A supposed software gap

I’m just a little taller than the Golden State Warriors’ MVP point guard, Steph Curry. But I weigh the same as he does, so I figure if I work out a lot and get my percentages of muscle and body fat to be the same as his, and if I start dressing the way he does and cutting my hair short like he does, then I’ll have a pretty good chance of making US$12 million a year like he does, too. (Shockingly, Curry ranks only fourth on the Warriors’ payroll, behind Kevin Durant, Klay Thompson and Draymond Green.)

OK, so that’s a crazy idea and not just because I’m a few decades older than Steph Curry. It’s crazy because surface similarities like height, weight, bodyfat indexes, hairstyle and so on don’t explain what makes a one-in-a-billion basketball shooter like Steph Curry. What does is pretty much a mystery, which is another way of saying: It’s so complex we’re never likely to understand it completely.

My basketball ambition came to mind as I read this week’s report from the Ottawa-based Centre for the Study of Living Standards on recent trends in Canadian ICT (information, communication and technology) investment and how they compare with what has been going on in the U.S. To make a long story short, Canadian ICT investment is lower than U.S. ICT investment and has been falling relative to it. In 2008, Canadian ICT spending per worker was 68.4 per cent what it was in the U.S., but by 2014 it was down to 56.3 per cent of the U.S. value (US$1,965 here versus US$3,493 there). The difference is not in computer hardware or telecom equipment, where we spend close to what the Americans do per worker, but mostly in software, which constituted three-quarters of the entire Canada-U.S. ICT gap.

Another well-known economic fact about Canada and the U.S. is that overall labour productivity is a lot lower here than there. We were at more than 90 per cent of their labour productivity in the mid-1980s but are down to just 75 per cent now. Over time, a productivity gap that big has dramatic effects on living standards. Last month, some of my relatives visited Stanford University, in the Bay Area, near where Steph Curry plays, and were blown away, as Canadians often are, by the manifest wealth. Some of us argue, rather like Cubans, that at least where we are, the wealth is more equally distributed. Well, maybe. And maybe that’s sufficient compensation. But maybe not.

Here’s the Steph Curry temptation. Our labour productivity is a lot lower than in the U.S. And our investment in software is also a lot lower in the U.S. Therefore, if we invest more in software, like they do, we’ll get higher labour productivity, like they have.

The correlation is not quite spurious. Spurious correlation would be: U.S. politics is much more ridiculous than ours and U.S. incomes are higher, too, so if we get more ridiculous politics (which might be on its way) we’ll also get higher incomes. There’s no plausible causal connection there (unless higher incomes make your economy more resilient to ridiculous politics). By contrast, we feel almost instinctively that there must be a causal connection between software investments and economic growth.

It’s certainly a plausible connection. But it’s by no means certain. The Nobel Prize-winning MIT economist Robert Solow once remarked, “You can see the computer age everywhere but in the productivity statistics.” It’s suspicion-making that the sector in which we most outspend the U.S. in ICT investment per worker — by 22 per cent — is public administration and defence. Do we really have more efficient public administration and defence than they do? Or is buying ICT simply something that’s more fun to do when you’re spending other people’s money?

The CSLS report doesn’t fall prey to the Steph Curry temptation. Instead it ends with a number of good observations about how ICT might be measured better. But it was written for the federal Department of Innovation, Science and Economic Development Canada and you know the folks over there are just itching to plumb the policy implications of our twin gaps: productivity and software. If we crank up software subsidies, they’ll be thinking, we can raise our software per worker, which will increase worker productivity, which can lead to wage increases and higher federal revenues and all sorts of other good things. Maybe even an NBA franchise in Ottawa.

Instead, the department should start asking itself hard questions about what it is in the Canadian tax, legal, trade, regulatory and policy environments that causes presumably profit-seeking Canadian firms to conclude that American-level investments in technology in general and software in particular won’t help their bottom lines.