A SLOWER DEATH FOR FACTORY JOBS.

Question for Donald Trump: Over the last four decades, which rich country has lost manufacturing jobs most quickly?

OK, so the answer to the question is a fact and the president seems to believe he’s entitled to his own facts. But if you respect real facts and look them up, the answer is “not the U. S.” In fact, from 1973 to 2010, Germany’s manufacturing employment fell as a share of its total employment by 0.45 percentage points per year on average, compared to a decline of only 0.40 percentage points per year in the U. S. True, Germany had a bigger share of manufacturing employment to begin with — just over a third of its total employment versus just under a quarter in the U. S. — but Germany wasn’t as rich as the U. S. then, so it hadn’t got as far in the switch to services, which is what all countries go through as they get richer. Even China will.

The biggest manufacturing employment share loser from 1973 to 2010 was the U.K., at 0.60 percentage points of total employment per year. Canada had a slower rate of decline: 0.32 percentage points per year, which took manufacturing from 22.0 per cent of our total employment in 1973 to 10.27 per cent in 2010.

All these numbers are from a new paper by Robert Z. Lawrence of Harvard’s Kennedy School of Government. The paper is called Recent U. S. Manufacturing Employment: The Exception that Proves the Rule.

The rule Lawrence refers to is that for most of recent history, manufacturing’s employment share has been governed by two things mainly: faster productivity growth in manufacturing than in most other parts of rich- country economies, plus people’s relatively “inelastic” demand for goods as they get richer. As Lawrence puts it, exaggerating just a little, “Poor people spend more of their money on goods. They tend to buy more food, clothing, and motorcycles. Rich people, by contrast, spend more of their money on services. They eat in restaurants, have personal trainers, hire accountants and lawyers, and see psychiatrists.”

These two forces create a double whammy for manufacturing employment. As societies grow richer, they switch their consumption away from goods toward services. But as productivity grows in manufacturing, it takes fewer workers to produce the goods consumers do want. Slower growth of demand combined with more rapid growth of productivity equals a declining share of total employment.

Of course, the rise of a big manufactures exporter like China can affect all this. Standard estimates are that goods imports from China displaced a little under a million American workers in the first two decades of this century. Remember, though, that total non- farm employment in the U. S. is 148 million. So in the big picture China’s influence is essentially marginal.

U. S. manufacturing employment, like manufacturing employment in Germany, the U. K., Canada and all other rich countries, would have declined over the last 40 years even if China had continued on as the geopolitical equivalent of the far side of the moon. As Lawrence shows, in the 1960s and 1970s, when a turned- inward China was last under the thumb of a chairman- for- life, manufacturing’s share of U. S. employment was declining at exactly the same rate as it has done since 1979, when China started facing up to economic reality and the world economy.

So what’s the exception in Lawrence’s title, “The exception that proves the rule”? It’s the last few years. In the countries he looks at, the decline in manufacturing’s share of employment has slowed. Where we lost 0.32 percentage points a year from 1973 to 2010, from 2010 to 2016 we lost just 0.15 percentage points per year. Same thing in Germany, with the decline in share falling from 0.45 percentage points to 0.16. In the U. S., manufacturing’s share of employment actually rose very slightly between 2010 and 2016. So Donald Trump comes to the rescue just as he’s no longer needed.

Politicians will be tempted, Canute- like, to take credit for the current reversal of trend: See! The pro- manufacturing, pro- tech policies we’ve all instituted are working! Lawrence, focused resolutely on the fundamentals, is skeptical. What’s going on, he says, is a slowdown since 2010 in the switch to services, as well as — despite all the foofaraw about robots and AI — a slowdown in productivity growth in manufacturing.

If people aren’t switching away from goods quite as much as they were and if the number of workers required to produce a given output isn’t falling quite as much as it was, then manufacturing’s share of employment won’t decline as much as it would have and that won’t have anything to do with Trump or tariffs or trade.

But the productivity slowdown part of it won’t be good for any of us in the long run.

MANUFACTURING’S SHARE OF EMPLOYMENT ACTUALLY ROSE VERY SLIGHTLY. TRUMP COMES TO THE RESCUE JUST AS HE’S NO LONGER NEEDED.