Canada’s largely unremarkable April employment report did reveal an eyebrow-raising development: a divergence between goods-sector and services-sector growth.
The country’s goods-producing industries generated 24,500 net jobs in the month, while the services-producing industries lost 12,000. Yet over the past 12 months, the divergence has been the reverse – service-sector employment has increased 1.6 per cent in the past year, but goods-sector employment has shrunk 1.6 per cent.
Both sets of numbers got me pondering whether a divergence between service-sector and goods-sector employment was particularly uncommon, and/or particularly meaningful. A couple of hours of digging through Statistics Canada’s database revealed the answers to those two questions to be “yes” and “perhaps/sometimes.”
First, some perspective. The service sector side accounts for roughly three-quarters of the total labour force. It’s also the much more stable side of the labour market; quarterly swings in service-sector employment are typically much smaller, on a percentage basis, than in the goods sector.
While month-to-month divergences between services and labour job trends can crop up reasonably frequently, it’s worth noting that in the past 120 months there have only been 20 instances when jobs in the goods sector grew while the services sector contracted, as they did in April.
If you look at average employment numbers per quarter – a better indicator of trend – this has only happened in one quarter in the past 15 years. That, it turns out, was in the third quarter of 2008 – pretty much ground zero of the financial crisis and Great Recession.
Slightly more common are quarterly divergences in the other direction, i.e., rising service-sector employment and contracting goods-sector employment. Still, this has only happened five times since the beginning of 1998. Three of those five instances (the fourth quarter of 2005, fourth quarter of 2007 and the fourth quarter of 2012) either preceded or coincided with substantial downturns in Canadian gross domestic product (GDP) growth. On the other hand, the two others didn’t portend any meaningful economic pivot point.
What does appear generally more meaningful, however, are sharp declines in the quarterly averages in goods-sector employment – regardless of what the services sector is doing at the same time. Every substantial GDP slump in the past decade (there have been seven of them) has been presaged by a significant slowdown in goods-producing jobs (which we’ll define as a drop of 4 per cent or more in the monthly average for the quarter). Only twice has a goods-sector job drop of that magnitude sent a false signal, i.e., the pace of economic growth did not slow meaningfully within a quarter or two.
On the one hand, it’s curious that the smaller and less predictable side of the labour market would be the one providing the more useful signal for economic turning points. However, the volatility in goods-sector employment may well be evidence of its heightened sensitivity to the ebb and flow of demand in the economy – thus putting it closer to the pulse of key underlying growth drivers.
Should any of this matter to us right now? Well, only if you take a look at the average monthly goods-sector job count in the 2013 first quarter. It was down 5.4 per cent – the second straight quarter topping the 4 per cent mark. The last time we had two consecutive goods-sector quarterly declines above 4 per cent? The 2008-2009 recession.