Washington, D.C., is the first local government we identified as having an open data policy. The Office of the City Administrator issued a memorandum in 2006 listing certain datasets that were to be posted online in a timely manner to help the public better understand government activities and “become knowledgeable participants in improving the quality of life and promoting economic development in the city.”
DC’s open data policy is considered to be among the first in a long line of more than 40 such policies now in place in states and municipalities around the country. These policies have grown significantly beyond the vision outlined by DC’s 2006 memorandum, expanding to include strong oversight, clear timelines, and, perhaps most importantly, language about “setting the default to open” rather than just listing certain datasets to be posted online.
The evolution of open data policies since 2006 provides a chance for stakeholders to learn from and build on what’s been accomplished so far. This summer, a new executive directive from Mayor Vincent Gray’s office could have taken advantage of that opportunity for growth. It fell far short, however. The scope, level of detail, and enforceability of the policy seem to reveal a lack of seriousness about making a significant improvement on DC’s 2006 memorandum.
Here’s how DC missed an opportunity for improvement and what it can do next.
How DC’s old and new policies compare
DC’s new open data policy did make a few improvements worth noting. Compared to the 2006 policy, the new policy:
- Mentions proactive release;
- References and builds on an existing access to information policy, the Freedom of Information Act;
- Specifies some criteria for prioritization of data release;
- Stipulates that the provisions apply to contractors;
- Mandates open formats;
- Encourages formats “fully accessible to the broadest range of users, for varying purposes”;
- Calls for the inclusion of public perspectives;
- Sets timelines for identifying the first round of high-value datasets;
- Includes language about exploring potential partnerships “among departments, other governmental agencies, the public, and non-profit and private entities in fulfilling its obligations”;
- and mandates an annual Open Government Report with updates on the progress of releasing open data and future plans for data release.
The new policy also takes some steps back for open data, however. The new policy removes the language about including “[h]istorical records that represent longer time periods beyond the current dataset” and removes language about data quality and accuracy.
- Narrowly setting the default to open for public information by focusing only on a subset of datasets,
- Not including a balance-test for safeguarding sensitive information,
- Not creating a comprehensive list of information holdings, instead only creating a list of data being shared online,
- Not explicitly asking for or considering public feedback in the process for determining high-value datasets,
- Not mandating the use of unique identifiers,
- Not requiring code sharing or using open source methods,
- Not optimizing methods of data collection by calling for electronic filing,
- Not ensuring permanent, lasting access to data,
- Not clearly mandating ongoing data publication, instead leaving it to each agency in collaboration with the Office of the Chief Technology Officer to determine appropriate updates,
- Not creating binding guidance or regulations for implementation of the policy,
- and not creating processes to ensure data quality.
That’s not the end of the list of what DC could have achieved with its new policy but didn’t do at all or didn’t do very strongly.
How did this happen?
What should have been done differently
There are two primary things that should have been done differently and could have helped significantly strengthen DC’s new open data policy.
1. The community’s input should have helped shape more of the policy
In the nearly 10 years worth of open data policies that we’ve observed, including more than 40 original state and municipal policies and various updates to several of those, the strongest iterations have been informed by community engagement and feedback. DC seemed poised to take community input into account when it put out a call for comments in the fall of 2013 about its transparency and open government plan. Comments came in from the DC Open Government Coalition, Socrata, Code for DC and Sunlight, among several others.
Specific suggestions in those comments varied, but a common refrain is evident: people asked for better access to information, including low barriers to access, machine-readable and searchable formats, more public engagement around information, and clear explanations any time information is withheld from the public.
Some of these suggestions seem to have been taken into consideration for making parts of the new policy stronger than the 2006 version. The new policy does, for instance, specify open formats and call generically for more public engagement. It has some big problems with the key open data principle of reuse, however, as Joshua Tauberer explored in detail, not to mention the best practices we listed above that were missed.
The policy could have made up for these shortfalls by requiring the creation of binding guidance or regulations for implementation that would more fully embody the public comments and recommendations. How successfully any of the new open data policy will be implemented, in the shadow of having no clear mandate for implementation guidance beyond technical standards, remains to be seen. The real impact of the policy will come from how it is implemented, but DC’s commitment to embracing an open data culture seems shaky at best after this year’s executive directive was first issued as a non-searchable PDF document, “itself embodying the way not to release text online in the 21st century,” as Alex Howard put it.
If DC’s open data portal to date is any indicator, there’s still much room for improvement with implementation of the government’s open data ambitions. It’s a credit to the community that, despite the barriers so far, numerous tools and stories have been created and shared openly using DC’s data. It’s all the better reason to draw on the community knowledge and input for execution of the policy.
2. The policy should have been passed through the legislative, not executive, process
The other big problem with DC’s new open data policy is nearly equally as concerning as the lack of embracing community suggestions.
The new policy is an executive directive. In some local governments, anything done through the executive branch does not hold the same force of law as something passed through a legislative body. The enforceability of an executive directive may be limited to the administration that issued it.
DC is no exception. The executive directive issued this year is enforceable only so long as those in the executive branch should choose that it be so, and it comes at a time when the administration is about to change. Voters will decide the next mayor of DC in the general election on Nov. 4. It’s certain that Mayor Gray will be out since he lost the primary election. Will the new administration choose to enforce Mayor Gray’s executive directive and carry out its implementation? The implementation of a policy aimed at improving openness doesn’t seem like something that should be left subject to deterioration, but that’s where this policy is at for now.
DC missed an opportunity this time, but there are still ways to bring the local government with one of the first open data policies into the leadership role it should have seized.
The new administration and Council should work together with the community to craft a strong open data policy that builds on community feedback, will continue to have the force of law and will have a strong, detailed guidance component for implementation. The suggestions made so far remain public for anyone to read and consider. The community is clearly ready to engage around this topic and make DC’s open data policy a truly iterative process that seeks constant refinement and improvement.
In the meantime, the good news is that there are ways to engage DC government on this issue. The 2014 open data executive directive is on Madison, a platform that allows for public feedback on specific aspects of the text and a generic up or down vote for the entire document. It’s one way to send a message about what should have been done better and what should be done next.
DC’s new open data policy rightly touts that it was one of the first local governments with such a policy. DC should now step up to take advantage of the lessons learned since then and the best practices available to set a high threshold for this crucial open government initiative.