Influenza and Productivity
This week didn’t seem great for productivity, but really, the situation was much milder than first thought.
The government reported their indicator for non-agricultural productivity. It’s about worker output per hour, which increased at an annual rate of 1.8% last quarter, as opposed to the 2.8% growth that had been initially projected. That’s a bit worse than the economists’ adjusted expectation of 2%.
Unit labor cost has raised some concerns. There’s a fear that the cost for employers to produce each unit might be adjusted from 2.8% to 4.4%, potentially triggering an inflationary cycle that would drive wages higher.
Yet, this downturn wasn’t as alarming as it might seem; workers didn’t suddenly become less productive. It’s more about lower productivity versus increased costs.
The adjustment mostly echoed a previous resolution by government statisticians, who revised the economic output for the fourth quarter downwards from 1.4% to 0.7%. When you decrease production without cutting working hours, you naturally end up with reduced productivity.
The concept is straightforward. Productivity measures how much output is generated each hour. If the government finds the economy is creating less than previously estimated, while labor input stays about the same, it naturally leads to a decrease in productivity. That’s essentially what occurred. The U.S. Bureau of Labor Statistics adjusted the non-farm output growth for the fourth quarter down from 2.6% to 1.5%, while the reported hours didn’t change in this updated calculation. If productivity drops significantly, unit labor costs will automatically rise unless compensation decreases enough to offset it. But it didn’t. So the perceived new issue with labor costs was largely an arithmetic outcome of the noted drop in output.
A more intriguing question then is why production fell so much.
A Little Bad Data Goes a Long Way
The answer is actually more revealing than the usual economic downcast. The BEA pointed out that the most substantial downward revisions on the consumer side were from sectors like services, particularly health care—including hospital and nursing home services. Additionally, some estimates focused on factory construction and software were also adjusted downward. Essentially, the economy didn’t experience widespread declines all at once; the price reductions were limited to a few areas, with health care seeing the largest drop.
So, why did medical production take a hit towards the end of last year? It wasn’t because health care quality diminished or that people suddenly became unwell. In fact, it’s quite the opposite—demand for healthcare decreased more than initially thought. According to the Census Bureau’s Quarterly Services Survey, receipts from health and social assistance grew just 0.5% in the fourth quarter, dropping from a 3.0% increase in the third quarter. BEA specifically mentioned the new QSS data as the reason for the downward adjustment in health services.
The lower demand indicated we were healthier than expected. Data from the Centers for Disease Control showed a decline in hospitalizations for severe respiratory illnesses during the latter half of the fourth quarter, which aligned with a milder illness season than typical for winter. Fewer sick individuals mean fewer hospital and outpatient visits, which ultimately affects health output metrics. So, while a gentler holiday season due to flu and coronavirus doesn’t look good for GDP calculations, it’s not necessarily bad for everyday people.
Interestingly, employment in health care didn’t collapse. In fact, health worker salaries kept rising throughout the quarter. That’s what you’d expect in the medical field. Hospitals and clinics don’t typically start laying off staff just because the respiratory season calms down. Workers are still on the job, and they continue getting paid. But with fewer people needing treatment, the output measured per worker falls. This sets up a situation for lower productivity and higher unit labor costs, even if the underlying labor situation remains solid.
Putting this all together, it feels more like a statistical domino effect than a substantial macroeconomic failure. Initially, the BEA revised output down, primarily due to less-than-expected medical utilization. The BLS then took that weaker output and applied it to a productivity calculation. Consequently, productivity numbers dipped while labor costs increased. These figures can seem concerning on their own, but understanding the context makes them less alarming.
This batch of bad macro news didn’t indicate that American workers suddenly forgot how to work. Instead, it pointed out that fourth-quarter output—particularly in health care—had been oversold to begin with. Sometimes, the economy can look weak simply because people are sick. In this case, certain parts of the economy appeared to falter, as many were not as unwell.
This irony is hard to overlook. A major factor contributing to reduced output may well have been that the nation was healthier than statisticians initially projected.
