• publish data about government deliberations, management, and results in standardized, open formats; and
• fully implement the federal DATA Act (Public Law No. 113-101), expanding it to additional data sets.
Americans widely approve of government transparency, but it is difficult to deliver. President Barack Obama called for transparent government, and his administration began with promises of great strides in transparency. After eight years, however, the federal government achieved only marginal improvements in some areas, with backsliding in others. Passage of the Digital Accountability and Transparency Act (DATA Act) was an important but incomplete step toward the promise of transparency in government spending. What remains to be done at the administration’s end is to fully implement the act.
Government transparency is difficult to deliver in part because the practices that produce transparency are not well understood and in part because transparency is not in the interest of some parts of government. Transparency shifts power from government agencies to their overseers in legislatures and the public. Improvements in freedom of information laws are welcome. But modern transparent government will result from the extension of consistent data publication practices to government deliberations, management, and results. Good data publication practices will make those types of information available for processing and wide redistribution on websites and through information services and apps, so people will be fully informed about the activities and expenditures of government.
What Is Transparency?
Everyone agrees that governments at all levels should be transparent, but how to achieve that goal has not been clear. Many people over many years have equated government transparency with the ability to access information about government via “freedom of information” laws. The federal Freedom of Information Act (FOIA) and state counterparts allow the public to demand access to data held by governments. These laws generally establish processes by which citizens may request government-held information, paying minimal costs for accessing it. There are exceptions to disclosure for the sake of protecting individuals’ privacy and national security, and many executive branch officers and legislators are exempt.
A significant hindrance to transparency is the low priority that government agencies often give to FOIA requests. The government-wide backlog of FOIA requests at the federal level went from 77,000 in 2009 to nearly 160,000 in 2014, according to a Justice Department report. The numbers have improved somewhat since then.
In 2016, Congress passed the FOIA Improvement Act (Public Law 114-185), seeking to improve FOIA processes at the federal level in various ways. Among them were requirements to disclose records and documents in electronic formats, to publish widely any frequently requested records, to limit fees, to create a presumption of openness, and to permit FOIA requests to be made of any agency from a single website.
Bringing government up to modern standards requires improving on FOIA by making information about government activities available as a matter of course, not based on request. Governments should publish data about their activities as a matter of routine and in digital formats that make it easy for the public, and governments themselves, to access and use.
A 2011 Cato Institute study identified four key data practices that support government transparency. They are authoritative sourcing, availability, machine discoverability, and machine readability. The first, “authoritative sourcing,” means producing data as near to its origination as possible — and promptly — so that the public uniformly comes to rely on the best sources of data. The second, “availability,” is another set of practices that ensure consistency and confidence in data the government publishes.
The third transparent data practice, “machine discoverability,” occurs when information is arranged so that a computer can discover the data and follow linkages within the data. Machine discoverability is produced when data are presented in a manner consistent with a host of customs about how data are identified and referenced: the naming of documents and files, the protocols for communicating data, and the organization of data within files.
The fourth practice, “machine readability,” is the heart of transparency: it allows the many meanings of data to be discovered. Machine-readable data are logically structured so that computers can automatically generate the myriad stories that the data have to tell and put the data to the hundreds of uses the public would make of it in government oversight.
The Subject Matter of Government Transparency Data
Experience with the Obama administration’s transparency efforts shows the importance of carefully defining what makes government transparent. The administration can easily say the publication of any data is helpfully transparent. However, only access to the types of information already mentioned — data about government deliberations, management, and results — makes for true government transparency.
In 2009, the Obama White House instructed agencies to produce data for data.gov. The instruction — produce three “high-value data sets” per agency — was broad, and its definition of “high-value” was equally broad. That broad definition permitted agencies to choose data feeds with the greatest likelihood of increasing their discretionary budgets or the least likelihood of shrinking them. The Agriculture Department, for example, published its calorie counts, not its check register.
Digital scholars David Robinson and Harlan Yu identified this shift in policy focus in their paper “The New Ambiguity of ‘Open Government’ ”:
Recent public policies have stretched the label “open government” to reach any public sector use of [open] technologies. Thus, “open government data” might refer to data that makes the government as a whole more open (that is, more transparent), but might equally well refer to politically neutral public sector disclosures that are easy to reuse, but that may have nothing to do with public accountability.
The heart of the government transparency effort is getting information about the functioning of government: government deliberations, management, and results. Making “deliberations” transparent means publishing the bills in Congress and state legislatures in structured formats with semantically rich markup or metadata. The rulemaking documents published by agencies should be given similar treatment. Along with the documents, the legislative and regulatory processes followed in rulemaking bodies should be equally transparent: hearings and other meetings, amendments, votes, and other decisions should all be made available in well-structured digital formats.
“Management” essentially refers to the flow of funds. Thus, the published data should cover budgeting, appropriations, allocation of spending authority, obligations, and actual outlays of money — all in formats that tell the story of who was responsible for what spending. “Results” means reporting on what the government activity, including spending, actually produced, hopefully for the betterment of the jurisdiction. Reporting on results is a difficult challenge, but it has been required at the federal level, at least, since Congress passed the Government Performance and Results Act in 1993.
Sound publication of data about deliberations, management, and results will make for a more open, more transparent government. It is important that the subject matter of government data publication remain focused on what makes the government itself transparent.
The DATA Act
One of the most important recent developments in transparency at the federal level was the 2014 passage of the Digital Accountability and Transparency Act of 2014 (DATA Act) (P.L. 113-101). The DATA Act required the federal government to increase the amount of information available on the USASpending.gov website. More important, it required the Treasury Department and Office of Management and Budget in the White House to establish standard data elements and formats to be used government-wide for all federal spending data.
Public access to such data could revolutionize both public oversight and internal management of the federal government. With spending data published in standard formats, a variety of websites, information services, and apps will be able to reprocess and republish the data, making the data available as information to various segments of the public. Data about the activities of agencies may be reported as straightforward statistics, much like the sports statistics found in newspapers, rather than described in “news” articles that contain few facts but many opinions of the authors and their editors. People may be able to draw relationships between bills introduced in Congress, the votes on them, and actual outcomes, whether those outcomes are beneficial infrastructure developments, such as a repaved highway, or damaging and counterproductive excess in an overseas military operation. And they will be able to tie those outcomes to specific legislators’ specific actions.
There is no guarantee that the DATA Act will be implemented well enough to produce such results. The law does not specify, for example, that each organizational unit of the federal government should have a unique identifier. Strikingly, the federal government today lacks a machine-readable organization chart. There is no single, consistent account of what agencies, bureaus, programs, and projects exist in the federal government. Having that information is as essential for government management and public oversight as having street addresses is for delivering mail. The lack of a coherent map of the federal government is part of the reason why it is so large (due to duplication) and so wasteful (due to opacity and failed oversight).
Full — and fully informative — implementation of the DATA Act will help cure the significant ailments of the federal government, allowing Congress to more effectively oversee the executive branch, the executive branch to more effectively manage itself, and the people to more effectively oversee both. But even the full implementation of the DATA Act and transparent data publication practices in general will not automatically deliver government transparency. Transparency also relies on the development of a community of websites, information services, and apps that process and broadcast data for public consumption. That type of robust transparency depends on the publication of data about government deliberations, management, and results in standardized, open formats.