, , , , , , , , , , , , , , ,

Wow! There were so many important results flowing from last week’s final (for now) government data on third quarter U.S. labor productivity, I hardly know where to begin. (I’m also feeling a little sheepish about waiting so long to report on these data, but it’s just another sign that we’re living in a target-rich commentary environment, as RealityChek‘s motto suggests.) But after finishing this post, I feel confident you’ll agree that the big downward historical revisions to manufacturing labor productivity growth deserve the most attention.

Not that the new figures on overall labor productivity (for the so-called non-farm business sector) were anything to sneeze at. These new numbers cover the narrowest measure of productivity (gauging only output per hour worked by an individual American employee) but as known by RealityChek regulars, they’re issued on a much more timely basis than the multi-factor or total factor data – which measure the output generated by a wide range of inputs.

And this update on third quarter labor productivity confirmed that it grew at its highest sequential annualized rate (2.95 percent) since the third quarter of 2014 (4.34 percent). The revised labor productivity gain was actually a touch smaller than the originally reported 2.97 percent rise, but not nearly enough to change the overall story. If this rate of improvement continues, that would be excellent news, since strong productivity growth is an economy’s best bet for a sustainable increase in economic growth and living standards.

At the same time, analyst David P. Goldman has noted that the new data add to compelling evidence that recent years have seen a reversal in the relationship most economists have long assumed (and that was borne out by by these same statistics) between unit labor costs (a main labor component of the productivity statistics) and the broadest measures of unemployment. As Goldman just observed, normally, they move in opposite directions – i.e., when joblessness is rising, the price of labor generally (and logically) falls, and vice versa. But since 2014, unemployment has kept tumbling, but labor costs have fallen as well. If this trend continues, that would be much worse news, since it would undermine the portrayal of productivity growth as a boon to the nation’s workers. (And a richly deserved hat-tip to a Twitter follower of mine, who goes by “Field Roamer” for calling my attention to this post.)

It’s been clear throughout this current U.S. economic recovery that wage growth has been unusually weak, but Goldman’s post paints the paycheck picture in a much grimmer light, and that’s definitely worth exploring further.

But to me, the manufacturing revisions deserve center stage, both because of their magnitude and the long time frame they cover – all the way back to 1987, when manufacturing labor productivity began to be tracked. I’ll let the Bureau of Labor Statistics (BLS), which calculates productivity for the U.S. government, summarize its dreary conclusions:

A large upward revision to the change in the annual manufacturing productivity index from 2008 to 2009 was more than offset by downward revisions in adjacent years, and the average annual rate of growth from 2007 to 2012 was revised down from 2.9 percent to 1.2 percent. The average annual rate of manufacturing productivity growth during the current business cycle from 2007 to 2016 was revised down from 1.6 percent to 0.9 percent, and the long-term rate for the entire series from 1987 to 2016 is now 2.8 percent, compared to the previous estimate of 3.2 percent.”

In other words, over roughly the last thirty years, labor productivity in industry has risen 12.50 percent more slowly than previously reported. And manufacturing’s performance on this crucial front wasn’t great to start with.

Another way to look at the new numbers is to see how they affect what we know of America’s manufacturing labor productivity performance during the most recent economic expansions – a method that gives us the best apples-to-apples data. If your jaw doesn’t drop, it should.

The 1990s expansion still comes across as a period of robust manufacturing labor productivity growth. The cumulative increase was downgraded only from 46.81 percent to 45.94 percent.

But check out the new results for the previous decade’s recovery. Viewed through the lens of the old productivity data, its performance was excellent, and surprisingly so. After all, this expansion was fueled by the inflation of the credit and housing bubbles whose bursting led to the global financial crisis and the Great Recession. Yet the BLS had been saying that its cumulative productivity gain was 41.23 percent – just about as good as the 1990s advance factoring in this recovery’s shorter duration.

The new numbers – only 30.08 percent manufacturing labor productivity growth – are much more consistent with the idea that the previous economic recovery was marked largely by phony, unsustainable growth.

And as for the present recovery? The old data already made clear what a productivity disaster it’s been. Though it’s lasted nearly as long as the 1990s expansion, the previous BLS data pegged its total manufacturing labor productivity growth at only 20.93 percent – just about half the rate generated during the 2000s expansion.

The new rate? Only 9.41 percent, meaning its been cut nearly in half. Moreover, according to the new figures, the current recovery’s manufacturing labor productivity growth rate represents a much greater deterioration from the performance of the bubble recovery than had been reported. At least by this measure, American economic growth was already appearing even less healthy these days than it was leading up to the last meltdown produced by fake prosperity. Now this problem looks much worse. 

In addition, don’t forget:  Even these dreadful numbers probably overstate manufacturing labor productivity’s advances. Why? Because as the BLS acknowledges, its methodology for calculating this indicator include the effects of offshoring: simply substituting foreign workers for American workers. Since the total number of workers doesn’t change, the productivity figures for U.S. factories and related facilities are artificially inflated – and for reasons having nothing to do with greater efficiency. 

And these results raise all sorts of perplexing questions. For example, I’ve been arguing for quite some time that the overall slowdown in American labor productivity growth must surely stem from the trade- and offshoring-related losses of so much domestic industry – which has generally been the economy’s productivity growth leader. But the new BLS statistics indicates that there could be a bigger labor productivity growth problem within manufacturing itself. Alternatively, these losses could have been concentrated in especially high-productivity sectors of manufacturing – or trade and offshoring have had little or nothing to do with the problem to begin with.

More light could be shed on these questions by comparing America’s manufacturing labor productivity performance with that of other countries. Has it been better? Worse? Some short-range data I’ve seen indicate that the slowdown has been widespread across the globe, at least between 2015 and 2016. But I need to dive much deeper into these statistics to draw firmer conclusions.

Further, how significantly will these new labor productivity results affect the broader multi-factor productivity results? BLS hasn’t scheduled its next report on this indicator, so my oft-used advice to “stay tuned” applies to me, too, in this case.