LOOK FOR THE UNION LEVEL. WATSON,

How much, if at all, does declining union membership explain rising inequality in North America? And how much, if we started favouring unions more in our labour-market legislation and regulation, would inequality fall? “Lots” and “lots” are the answers most anti-inequality activists give.

By contrast, “Probably some, but it’s not so easy to say” is more like the answer emerging from a new study by three blue-chip Canadian labour economists: David Card of the University of California at Berkeley and Craig Riddell and Thomas Lemieux of the Vancouver School of Economics at the University of British Columbia. Apart from its delightful characterization of Nobel Prize-winning economist Joseph Stiglitz as a “pundit,” their paper is noteworthy for emphasizing how unions’ effects depend on whether you’re talking public sector or private.

If you subscribe to hoary union lore, as celebrated most often in this country on the CBC, the union wage effect is all good and all equalizing. Unions raise up the downtrodden and give them fair value for what they add to output. According to the folk songs, the effects are only beneficial, which, if that were true, would make unions unique among human institutions. But it’s actually more complicated than that.

Suppose you organize teachers and they do manage to raise their wages. Does that produce more or less wage inequality in society? It depends. If teachers are already higher up the wage scale, their unionization may well increase inequality, which is not what its proponents (including pundits like Stiglitz) project. In fact, some studies suggest it does raise inequality among women, since higher-income women working in education and medicine unionize the most.

On the other hand (this is economics, there’s always another hand) what if the teachers’ union ends up compressing wages within teaching? What if, as many unions seem to want to do, it ends up levelling: raising the wages of those lower down the teachers’ wage ladder and, not actually reducing, but slowing the relative growth of those higher up? That reduces wage inequality within teaching.

So the overall effect depends on who’s unionizing or, more likely in recent years, who’s de-unionizing and what effect unionizing or de-unionizing has on the occupations that experience one or the other fate.

In North America, unions are now mainly a public sector affair. In fact, although the rate of unionization in the U.S. has fallen in the private sector, from 24.2 per cent in 1973 to 6.5 per cent in 2017, the rate in the public sector is up, from 23.0 per cent in 1973 to 34.4 per cent in 2015. In Canada, the story is essentially the same: private sector unionization has been falling for decades and at 2017 was just 16.4 per cent, but public sector unionization in 2017 was 75.5 per cent. This switchover from private to public sector has effectively feminized unionism. In the U.S., 47 per cent of union members are women; in Canada, it’s 53 per cent.

What about the union effect on inequality? It has diminished, although it wasn’t that big to begin with in the private sector, the three labour economists find. In the U.S., unions now reduce wage inequality by just 3.5 per cent for the economy as a whole, and in Canada just 7.9 per cent, which means that in the U.S., more than 95 per cent of the inequality story, and here, more than 90 per cent, is from other stuff — technology mainly and to a lesser extent trade.

In the two countries’ public sectors, however, unions’ inequality-reducing effects are big: 16.2 per cent for male public-sector inequality in the U.S. and 48.5 per cent in Canada. And 10.7 per cent for female public-sector inequality in the U.S. and a full 50 per cent in Canada.

These big effects in the two countries’ public sectors raise two questions: One: Is public sector pay really a big part of the inequality story? Only one in six Americans and one in five Canadians work in the public sector. And intersectoral comparisons suggest public sector workers aren’t lagging wage-wise. Quite the opposite. If anyone’s hurting in this society, it’s not them.

Beyond that, the other question to wonder about is whether union levelling is a good thing. Part of it may even be spurious. When a university professor becomes a dean, she gets a wage bump but also has to quit the union since she’s joining “the bosses.” The resulting correlation? Union membership seems to depress the wages of highly skilled workers. But there may also be a non-spurious effect. It’s not uncommon for star professors in one-salary-fits-all unionized shops to escape to the few non-unionized universities that remain in Canada and are able to recognize their abilities with higher salaries. Which constitutes both economic and social justice even if it does raise inequality.

IS PUBLIC SECTOR PAY REALLY A BIG PART OF THE INEQUALITY STORY?