Science and Democratic Policy in a Data-Driven World

Public access to data behind regulations should not be a political question.

Data have no politics.

It is, for this reason, lamentable that a recent proposed rule from the U.S. Environmental Protection Agency (EPA) dealing with the transparency of data used to develop regulations is being evaluated for its effect on political battles. The rule would require EPA to make regulatory decisions only based on research with publicly available data, unless EPA’s Administrator grants a waiver when public data access is impossible to provide, consistent with law, privacy, or security.

Ideally, public policies will be factually based and non-partisan. However, politicians will always hold the reigns of regulatory agencies, and politicians will seek to influence regulatory decision-making for political purposes. As a check on the influences of politics and ideology that result from political control of regulatory agencies, citizens need for agency officials to rely on transparent, data-driven approaches to policymaking.

Public access to the data underlying regulatory policies allows the public to see how much politics may be influencing those policies and to challenge them if they become unduly political. Public access would also facilitate meaningful public evaluation, collaboration, and input into the policymaking process, which is inherently messier in a democratic, participatory society than in a technocratic society.

Even if EPA’s recent proposal is itself politically motivated, it is consistent with the transparency initiative of the previous Administration. President Barack Obama’s Memorandum on Scientific Integrity declared, “If scientific and technological information is developed and used by the federal government, it should ordinarily be made available to the public. To the extent permitted by law, there should be transparency in the preparation, identification, and use of scientific and technological information in policymaking.”

To implement the President’s policy, the director of the Office of Science and Technology Policy provided guidance to federal agencies, saying that they “should expand and promote access to scientific and technological information by making it available online in open formats. Where appropriate, this should include data and models underlying regulatory proposals and policy decisions.”

EPA’s proposed policy is also consistent with independent judgments about regulatory policy. The Bipartisan Policy Center recommended that “studies used in the formulation of regulation should be subject to data access requirements equivalent to those under the Data Access Act…regardless of who funded the study.” The Data Access Act requires data produced using federal funding to be made publicly available through the Freedom of Information Act (FOIA). The Bipartisan Policy Center’s recommendation would expand the pool of data available upon request to the public to studies supported with nonfederal funding.

In July 2013, the Administrative Conference of the United States (ACUS), an independent agency that recommends improvements to administrative procedure, recommended that, “to the extent practicable and permitted by law and applicable policies, each agency should identify and make publicly available (on the agency Web site or some other widely available forum) references to the scientific literature, underlying data, models, and research results that it considered.” Although agencies must follow the Information Quality Act Guidelines from the Office of Management and Budget, the ACUS recommendations provided that “each agency should ensure that members of the public have access to the information necessary to reproduce or assess the agency’s technical or scientific conclusions.”

None of these directives or recommendations have been implemented by any federal agencies. EPA is the first agency to propose implementation of a data access standard for policymaking.

Data access should not be the rule for only EPA. After all, it does not use scientific research in any unique way. All federal regulatory agencies should operate under a data access rule for their significant regulatory policies as recommended by the various entities mentioned above.

Opponents claim that EPA’s proposed policy will lead to policies with weaker scientific justification. The glaring implication of this claim is that there is a subset of science that is incompatible with transparency. Yet many of the most prestigious scientific journals have policies very similar to the EPA proposal, requiring data access as a condition of publication, with special provisions when doing so is impractical. The journals require data archiving because the only way to replicate research results is to have access to the data.

Why should professional scientific journals require more transparency than federal regulatory bodies? Research has demonstrated a relationship between the willingness of scientists to share data and the strength and quality of the research results.

Much of the political debate over the recent EPA proposal relates to concerns about the ability of the agency to use the results of controversial, federally funded, air-quality health research conducted in the 1970s and 1980s—known as the Harvard Six-Cities Studies. Despite years of requests, the authors of that research have refused to make their data publicly available or even to turn the data over to EPA, which funded the research. The debate over that 25-plus-year-old research is, however, somewhat moot because recently researchers have been able to use publicly available Medicare data to measure the effect of air quality on the health of older adults.

The authors of the Harvard air quality health studies have two reasons for not providing access to the data. First, they point to legitimate concerns about the personal privacy of the study participants and their personally identifiable information. The National Research Council and the American Statistical Association, though, have identified a number of strategies to mitigate privacy problems with data access. Furthermore, EPA’s proposal allows exceptions to data access. Just because data access may be difficult or impossible for some research, that should not mean that data access should never be expected for any research.

A second reason the air quality researchers give for not making the data accessible is that it would open the research results to inconvenient questions and challenges. When the authors of a Harvard Public Health Magazine article interviewed the researchers of the Six-Cities Studies, the researchers were candid about the reservations that they had about data access beyond privacy issues. One person involved in the research stated that “we knew that if we released the data, it would be endless aggravation and defending against attacks…To have a hostile group combing through your data looking for anything to attack you about was not something any of us relished.” Another researcher allegedly told the head of the study that “to release the raw data would be to allow ‘biased groups’ to manipulate it and to set a precedent that ‘will undermine future research by academic institutions.’”

Scientists understandably have reservations about having their analyses scrutinized and second-guessed, but the years of controversy about the Six-Cities Studies demonstrate that prohibiting data access does not protect research from questions, criticisms, and attacks.

The great arc of history is bent toward transparency, openness, science, and data. Civilization has advanced to where it is today because people, after gaining access to information previously withheld from them, have questioned the conclusions drawn from the previously collected information. In the long run, transparent data will benefit society far more than it will benefit partisan agendas in the short run.