In light of recent disappointing job numbers and revised previous statistics, President Trump has opted to let go of Erica Mantelfer, the director of the Bureau of Labor Statistics. However, many believe that firing someone like Mantelfer, primarily for political reasons, is unlikely to aid the economy or advance the president’s agenda.
The rationale behind this decision seemed to hinge on a perceived political bias in the reported job growth figures. The Director of the National Economic Council even labeled the job data as “bad.” But, if you delve into how this data is actually gathered, and consider recent adjustments, it becomes apparent that such criticisms may not be fully justified.
Much like the BLS, other federal statistical agencies also maintain a clear and transparent methodology for collecting data. They survey a sample of workplaces about wages and hours worked on the 12th of each month, providing an initial estimate about three weeks later. Since not all participants respond promptly, the first release is considered tentative. Over the next couple of months, the BLS revises this data as more responses come in.
So, what’s next? Historically, if we analyze the initial and second batch of payroll revisions, the BLS, on average, performs quite well. For instance, since November 1964, the average increase reported in revisions is about 5,400 jobs—barely a blip in the grand scheme of non-farm payrolls.
Yes, recent years have seen average negative revisions, which isn’t extraordinarily unusual, and most of these tweaks tend to be minor. The June revision, while notable with a drop of around 133,000 jobs, doesn’t make it stand out among the top 20 downward revisions since the 1960s.
It’s noteworthy to mention that if you’ve never faced a dramatic change in your payroll data, you might be surprised. Data collection challenges are continually mounting. Budget constraints and shifts in federal employment mean the department struggles to gather the necessary information. Moreover, response rates to surveys have been declining for years, which will inevitably lead to less reliable data and more erratic estimates.
The politically charged move to dismiss the BLS commissioner could dissuade even the most dedicated employees from enhancing survey response rates. If you believe the figures are fabricated, why would you take the time to fill out the Labor Department’s survey?
Moreover, questioning federal statistics could backfire on the Trump administration’s economic initiatives. Businesses often rely on government data for decisions involving hiring, wages, and investments. If data integrity is in doubt, firms tend to act with caution rather than confidence. Accurate estimates from the BLS and other agencies are essential for guiding decisions that allow policymakers to understand the economy better and evaluate the potential effects of legislation.
Perhaps the president thinks that by replacing the commissioner, he’ll get more favorable job numbers in future reports. But managing perceptions without fundamentally reworking how data is collected and calculated is quite a challenge—one that the public will likely notice.
A troubling example to consider is Argentina. In the wake of soaring inflation in 2006 and 2007, their government dismissed statisticians responsible for calculating consumer price indices, skewing data to fit a more favorable narrative. This manipulation backfired, as it did little to convince the public or financial markets. Argentinians soon ceased to trust official statistics altogether.
Relying on rosy forecasts won’t help the administration gain public confidence. The dedicated professionals who generate these statistics value public trust deeply and might respond similarly to their Argentine counterparts.
Firing Erica Mantelfer was a misstep. Although it’s unlikely Trump will change his mind, Congress should take action. Federal statistical agencies require adequate funding to enhance their capabilities and maintain their independence from political pressures. However, meaningful change is unlikely without advocacy from business leaders, local governments, and the communities that utilize these statistics. In essence, they need to hear from us.
