Erosion of Federal Statistical Agencies Puts Sound Policy at Risk | Commentary
Government statistics not believed by the public are useless. It is in the interest of no one — no interest group, no party — to threaten their credibility. Hence, governments around the world strive to choose technically accomplished, independent leadership to fend off any hint of political interference in government statistics. It is a bedrock principle of democracies.
In the United States, a key tool to achieve such independence is to assure that heads of agencies are both appointed by the president and are able to survive the review of 100 senators elected by the people. Letting appointments be controlled only by the administration in power risks perceived lack of independence, or worse, real lack of independence. The U.S. may be in trouble on this score.
During the frenetic few days Congress was in session in September, the Senate Health, Education, Labor and Pensions Committee rubber-stamped HR 4366, the House-passed bill reauthorizing the Department of Education’s Institute of Education Sciences, which includes the National Center for Education Statistics and three research centers. In an attempt to streamline the IES bureaucracy, both chambers sacrificed the NCES’ autonomy and stature by removing presidential appointment of the commissioner and relegating control of key operations and decisions to IES.
If, as expected, the bill is signed into law in the lame-duck session, Congress and the administration would complete the weakening of the NCES that started in 2002, when it was first placed under IES and further undermined in 2012 with removal of Senate confirmation of its head.
Why the concern? For statistical data to be perceived as sound and independent, its producing agencies must completely control the data’s collection, analysis, and publication, including control over its planning, budget, press releases, and information technology. Such autonomy affords the agency the ability not only to withstand improper outside influence but also, just as importantly, to deflect the perception of such.
For an extreme example, consider President Richard M. Nixon’s attempts to influence the unemployment figures released by the Bureau of Labor Statistics, which failed only because of the independence of the commissioner at the time. Indeed, based on our experiences heading the nation’s largest statistical agency, we can affirm that the career statisticians, economists, social scientists and other professionals in the federal statistical agencies are devoted to producing high-quality, objective statistical information for the country. They need strong, independent leaders to enhance this culture of excellence and buffer it from intrusion.
Having an independently Senate-confirmed statistical agency head not only benefits Congress and the public, it also benefits the administration lest detractors of their proposals and policies try to suggest the administration had influence over the data. Further, it gives the agency itself sound statistics on which to build priorities, including priorities for research and development investment.
Autonomy without relevance is wasted, however. Presidential appointment and Senate confirmation (PASC) of the statistical agency head, along with inclusion in secretary-level discussions, is important. The prestige and vetting provided by PASC strengthen an agency head’s position to stand by the integrity of the data and to inform policy making with useful data. By involving a federal statistical agency head in secretary-level discussions, the head can lead his/her agency to provide the most useful statistical information to serve the country.
We urge Congress to approach the pending bill with deep devotion to the quality and independence of the federal government’s statistical agencies. A small, bipartisan legislative modification could make a difference for education and statistics. It also would mean one less blow to the credibility of the statistical system worldwide.
Robert M. Groves is provost at Georgetown University and served as director of the Census Bureau from 2009 to 2012. Kenneth Prewitt is Carnegie professor of public affairs and vice president for global centers at Columbia University, and served as director of the Census Bureau from 1998 to 2000.