Last month our Coalition and other transparency advocates cheered the release of President Obama’s Open Data Policy (and executive order). The Open Data Policy recognizes that government data is a public asset. Whenever possible, government data should be published, easy to find and use, machine-readable and interoperable. And the President is calling on agencies to pursue those principles – by designing new information collection processes and systems around data standards (sections 1 and 2), taking inventories of existing data resources (section 3), and creating strategic plans to build interoperability and openness into core processes (section 5).
But the Open Data Policy will not self-execute. Some federal data domains require specialized knowledge and are beyond the control of the chief information officers (CIOs) who will take the lead in implementing the Open Data Policy. In these domains, there is a wide gulf between technology policymakers and the agency staff who are actually collecting and disseminating the information.
Want a prime example? Federal spending. Financial, procurement, and assistance data systems are managed by federal CFOs, purchasing officers, and grant managers – not CIOs. Even after years of progress toward open data in other domains, like health and energy, federal spending is still a mess: incomplete, inaccurate and late; suffering by comparison to state and local efforts; unable to support government-wide analytics; and prone to embarrassing omissions. Nobody’s in charge of data standards for federal spending, so there are none.
If, as the President has said (page 5), open data is (1) public, (2) accessible, (3) fully described, (4) reusable, (5) complete, (6) timely, and (7) managed post-release, then federal spending isn’t open data. It flunks all seven of those characteristics.
That’s why the Digital Accountability and Transparency Act (DATA Act) is so important. The DATA Act, recently reintroduced in both the House and the Senate, would establish government-wide data standards for federal spending and require the publication of the executive branch’s grants, contracts, and internal expenditures in transaction-level detail. Without the DATA Act – which puts someone in charge (the Treasury) of turning inaccessible spending documents into open data – the spending domain will remain closed.
Spending is not the only area where open data advocates need legislative air support. Many other domains need legislative reform too – financial regulation, federal operations and management, Congressional actions, the judiciary, and more. It’s a challenge. Our Coalition is on it.
We’re developing a comprehensive legislative agenda for federal government data transparency, starting with the DATA Act for spending and the Financial Industry Transparency Act (FIT Act), expected to be introduced later this year, for financial regulation. Our goal? A thorough yet flexible roadmap of legislative actions to promote data transparency throughout the federal government, especially in those stubbornly change-resistant domains.
In early May, the Coalition had the opportunity to seek input from some of the nation’s leading transparency advocates on this ambitious project. At the Sunlight Foundation‘s annual Transparency Camp, Executive Director Hudson Hollister led a session titled “Creating a Legislative Agenda for Open Data.” Dozens of transparency advocates brainstormed about what the full landscape of open data mandates should look like.
That lively discussion was our first step. It produced some great ideas and valuable suggestions for a comprehensive legislative agenda. Here’s what the Transparency Campers told us.
Principles of Data Transparency
Participants agreed: government data belongs to citizens and general principles of openness should apply to all data transparency efforts. We need to continue to develop a culture of open government that includes open data. The presumption should be that the data is open and published, with effective privacy protections for nonpublic personal data. The public should be given enough information to understand the context and have access to people who can provide more information. For government data to be truly open, it needs to be in a format that is usable, which means it needs to be in a standard format with consistent identifiers.
Legislative Agenda Domains
Talking about federal government open data policy reform is a big topic, so we started with how to categorize the effort. The Coalition generally organizes legislative planning into five change-resistant domains: Spending, Operations & Management, Regulation, Legislation, and Judicial. A wide-ranging discussion of government data led us to the addition of three more categories: Quasi-Government, Elections, and Meta. With these categories in place for now, we dug deeper into both immediate legislative concerns and long term wish lists.
Meta: The purpose of the Meta category was to identify standards for government data standards. (You can see why we used the term “meta.”) An important aspect of standardizing government data is using standard names for the same concept or entity, such as Legal Entity Identifiers (LEI). Similarly, open data should be permanently open by providing a permanent URL (PURL) for all information published online.
Management and Operations: Clarity of federal operations is essential to establishing an open government. In a democracy, citizens must be able to understand what the government does before they can be said to authorize the government actions. The first step is to create a machine-readable government organization chart, which will allow a computer to show the breadth of the federal government and can connect agencies or projects with authorizing legislation, federal spending, regulations, and other important data to give the user a better understanding of specific parts of the federal government. Additional legislative efforts identified include the Access to Congressionally Mandated Reports Act, the Public Online Information Act and FOIA reform.
Spending: Understanding government spending is an important key to understanding what the government does. Currently, spending is hidden by disconnected agencies, incomplete data, and unreported federal spending. The DATA Act, by standardizing government spending data, will bring much needed clarity to how the government spends its money. Federal spending should be reported in a standardize data format through the entire lifecycle from appropriations through disbursement. Data should be standardized and published with standard identifiers to track each dollar for every authorization, budget submission, award, and outlay.
Regulation: The federal government’s regulations and regulatory process are opaque with complexity. Data standards such as standard identifiers and structured data formats can transform regulatory filings into a manageable dataset to be analyzed for insight into the industry or filers. Machine-readable regulations can simplify and minimize the regulatory burden for industries. Current legislative proposals to support include the FIT Act and the President’s electronic filing of IRS form 990 proposal. Participants agreed that regulatory information and data, including the CFR and Federal Register, should be machine-readable and easily navigable. Regulations should be electronically connected to the authorizing legislation, so the user can easily analyze what authority was actually given to the agency.
Quasi-Government: Quasi-governmental entities, entities with both private and governmental characteristics, are increasingly relied upon to implement federal government functions and it is important that public data from quasi-governmental entities is included in data transparency reforms. Participants suggested strengthening the Federal Advisory Committees Act (FACA) and ensuring open licensing. Like many other areas, participants were concerned with accessibility and easily navigable publication of quasi-governmental data.
Laws & Legislation: Like regulations, legislation should be machine-readable and legislation should be easily navigable through the lifecycle of the law. CRS reports should be accessible to everyone, not just congressional staff. Additionally all Congressional rules and committee information, including votes, must be open and accessible. Ideally, navigation of legislative material will be improved. Participants also were interested in automatic redlines for all proposed legislation.
Elections & Government Officials: The transparency community has a strong interest in information regarding government officials and financial incentives. The best way to get that information is to standardize and publish the donation data. Specific legislative initiatives identified included the Senate Campaign Disclosure Parity Act, Lobbyist Disclosure Enhancement Act, and Presidential Library Donations Reform Act. Data transparency principles should be applied to all aspects of lobbyist and donor data including standardized identifiers for lobbyists and donors, permanent URLs for data regarding elections and candidates, information about ex parte communication, tax credit information for lobbying groups and trade associations, and all data should be updated in real-time.
Continuing the Discussion
As the Data Transparency Agenda continues to develop, the Coalition recognizes the importance of contributions from data transparency advocates. We want to hear from you!
What do you think of the suggestions from Transparency Camp? Are we missing a major domain of federal data or do you have an alternative organizational structure to suggest? What initiatives are urgent and what are long term goals? What else should we include?