This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American
Members of the Commission on Evidenced-Based Policymaking appear before Congress this week to present their final report (pdf). In a city otherwise riven by partisan battles, the commission’s work is good news and deserves attention. It’s the brainchild of House Speaker Paul Ryan (R–Wis.) and Sen. Patty Murray (D–Wash.). The premise is simple: Federal government data can help produce better informed decisions. “For example, according to the report, “access to administrative data has improved federal investment in workforce training by making possible comparisons in changes in quarterly earnings between different training and employment programs.”
There are some caveats, though. Much of the information in question is personal data, and there are real privacy concerns when particulars about people are combined across federal agencies. That could lead to secretive profiling and unfair and unaccountable decision-making. In fact, a proposal in the mid-1960s to create a national data center—also to ensure better policy-making by the federal government—produced such a backlash that Congress enacted comprehensive legislation to limit federal agencies’ ability to gather and collect personal information.
In passing the 1974 Privacy Act Congress recognized that statistical data, such as the U.S. census, can help inform government policy while minimizing privacy risks. And so, the same law that limits use of personal data recognizes the critical role of statistical information to promote smart government.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Over time technology has chipped away at the ability to separate statistical from personal data. The use of microdata, new techniques for “reidentifying” people whose data had supposedly been made anonymous, commercial incentives and even criminal hacking could foil even the best attempts to prevent individuals from being identified. The commission urged the adoption of privacy enhancement and preservation techniques—such as the use of differential privacy algorithms that glean information from a person’s data without identifying that person—to minimize privacy risks. It also recommended the creation of a national secure data service to manage multisource data projects without creating a centralized repository. And in a related report the National Academies of Sciences, Engineering and Medicine has stated that agencies must be responsible for ensuring that privacy techniques work as advertised.
Still, the larger goal of promoting better use of government data is laudable. Remarkably, in the privacy field some of the most sensitive data in government is made available to the public. For almost 50 years the Administrative Office of the United States Courts has provided detailed reports on the use of surveillance authority in the U.S. Those reports document the use of wiretapping by the federal government—its frequency, cost, reasons for use (most wiretaps concern drug investigations not terrorism) and effectiveness. The data makes possible meaningful oversight of law enforcement activities. And even when law enforcement agencies and civil liberties advocates debate the scope of government surveillance, there is a common data set that informs the discussion. Remarkably, the data is made available without compromising investigations or revealing the names of individuals
Of course, a lot of the data collected by federal agencies does not threaten privacy at all. Consider the important work of the National Oceanic and Atmospheric Administration (NOAA) and its recent tracking of hurricanes across the Atlantic Ocean. Meteorological data is becoming increasingly important as the severity of storms increases. State and local officials need the best information possible to make difficult decisions about how to deploy resources and when to evacuate residents of coastal communities. Our ability to forecast storms has improved dramatically from 1900, when a hurricane—which still ranks as the most deadly natural disaster in our nation’s history—took the lives of approximately 10,000 people in Galveston, Texas. The reason for the high death toll was the lack of accurate forecasting and therefore the lack of preparation. But there is more to do.
Even today, U.S. hurricane forecasting lags others. The European Center for Medium-Range Weather Forecasting (ECMWF) model is now considered the premier global model in the world. And so, U.S. weather forecasters often superimpose the ECMWF tracks when discussing recent storm activity in the U.S. The U.S. needs to do better, and data in the federal government is a good starting point.
With the report on Evidence-Based Policymaking completed, the work of the commission will now be transferred to the Bipartisan Policy Center. That is also a hopeful sign. The formulation of evidence-based policies should remain a priority, regardless of party affiliation.