staff understand the why behind data initiatives. To roll out the Online GIS Data Catalog, DVRPC arranged for Esri to meet with IT and go through all the technical "nuts and bolts" to make the initiative successful.