Without independence from political control, official statistics lack credibility and value.
W e must repair our nation’s trust in the Bureau of Labor Statistics (BLS) and in federal statistics generally. That work must begin immediately to avoid the U.S. having to bear the cost of the level of economic uncertainty normal in China and other countries where businesses, investors, or ordinary people have lost confidence in their official statistics. Fortunately, this repair is doable. As recent BLS commissioners, we offer some ideas on how to do this. In our view, Congress, the White House, the advocacy community, employers, and the courts all have a role.
Repair is needed because, after BLS (following its usual objective methods, which always includes revisions of the two previous months’ numbers) reported evidence from hundreds of thousands of U.S. employers that the labor market is weakening, President Trump shot the messenger. He ordered that the BLS commissioner be fired, an unprecedented action, asserting without evidence that she “rigged” the data. These false allegations damage the credibility of the agency in the eyes of Americans who trust the president. His attempt to subject BLS to political pressure can only damage its credibility.
Without independence from political control, official statistics lack credibility and value. For an example of life without statistical agency independence, think of China. The numbers produced by its government’s statistical agencies bear little relation to reality and are not based on scientific principles. Their “statistics” match the plans and forecasts political leaders establish in advance. If the agency releases estimates that contradict the party line, the staff responsible would likely lose their jobs. Consequently, Chinese official statistics are almost worthless; the economy is difficult to understand; and employers, investors, and citizens cannot make well-informed economic and social decisions.
Our remedies begin with an important role for the Senate. The next BLS commissioner and the nation both deserve a rigorous and rapid confirmation process. The confirmed candidate should have the qualifications needed to reinforce trust in BLS’s credibility and excellence. The candidate must have strong management experience, deep expertise in economic statistics and the respect of other experts, extensive engagement with the federal statistical agencies, familiarity with BLS and an understanding of its products and broader relevance, visibility in the statistical community, an ability to interact effectively with both Congress and senior Department of Labor staff, and a commitment to follow the National Academies’ Principles and Practices for a Federal Statistical Agency. They should also have a record of understanding that trust and data integrity are mission-critical for BLS. Such a candidate would garner broad, bipartisan support in the Senate, and such support would demonstrate their suitability. The current nominee, E. J. Antoni, should be closely scrutinized on these important characteristics. By following these time-tested, nonpolitical guidelines, working thoroughly and transparently, and considering the Bureau’s restoration and future, the Senate can help restore public trust in the Bureau’s work.
We also have suggestions for the White House. The administration must reverse the impact of policies in place since January that have forced BLS to cut data collection efforts, reducing data quality and breadth. The agency must be allowed to hire staff to replace fired recent hires, early retirees, or people who took a deferred resignation. Below the commissioner, eleven of BLS’s 35 top leadership positions are vacant. Overall, BLS has probably lost about 20 percent of its staff, and the number is mounting. BLS must be permitted to sign contracts for needed services and to reinstate the advisory committees that gave external experts visibility into changes at the agency. The administration should also support a BLS budget that would allow the agency to modernize and improve its products, rather than suggesting an 8 percent cut.
Congress as a whole should also take several steps. It should exempt the federal statistical system from ongoing administration efforts to bring more federal staff under its direct control. The administration is moving a personnel reform agenda through a rulemaking process that would convert many civil service roles into political roles, creating the ability for political leaders to fire far more agency staff for failure to toe a political line. This would undermine the independence and integrity of the official statistical system.
Second, Congress should investigate the circumstances leading to the commissioner’s firing.
Third, Congress should invest in improving and modernizing BLS methods. To maintain trust, statistical agencies must continually modernize. The Information Age provides new opportunities to use non-survey data to improve products and the reliability and efficiency of agency processes. Improvement requires designing, standing up, and running a parallel system to current methods during a testing period. That process requires new up-front funding.
Fourth, Congress should clarify the limited set of conditions under which a term appointee may be removed by the president. Legislation appears to be needed to spell out how term positions differ from serving at the pleasure of the president to advance the president’s policy agenda.
Advocacy groups — including business and industry associations, donors, PACs, and others — must recognize their responsibility to protect federal statistics. They rely on official statistics to support their interests, advocate effectively, and to judge policies and policymakers. These influential groups must expand their advocacy agendas to include support for the independence and funding of statistical agencies that will ensure the integrity, accuracy, and modernization of the evidence they need.
Employers can and should help make revisions and benchmarking adjustments smaller right now. Every C-suite leader should tell their staff that responding to federal statistics surveys is a priority, especially for the Current Economic Statistics survey. Too many employers, including some of our largest companies, either do not perform or do not prioritize this essential but voluntary public service. This support from business leaders would directly improve data quality.
Finally, cases involving presidential power to dismiss presidentially nominated, Senate-confirmed leaders of independent agencies without cause are currently working their way through the courts. We do not fully agree on the ideal policy here. We do, however, agree that the courts should recognize the principle that statistical agencies’ effectiveness hinges crucially on their credibility and, thus, their protection from political interference. These agencies have no policymaking power, only the power to credibly inform. If they aren’t credible, they are pointless.
For over a century, congresses, presidents, the courts, advocacy groups, and agency leaders have jointly built federal statistical agencies like the BLS with features that assure their independence and credibility. These agencies have served the nation well by expanding evidence-based decision-making. They have been emulated across the world. To avoid turning away from an approach that has served this nation well, the White House, Congress, the courts, employers, and advocacy groups must take action now.
William Beach, former Bureau of Labor of Statistics commissioner, is a fellow at the Calvin Coolidge Presidential Foundation. Erica L. Groshen, senior economic adviser at the Cornell University-ILR School, was the 14th Commissioner of the US Bureau of Labor Statistics