Follow by Email

Friday, March 30, 2018

Jared Bernstein: NYT oped: When it comes to trade-induced job loss, “don’t worry, be happy!” [feedly]

NYT oped: When it comes to trade-induced job loss, "don't worry, be happy!"

I've long hoped, probably naïvely, that one of the benefits of team Trump's promotion of generally ineffective (or worse) solutions to the downsides of trade could engender a debate about better ideas. Of course, the debate will also generate some really bad arguments, like this one from economist David Boudreaux in this AM's NYT.

Boudreaux argues that trade (and, implicitly, anything else) can't be a problem for jobs because the US economy creates and destroys tons of jobs all the time. The nub of his case comes down to:

"…estimates of jobs destroyed by trade sound big, but they're actually tiny. Relative to overall routine job destruction and creation — "job churn" — the number of American jobs destroyed by trade is minuscule.

In January alone, the number of American workers who were laid off or dismissed from their jobs was 1.8 million. The number of workers who quit their jobs that month was 3.3 million. Adding in workers who left their jobs for other reasons, such as retirement and disability, the number of job separations in January was 5.4 million. But there were 5.6 million hires in January, too. Those numbers are typical of most months.

Awareness of job churn should calm Americans' fears about imports [good luck with that–JB]…Compared with the number of total annual job losses…job losses from trade shrink into insignificance."

He then cites some estimates of trade-induced job losses:

"Ms. Wallach's estimate that Nafta destroyed one million jobs in its first 20 years means that it took freer trade with Mexico two decades to destroy as many American jobs as are now destroyed every 18 days on average. Mr. Autor, Mr. Dorn and Mr. Hanson's calculation that 2.4 million American jobs were ended by trade with China from 1999 through 2011 implies that the 13-year "China shock," as the paper called it, eliminated as many jobs as are eliminated, on average, every 41 days."

By this measure, almost any amount of job loss attributed to any cause will be insignificant. Boudreaux has taken Panglossian economics ("don't worry—be happy!") to a new high. His trick, if you didn't notice, is a) conflating gross with net flows, and b) not giving a crap about the pain of job loss, dislocation, and the damage done to whole communities that found themselves on the wrong side of these global dynamics.

I asked David Autor—he's one of the economists whose work Boudreaux critically cites—what he thought about this argument that job churn somehow negates job loss. His response follows:

"It's unfortunate that a Ph.D. economist would not recognize the crucial difference between gross and net job losses. By Boudreaux's logic, since "in a normal year, then, the number of workers laid off or dismissed averages 21 million," the U.S. Great Recession was a negligible event: the U.S. lost fewer than 4 million jobs in the first year (a mere one-quarter's worth of job losses) and no more than another 2 million in the second year (only a month's worth). It's remarkable that we even noticed!

Yes, when the U.S. loses and gains 21 million jobs in a year, this is the normal ebb and flow of the labor market. Large gross job flows need not imply any net loss of employment. But when sharp changes in world trading conditions cause the U.S. manufacturing sector to close up shop on 14 percent of its base employment level (2.4 million of 17.3 million manufacturing jobs) in the space of a few years, and many of these displaced workers leave the labor force, that's a huge rise in concentrated net job loss that is not part of the normal ebb and flow. (By the way, 2.4 million is the conservatively estimated trade-induced fall. U.S. manufacturing jobs plummeted from 17.3 million in 1999 to 13.8 million in 2007, a net reduction of 3.5 million, followed by another 1.9 million net fall between 2007 and 2010)."

So, if you happen to read Boudreaux's oped and it seemed inconceivable to you that millions of net job losses magically "shrink into insignificance," be assured that you were right and he was very wrong.

As I've tried to stress in much recent work, this moment does, at least it should, create a moment to talk about what we should do for those hurt by trade.

I've argued:

–Much better work supports for job losers, including direct job creation in places with persistently weak labor demand.

–Improve the quality of existing jobs through much better labor standards (see Heidi Shierholz's recent work on this). Though there are definitely pockets of weak labor demand, even today, broadly speaking, our labor market problem is less job quantity than quality.

–Help our smaller manufacturers by expanding the Manufacturing Extension Partnership (it's a small but venerable part of the solution—I've got a piece coming out soon on this with the details).

–Push back on currency intervention by trading partners with "countervailing currency purchases" (see Gagnon/Bergsten on this). Trump's new South Korean trade deal relegated currency rules to a toothless side agreement.

–See Lori Wallach and my agenda for more inclusive trade deals, including taking ISDS out of these agreements, putting a currency chapter in the deal with enforceable disciplines, and ensuring a much more balanced set of interests around the table when these deals are cast.

 -- via my feedly newsfeed