DO THE UNSKILLED DRIVE TRADE POLICY?

For economic liberals, it’s heartening to see widespread public support for free trade. An Abacus poll last month reported that 80 per cent of Canadians oppose Donald Trump’s tariffs on steel and aluminum. Over 70 per cent also back this country’s retaliatory tariffs, albeit probably with little knowledge yet of how much they’ll hurt.

Strong support for our own tariffs doesn’t necessarily indicate support for trade, of course. Some of the 70 per cent who like our retaliation have probably always favoured higher Canadian tariffs. But, given the current context, polls such as this are effectively a referendum on whether people want freer trade or a more protectionist Trump-world. For anyone who remembers the political civil war this country endured from 1985 through 1988 over Canada-u.s. free trade, to see such solid support for the trade status quo is gratifying.

But isn’t it a given these days that support for trade and globalization has declined in the last couple of decades? Isn’t the Trump presidency itself a result of growing hostility to imports, immigrants, open borders, Chinese competition — you name it — especially on the part of what Ken Dryden once called “the moved and the shaken,” i.e., the victims of globalist movers and shakers?

Maybe that’s true in Ohio, Pennsylvania and two or three other states whose Electoral College votes the whole world now has to fret about. But an impressive new study from the European Bank for Reconstruction and Development suggests it’s not generally true. We may have to rethink our assumptions about trade politics.

(Given how many hundreds of millions of people around the world stand to be hurt by trade policies invoked in the name of 500,000 or so swing voters in the U.S. rust belt, maybe the WTO could organize a Gofundme campaign so the one- or two- billion victims of Trump policies worldwide could pay the voters in question to knock it off.)

The study — by Cevat Aksoy of the EBRD, Sergei Guriev at Sciences-po in Paris, and Daniel Treisman at UCLA — is most impressive for its scale. It looks at Gallup surveys of 450,000 people in 118 countries over 10 years (2005–15) to see how their feelings about their political leaders and governments varied according to the ups and downs of their countries’ trade. If imports and exports rose or fell, did that affect governments’ and prime ministers’ and presidents’ approval ratings?

It turns out trade does have an effect, but not in quite the way most people assume these days. The rawest correlations indicate that, across the 118 countries, a rise in imports actually increased people’s approval of their governments and leaders, and that was true among both skilled and unskilled workers. (Reflecting both data necessity and possibly also the bias of the academics who do these kinds of studies, the division between skilled and unskilled labour is whether a person has or doesn’t have any post-secondary education. Personally, I’ve got as much post-secondary education as you can get — a PHD — but as for skills, I can barely fix a flat tire.)

Imports are correlated with lots of things, so it could be, for instance, that the rise in national income that usually sucks more imports into a growing economy is really the reason why people’s opinions of their government rise at the same time imports do (which isn’t to say people don’t also like the imports; they wouldn’t buy them if they didn’t.) When you do control for national income and several other things, the positive impact of imports on approval ratings loses its oomph and is no longer statistically different from zero. On the other hand, it doesn’t go negative. Even when the pure import impact is isolated, governments generally don’t get punished for higher imports. Hear that, Mr. Trump?

How about the pattern of trade? Economic theory says that if you’re skilled, you may not appreciate imports with high skill content since they’re competitive with what you yourself produce. With all the appropriate statistical controls in place, the data bear this out: “A 10 per cent increase in highskill-intensive imports results in a 0.53 percentage point fall in approval of the leader among skilled individuals (relative to unskilled ones).” By the same token, the popularity gain from a 10-per-cent rise in exports of such goods is 0.46 percentage points. For many political races, of course, a half point can be decisive.

In sum, around the world the movers and shakers of political opinion and therefore presumably of elections, too, tend to be high-skilled workers, not the low-skilled workers that have received so much attention lately. And these movers and shakers reward governments and leaders when things go their way. If you’re comfortable with the old Smithian notion that people generally act in their own interest, that’s greatly reassuring.

WE MAY HAVE TO RETHINK OUR ASSUMPTIONS.