Sectoral shifts and the Canadian recession

Last December, I wrote a post on the effects of the recession on various sectors of the Canadian economy. The main results were:

• The service sector was barely affected; most of the losses were in the goods sector.
• Within the goods sector, manufacturing took the brunt of the job losses.

Of course, the losses in the manufacturing sector had to be put into the context of a broader trend of a shift away from manufacturing. I concluded with this:

Given these trends, it's hard to see how the manufacturing sector will be able to make up much more than a small fraction of the 200,000 jobs it lost during the recession. Perhaps all we can hope for is that employment there stays steady until the downward trend catches up to the losses generated by the recession.

That seems to be what happened. Between July 2009 and July 2010, the manufacturing sector added 2,400 jobs. You probably wouldn't call that much of a recovery, except for the fact that this was the strongest year-over-year number in more than five years.

The story of the recovery is below the fold.

Here is a graph that charts the changes in the Canadian labour market. Recall that the labour force is defined to be the sum of those who are employed and those who are unemployed (ie, not employed but seeking work):

This illustrates why we are still talking about the recession: even though we have almost recovered the jobs lost during the recession, employment growth has not been strong enough to absorb the flow of new entrants to the job market. The number of those unemployed is still some 300,000 more than it was when the downturn started.

As I mentioned earlier, the recession didn't affect all sectors equally. The services sector came through relatively unscathed:

The services sector has more than made up for the losses incurred during the recession. But since employment growth hasn't been strong enough to absorb the growth in the labour force, unemployment remains high.

Things are very different in the goods-producing sector:

Here we start talking about sectoral shifts. Employment in in the goods-producing sector is still more than 200,000 below the pre-recession peak, but unemployment has returned to what it was in October 2008. The only way to make sense of these numbers is to conclude that 200,000 workers have simply left this sector. Sadly, we don't know how many have left the labour force entirely or have simply moved to the services sector.

Let's break the goods-producing sector down:

Thanks to a combination of low interest rates and the federal stimulus program, construction employment has recovered almost all of its losses. On the other hand, although manufacturing employment has stopped falling, it hasn't recovered, either.

Let's look at unemployment in the goods-producing sector:

Now this is a surprising fact: of all the sectors we're looking at, the manufacturing sector is the only one to see a decrease in unemployment over the past 21 months. On the other hand, this result has been mainly achieved by former manufacturing workers leaving the labour force or moving to another sector:

So it would appear that not only has manufacturing employment not come back, former workers in that sector don't expect to find work in that sector, either.

I mentioned earlier the fall in manufacturing employment should be put in the context of a secular decline:

The dashed line is just a straight line I drew, but it seems to be enough to tell a story in which the effect of the recession was to accelerate manufacturing job losses that were likely to occur anyway. If employment in that sector holds steady for a few more months, it will revert to the (declining) trend. And it's hard to see why that general downward trend won't continue.

Some implications:

• If we thought that those manufacturing jobs were going to come back, we'd interpret those 200,000 lost jobs as unused productive capacity, and the Bank of Canada would view this as a signal for deflationary pressures. But if those jobs aren't coming back, estimates for capacity - and for idle capacity - will be revised downward. This suggests continuing on a path of increasing interest rates.
• Unless a fall in commodity prices brings about a sharp depreciation of the Canadian dollar - a development that would be unwelcome - the manufacturing sector is likely to continue its secular decline. But that doesn't mean that those who used to work there should be left to their own devices. Programs to provide training and other forms of support should be made a priority.

 

 

Comments

[Mike Moffatt]

Here's a data point I wish was available, though I doubt the data is available:

1. What percentage of manufacturing companies are primarily exporters?
2. What percentage of service companies are primarily exporters?

Given the timing of the secular decline, I believe Dutch Disease could be playing a significant role here.

[Mike Moffatt]

And interesting stuff, as always!

[Wendy]

Mike,

Good question. But I wonder if in today's global economy it is possible to separate "Dutch Disease" -- or the impact of a high petro-dollar currency--from the general phenomenon of lower cost developing economies like China, India, etc. being able to do the manufacturing work.

Even places like Britain that have seen a significant depreciation in their currency value are not seeing a manufacturing resurgiance as far as I know (not that I follow the UK economy very closely, you might, so feel free to correct).

[Nick Rowe]

I wonder how they define what sector an unemployed worker is not working in? I expect it's the sector he last worked in, rather than the sector in which he's looking for work?

[Stephen Gordon]

I'm pretty sure that's it. I looked up the LFS questionnaire, and it asks unemployed people where they used to work. It would be hard to ask what sector they're looking for, since someone with certain skills (ex: IT) could conceivably work in many different sectors.

[JP Koning]

Stephen, thanks for the update, interesting stuff.

One comment. Last time I checked (April?), the manufacturing sector has been the largest net contributor to Canada's improvement in GDP since the May 2009 (larger than services). Interesting that it would provide the largest drag on employment but provide the largest push to GDP.

[James K]

Thanks professor - very interesting.

One quick question: does employment underlying raw materials extraction & refining get picked up in the 'other goods-producing sectors' category?

[Stephen Gordon]

Yes - you can see the breakdown for levels (the graphs plot changes) from the July LFS here.

JP - I guess productivity must be the story there. But I don't know if it's the kind we like, namely, based on better capital equipment and technology.

[Kosta]

Stephen,

Once again, great stuff!

In your closing points, you suggest that a drop in commodity prices might lower the Canadian dollar and lead to a resurgence in Canadian manufacturing. There's a second possibility that could lead to a similar result, namely that the U.S. dollar could strengthen for some reason which would make Canadian manufactured exports to the U.S. much more competitive.

Although Canada is susceptible to the Dutch Disease, the dynamics are different given that 75+% of our trade goes to one nation. Not only our currency must be considered, but also the U.S. currency.

[Wendy]

Someone in Toronto recently told me that the GDP contribution of manufacturing to the Canadian economy (or was it Ontario economy) has not changed over the past 8+ years, despite a significant decline in jobs.

They explained this as being a consequence of the shifting nature of manufacturing in Canada to include "pharmaceuticals/drugs," technology components, processed foods, and other items that might have significant export values (besides autos and auto parts).

Any comments from the WCI gurus?

[Andrew F]

I've also heard that many of the jobs that have been lost were not particularly good, high-value added jobs like generic plastic or metal parts manufacture where firms essentially compete primarily on price.

[IgnaciousPlunder]

It'd be interesting to see what effects this might have on labour productivity in manufacturing. I would argue employment in the manufacturing sector in most countries has declined because of substitution for new technologies, you know, robots and such. But in Canada I'd say the same trend has to do with people shipping out west where opportunity is.

Since capital is easier to move than labour, it slipped away years ago leaving too many workers controlling too few robots, er, tools. Labour wont leave because it's a pain in the ass to take your kid out of school and ship across country, especially when you already have a decent job. That opportunity cost is reduced considerably when you get laid off. I guess the lesson to be learned here is that Canada has way too many Ontarians and not nearly enough Albertans.

[Jamie]

It's starting to get scary how many bankruptcies and consumer proposals we're administrating these days.. our debt settlements seem to be on the decline. Never a good sign.