The problem with uneven economic growth

The headline statistics from Statistics Canada’s first release of income data from the 2016 census were moderately good news: median real incomes rose by just more than 10 per cent between 2005 and 2015. (The actual number depends on which income is measured — before or after taxes — and whether you look at individuals or households.) This may be even better than good news, since 2005 to 2015 excludes the first few years of the oil boom in the 2000s, and includes the sharp recession in 2008 and 2009.

Of course, as many were quick to point out, the statistics do not reveal uniformly good news. Incomes grew more slowly in Ontario than in other provinces, and according to some measures, median incomes in cities like London and Windsor actually declined during this decade.

Prime Minister Justin Trudeau often speaks about the importance of the benefits of economic growth being broadly shared: social cohesion is harder to maintain if pockets of society are left in poverty while the rest prospers. But it’s a mistake to start from that unobjectionable point and conclude that economic growth rates should always be uniform.

Relative changes shouldn’t be examined in isolation; starting points matter as well. Indeed, the University of Rochester economist Steven Landsburg calls this the “Grandfather Fallacy.” By looking only at variation in changes across subgroups, any iniquities in the original distribution are “grandfathered” into the discussion. An analysis that looks at dispersion in growth rates and concludes that inequalities are being created, or that some groups are falling behind, is implicitly assuming that the original situation was worth preserving.

Of course, this is not always the case. For example, median inflationadjusted men’s employment income increased by only 2.2 per cent between 2005 and 2015, while that of women increased by 11.6 per cent. It would be a mistake to conclude from this data that men are being “left behind” in the labour market. When you look at the initial starting points, you’d notice that median women’s employment income was 34 per cent lower than men’s in 2005, and that those higher growth rates in women’s income only managed to close the gap to 29 per cent in 2015. The only way that women’s earnings will converge with men’s is if women’s earnings grow faster — that is, if more of the benefits of economic growth go to women.

What about the provinces? The standard model of economic growth predicts that income levels in different economies will converge. Again, this can only happen if there’s dispersion in growth rates: if every region or country has the same growth rate, those with low incomes will never catch up.

The standard model doesn’t always fit the data, of course: there are many examples of low-income economies that have fallen farther behind, and of high-income economies that have pulled farther away. Economic growth is not predestined; it requires a minimal set of laws and institutions. But once you take that into account, it appears that income levels in countries that are similarly governed do tend to converge. So for a country like Canada, we’d expect to see income differences across provinces gradually disappear. (The fact that they’ve persisted for so long is a bit of a puzzle. Perhaps those barriers to interprovincial trade are even more important than we thought.)

Which brings us back to Ontario. Median real incomes in Ontario rose by 3.8 per cent (using total income for individuals), compared to 12.7 per cent for Canada as a whole. But in 2005, Ontario’s incomes were 6 per cent higher than for Canada as a whole. And when you look at the census metropolitan areas (CMA), 14 of the 17 Ontario CMAs had median incomes above the national number. Everything else being equal, you’d expect that a high-income economy like Ontario’s to have lower growth rates.

The good news for the convergence story is that all seven of the traditional “have-not” provinces — Atlantic Canada, Quebec, Manitoba and Saskatchewan — saw their median incomes increase more than the 12.7 per cent increase for Canada as a whole. Indeed, median incomes in Saskatchewan are now above the national number and are only behind Alberta among the provinces. (I’m taking care not to say the “national average,” because these are medians: the national median is not a weighted average of the regional medians.)

I’ll pass quickly over B.C., because median incomes and growth rates are reasonably close to the national numbers: convergence has already happened.

Finally, there’s Alberta, which confounds the narrative. It already started with the highest incomes in 2005, and its provincial median incomes have grown twice as fast as the Canadian median since then.

If it were up to policymakers in 2005 to choose where to locate economic growth over the next 10 years, they would have viewed the results as being mixed. Lower-income provinces had faster growth, and they were able to make some progress in closing the gap with Ontario and B.C.

Instead, the “problem” — if we can call it that — is that Alberta continued to pull away from the rest of Canada, at least as far as incomes go. And we probably should call it a problem: the widening gap between Alberta and the rest of the country has sharpened traditional grievances in Alberta about equalization.

The issue with uneven economic growth over the past decade isn’t that it left poorer provinces further below the Canadian median; they managed to catch up somewhat. The issue is that Alberta pulled even farther ahead. This is good news for Albertans, but could pose headaches for policymakers.