David Warsh: Time to build new public data infrastructure
SOMERVILLE, Mass.
“Philosophy? Philosophy? I’m a Christian and Democrat — that’s all.’’ Franklin D. Roosevelt responding to the question “What is your philosophy?” as quoted in Freedom From Fear: The American People in Depression and War 1929-1945, by David Kennedy.
It’s a commonplace that every new presidential administration arrives as a policy omnibus, with all kinds of venturesome policy entrepreneurs aboard. Campaign managers, financial backers, friends of the president and vice president, Cabinet members, Congressional committee chairpersons and their staffers, lobbyists, opinion shapers in the media – all bring agendas and, in the pell-mell of the first hundred days, seek to put them in motion.
It will be decades before we know what the Trump presidency was all about. But Biden, if he wins, will take office, with only one certainty. The Democrats will once again be the party of Innovation. Starting with candidate Barry Goldwater in 1964, the Republican Party managed to wrest away the mantle of change: deregulation, globalization, supply-side economics, all that. But after the monumental stumble that has been the Trump presidency, the GOP must rebuild itself as the party of conservatism or go out of business.
Trump himself may have arrived with a headful of ideas, but in the event he is defeated after a single term he is headed for obloquy greater than that of Herbert Hoover as a president who didn’t show up. Should Trump win a second term, of course, that’s another story.
Green New Deal? Tax restructuring? Pandemic crisis management? Health care and Social Security repair? Immigration status reform? Supreme Court appointments? Which of a thousand possibilities will blossom into fact? There is no way of knowing today. As a gauge of the Biden administration’s long-term success, however, consider a modest suggestion. What about commissioning a National Data Service to stand with the Army, Navy, Marine Corps, Air Force, Coast Guard, Space Force, Public Health Service and National Oceanic and Atmospheric Administration, as the nation’s ninth uniformed officer corps?
The pre-history of the Space Force is instructive. As far back as 1961, then Defense Secretary Robert McNamara designated the Air Force as the lead military service for space. But not until 1982 was an Air Force Space Command created, in connection with the Reagan administration’s Strategic Defense Initiative, a satellite-based laser weapons system abandoned after some testing. Satellite reconnaissance became part of the War on Terror after 9/11. And in March 2018, President Trump embraced the independent Space Corps idea in a public speech. He signed a statute in December 2019 establishing it. Around 16,000 active-duty Air Force personnel and civilians are assigned to it.
Public data collection in the United States is a much older service. Article I, Section 2 of the Constitution requires the enumeration of the nation’s population for purposes of establishing representation in Congress. The first Census was conducted in 1790, and at ten-year intervals ever since. New statistical agencies were added as needs arose, for banking, agriculture, labor, safety, manufacturing, transportation, health, medicine, and so on. When the War Department (as what is now the Defense Department was previously called) needed planning data for World War II, the Commerce Department turned to the National Bureau for Economic Research, where Simon Kuznets was developing National Income Accounts. Most countries have a single statistical agency; the U.S. has 13 major agencies and hundreds of smaller programs.
The current system has needed an overhaul since the mid-1990s, when Janet Norwood, then Commissioner of the Bureau of Labor Statistics, wrote Organizing to Count: Change in the Federal Statistical System. Today it is judged to be broken. Old-fashioned survey techniques are outdated in the age of the Internet. Concerns for privacy and confidentiality have sent costs soaring. The 2020 Census is expected to cost $48 a head to collect in constant dollars, as opposed to $5.50 in 1960. And political meddling has become a problem in the last four years. Manipulation of pandemic statistics has made headlines in recent months. And, having failed to exclude non-citizens by requiring them to identify themselves as such, the Trump administration decided to end the Census four weeks early, in August.
A high-level blueprint for building a new public data infrastructure appeared over the summer. Democratizing Our Data: A Manifesto (MIT Press), by Julia Lane, of New York University. A native New Zealander, now a U.S. and U.K. citizen as well, Lane earned a PhD from the University of Missouri in the 1980s, taught, worked for the World Bank for a decade, taught again for another 10 years at American University, managed the economics department at the University of Chicago’s National Opinion Research Center, and rotated through four years at the National Science Foundation as a senior program director. She subsequently founded a series of programs for the American Institutes for Research, and eventually the Coleridge Initiative, a rapidly growing research and training collaborative, before settling down as a high-end professor at New York University’s Wagner School of Public Policy. “It’s a golden moment,” she writes, to rethink the system, retaining its best aspects – trust, professionalization, continuity – while designing new systems of measurement, putting them to work, and making them widely available
Will it happen? It’s a long road, a matter for experts backed by diverse constituencies in government and private enterprise. But if ever a highly technical program was worth setting out, manifesto-fashion, in a short and readable book, it is this one. An age of Big Data is upon us. Shouldn’t a government run by a Party of Innovation bend to the task of creating a National Data Service?
David Warsh, a veteran columnist and an economic historian, is proprietor of Somerville-based economicprincipals.com, where this essay first ran.