Time for action to improve government data analysis

Even if we cannot agree whether government is too big or too small, we should be able to agree that we want government to be as effective as possible. Using taxpayer funding efficiently requires having reliable information about what government policies are actually accomplishing.

Lawmakers and administration officials come to Washington with the laudable intention of making government work better, but often soon realize just how complex many of the problems the nation faces are and how difficult it can be to know whether policy goals are being achieved. Over the last 20 years, better monitoring of agency performance has helped to identify and correct gaps in the services provided to the public, but performance metrics too often are indicators that reflect what can easily be measured rather than what is most important.

We can do better. There is strong bipartisan support for making more effective use of data already being collected to determine whether policies are having their desired impact. The Commission on Evidence Based Policymaking unanimously supported this goal in its 2017 report. Making better use of the data the government holds is something that was emphasized by former President Obama and recently highlighted in President Trump’s management agenda.

While there are considerable benefits to using government data to provide insights about how programs are working, some observers see the increased use of data for these purposes as a threat to individual privacy. Fortunately, advances in statistical methods and computing technology provide ways to protect data while extracting value.

The value of new analyses based on existing data can be considerable. Consider the efforts led by Stanford University researcher Raj Chetty to link together government collected data on income, education, housing, and patents for a series of landmark studies. His team removed personal information before analyzing the data. The findings suggest policy approaches to closing the innovation gap among the members of low income and minority communities. This work has changed the discussion around innovation as well as around economic mobility and inequality.

In another research project, Bruce Meyer of the University of Chicago and his former student Nikolas Mittag have linked administrative data for welfare programs to the current population survey, the data used for official poverty statistics. They have found that the survey greatly underestimates program benefits for low income families, making efforts to fight poverty appear much less successful than they in fact have been.

These are just two examples of research that would not have been possible without secure access to confidential government data. Both projects demonstrate the value of enabling this type of access and show that new insights can be realized while protecting individuals. Today, researchers finding ways to improve the effectiveness of government programs must navigate a daunting bureaucratic process. They also must work within a poorly defined legal framework that is not necessarily supportive of the government need for good information to guide policy.

Thanks to the leadership of Speaker Paul Ryan (R-Wis.) and Senator Patty Murray (D-Wash.), cosponsors of the Foundations of Evidence Based Policymaking Act, all of this soon could begin to change. The bill, which unanimously passed the House in 2017, now is awaiting consideration in the Senate. The legislation begins to implement recommendations that were made by the commission we formerly chaired and that we continue to champion because of the impact they would have on making important federal data more available for research.

The bill will strengthen privacy protections without expanding government bureaucracy, help researchers securely access the data they need with strong controls to protect personal information, increase transparency about what data government collects and how those data are used, encourage government to make data not sensitive more publicly available, establish processes for government to articulate where better information is needed to inform future decisions, and begin to change government culture to more readily engage in evidence based policy.

Without this legislation, our data infrastructure and evidence culture will remain a relic of the 20th century. The Foundations of Evidence Based Policymaking Act will help agencies achieve overdue improvements and provide government with the tools it needs to operate more effectively and efficiently. It will go a long way to achieving the accountability that the American public demands of its government. It is time for the Senate to take action and move the bill forward.