The Decline of Manufacturing in Canada - 1926-2011: Dutch Disease?

The debate about “Dutch Disease” is focused on the relationship between natural resource export booms, currency appreciation and the decline of Canadian manufacturing.  I decided it was worth hunting up some long-term data on manufacturing’s share of Canada's economy given that my economic history background tells me that over the long-term, the share of the economy in goods production has declined while that of services has risen. 

 

It turns out it is not a simple task to calculate a consistent long-term manufacturing to GDP ratio for Canada’s economy.  The best I was able to come up on short notice was three separate series with some overlap. The first series is from Historical Statistics of Canada (F56-75) for GDP at factor cost by industry.  The second series was GDP by industry at factor cost (durable and non-durable manufacturing) for the period 1961 to 1997 (v334558, v334560, v334561).  The third was a similar series but in constant dollars for the period 1981 to 2011 (v41881478, v41881488, v41881489).  What I want is the manufacturing share of GDP and I have plotted that in Figure 1 for the three series. 

 

What Figure 1 illustrates is that there has been a long-term decline in manufacturing’s share of Canadian GDP.  The M/GDP ratio rose from 1926 to the WWII period and then began a long-term decline from just over 25 percent of GDP.  The period from 1980 to 2000 sees a halt of that long-term trend but since 2000 the decline appears to have resumed and at a very steep pace. Figure 2 illustrates these trends a bit more clearly. I have combined the three series into one by taking the average during periods of overlap.

 

I think what has happened is that the period from 1980 to 2000 saw a stabilization of manufacturing’s share of GDP at about 17 percent.  This stabilization came after several decades of postwar decline in the ratio but has become the “new benchmark”.   This stabilization took place during a period where our currency depreciated substantially against the U.S. dollar.  From the end of the Second World War to the early 1970s, the value of the Canadian dollar relative to the US dollar was close to par.  The period from the 1970s to 2001 saw a period of depreciation but since 2001 the currency began to appreciate relative to the US dollar and is now where it was for much of the period 1945-1975.  The most recent plunge in the manufacturing to GDP ratio is associated with that most recent appreciation with much of that manufacturing loss in Ontario.

 

Is this Dutch Disease?  If it is it has been a very sustained and chronic long-term disease.  Manufacturing as a share of GDP in Canada has been in decline since the end of the Second World War from a peak of almost 30 percent that was due largely to wartime industrial policies.  Until the Second World War, the manufacturing to GDP ratio ranged from 20 to 25 percent and had actually been in that range since the 1870s based on the Green-Urquhart numbers.  Despite this long-term decline, there have been short-term reversals fueled in part by currency depreciation – the most sustained being that from 1980 to 2000.   Relative to the benchmark of the period from 1980 to 2000, the current decline in manufacturing is understandably a cause for concern. However, viewed over a longer time span, it is part of a long-term trend dating back over half a century much like the decline of agriculture as a share of both employment and output as we moved from the nineteenth and into the twentieth century.

 

Comments

[Stephen Gordon]

Nicely done. I was wondering what the longer-term perspective was.

[Mike moffatt]

Great stuff! Keep in mind this is output, not employment. With above average productivity gains in mfg of late, employment has fallen even faster than output.

[Tuzanor]

What about the fact that since 1926 there have been many new industries? Wouldn't this crowd out manufacturing as a percentage of GDP? Also the government gobbled up more and more of overall GDP between 1950-1980, could this not effect the stats as manufacturing as a percentage of GDP, especially as government stopped "growing" after that period?

[Simon van Norden]

"Is this Dutch Disease? If it is it has been a very sustained and chronic long-term disease."

Um....who precisely do you think is claiming that Dutch Disease is the only long-term influence on manufacturing's relative importance?

For that matter, just who is claiming that Dutch Disease is important on time-scales of over half a century?

"I think what has happened is that the period from 1980 to 2000 saw a stabilization of manufacturing’s share of GDP at about 17 percent."

Now, stop me if I'm wrong, but wasn't the global recession of '82 followed by a sharp and persistent drop in non-energy commodity prices? And didn't we see similarly large and persistent falls in energy prices starting around 1985? (Think Iran-Iraq war.) In the face of such events, the Dutch disease model predicts a boost to manufacturing....and we see a halt in its 40-yr trend decline...as the Canadian dollar declined steadily.

Two caveats to note:
1) FTA (and then NAFTA) was signed in the late 80s, which could have influenced manufacturing trends.
2) The Bank of Canada's tight monetary policy during part of this period probably made the CAD quite a bit stronger than usual.

[Normand]

Two questions.

How is this ratio compares with other countries?

How is the manufacturing output calculated? Exemple: A car is entirely built in Canada vs a car that is assembled in Canada.

Thank you!

[duncan cameron]

Note that the high point for manufacturing was industrial policies associated with WW II, when Canada adopted the CCF model of price controls and planned investment, and attained full employment. Obviously a special situation prevailed and cultural inhibitions to government intervention were attenuated.

[Erin Weir]

It’s certainly true that there has been a slow, long-term trend of relative decline in manufacturing, but that does not explain much of the sharp collapse during the past decade.

[Chris J]

A question: let's put aside the manufacturing buzzword: Is it significantly harder to start a business in Alberta now than it was 10 years ago because land prices/rent, labour costs and such are expensive?

And when the oil boom ends will the resulting bust be large enough that the hardship in future outweighs the benefit now?

A guy from a nickel town....

[The Keystone Garter]

I googled Que had 1/3 of employment in 1960s from manufacturing and 1/5 later. Couldn't find for Ont; tough to Separate other factors.
The branch plant economy was a deadend. No R+D in Canada. Obviously petro doesn't use R+D. I'm not sure if Irving winning shipping building contracting encourages diversification of petro or reinforces such a low R+D industry (building new ships is R+D).
Really need to index corporate tax increases and decreases to R+D. Our cars, petro and banks are all low R+D. R+D is new products; might not be manufacturing but at least a chance.
Mulroney took bags of cash to have helicopters made in Germany. TransCanada is somehow mobbed up is today's news; I thought they'd pipeline to Thunder Bay, Chicago or N.O....
I know sourcing contracts to chosen actors is inefficient from a trade perspective, but why don't we just chose a non-corrupt gov, a non-corrupt energy future, and subsidize a cdn manufacturing base: wind turbines in AB/SK and rails for instance?

[The Keystone Garter]

...Finance is 40% of profits. Petro now prolly same. These both tend to rise when CAD ^. Cars were around there pre-2003. Now if the economy is booming, health and education are prolly already too via tax base rises.
Before 2003, cars were a hedge against finance (not so much as insurance bigger back then). And it was solely USA-Can interaction. Makes the public sector dynamic tricky.
Now, the manufacturing hedge against finance, petro, education, health, is very uneven. If the dollar falls, we get cranked everywhere. Either need a new hedge or surpluses (savings, like a Trust by B of C or like provincial Heritage Trust). If the dollar falls we will all be unemployed, and there isn't a surplus now like there should be. IDK about Dutch disease, but I know we need investment in new sectors while resource revenues good.

[Brian]

Duncan Cameron's post illustrates what's wrong with the economic conversion in this country: partisan people only hear the arguments they want to hear, they only see the evidence they want to see, and anything else - consequences, facts, explanations, context - gets tuned out.

First, it wasn't the "CCF's model" of price controls and planned investment; it was a wartime model, much of it developed in the First World War when the CCF didn't even exist. Wartime production often mobilized the management skills, capital and creativity of private firms which the CCF would have cheerfully nationalized in their plans for a socialist utopia. The CCF's model was designed for peacetime, without, for example, the rationing of basic consumer goods. And I can't imagine the CCF would have been too cheerful about the wartime model if it were tried in peacetime, given the fact that high outputs were achieved partly because 'cultural inhibitions' against draconian labor practices were more easily suppressed in wartime. For example, you mention price controls, but cleverly (or ignorantly?) left out the part about wage controls. So much for the sanctity of collective bargaining!

Second, full employment was achieved in part because hundreds of thousands of Canadian workers in their prime were out of the industrial workforce, employed by the taxpayer in places like Dieppe and the Aleutians, Caen and Ortona, Hong Kong and the Rhineland. Some of those at home and a tiny few sent abroad were conscripted. How on earth anyone could refer to 'full employment' during the period without noting that mildly important distinction is beyond me.

Third, it wasn't a matter of 'cultural inhibitions,' but fiscal inhibitions. This level of output was achieved partly by borrowing every surplus dollar possible from workers, and taxing them on top of that. We were in a position to handle those debts after the war because Europe - with US aid - put itself in hock to buy goods and resources from our undamaged country after the war, not as a result of any particular act of Canadian genius.

So, let's quickly put the "Second World War proves the NDP is right!" idea into the bin and keep it there, forever.

[The Keystone Garter]

...insurance profits rise with booms; higher interest rates give them arbitrage to make profits. Used to be a hedge within the banks. Think are good investment as China been growing at very high 8% instead of 6%....

I pictured a teeter-totter with workers running to oil, banks in high dollar, and factories along Windsor to Mtl, insurance, during low dollar....with surpluses in booms the nursing and education should maybe be maintained in recession (contradicting myself), and are thus counter-cyclical. Without the boom surpluses, nursing and teachers will be cut in recessions, insurance has shrunk, the cars are in deep South...I guess we are supposed to go to USA or China or something if low CAD with hollowed manufacturing...

[The Keystone Garter]

...running red ink no matter what will always cut education and thus new industries. That's what I'm tripping over. Will we get in the black in recessions or when oil is profitable and powerful?

[Ginna]

I would note that your timeline quite neatly matches the rise of Japan, and then China as manufacturing powerhouses. Of particular note is the explosion of cheap goods from China starting in the late 1990's. You could make an argument that Canada's competitiveness fell during this period, and I would not disagree, but it is historical reality.

You could also make the argument that the artificially low $CND allowed domestic firms to stagnate, and not increase productivity in the 80's and 90's.

[Neil]

@The Keystone Garter
Obviously petro doesn't use R+D.

I keep seeing people make statements like this. How is this obvious at all? The energy industry and related petrochemical industries do invest quite a bit of money in R&D, trying to squeeze more out of each site. Indeed, there's an entire industry of secondary extractors that buy up spent fields cheap and try out new technologies to squeeze out more production. Per Stephen's ""Does Canada have Dutch disease?" is a question without a meaningful answer" R&D in the resource industry has tripled over the last decade, and is now bigger than pharmaceutical research.

[The Keystone Garter]

...it isn't obvious. Reading income statements and having a free UK government publication that broke down the sectors and biggest companies, is how I got my expertise.
http://en.wikipedia.org/wiki/Research_and_development
Avg industrial R+D in USA is 3.5% of revenues.
Merck and Novartis are 15%. You picked the worst sector example due to limited duration patents more than anything.
If you google finance the one yr income statement of husky (one of the big-3 here), you get revenue of $25B and R+D of just under 2%. Given our low tax-rates, it would be nice if they doubled this. And paid for pollution like everyone else does. Yep, if you develop a new straw, we can build them here maybe.

[The Keystone Garter]

http://www.bis.gov.uk/foresight/our-work/projects/current-projects/future-of-manufacturing
...ready for 2013 on topic. Mailed right to you in 2007.

[Jacques René Giguère]

That manufacturing has been returning to from its unnaturally high share due to the two unpleasantnesses down to its natural level doesn't preclude Dutch Diseasa from happening. A slowly leaking ship can get torpedoed...

[Bob Smith]

And sometime leaking ships continue to leak...

[Determinant]

Third, it wasn't a matter of 'cultural inhibitions,' but fiscal inhibitions. This level of output was achieved partly by borrowing every surplus dollar possible from workers, and taxing them on top of that. We were in a position to handle those debts after the war because Europe - with US aid - put itself in hock to buy goods and resources from our undamaged country after the war, not as a result of any particular act of Canadian genius.

So, let's quickly put the "Second World War proves the NDP is right!" idea into the bin and keep it there, forever.

Neither Canada nor the US "paid off" our war debts. The absolute nominal value of the debt did not decrease, revenue growth and inflation reduced the payable amount as a share of our national budgets to a trivial amount. And no, it wasn't European consumption that helped most of all, it was our own growing domestic consumption.

Second, it is very, very clear from materials circulating that the time of WWII that people realized that we don't have to suffer depressions if we don't want to. The esteemed Dr. Krugman has written an entire book on this to reiterate this point.

Really, the government flicked a switch, sucked up all the spare capacity in the economy and still wanted more. Private debt burdens withered away.

Of course in 1945 everyone was scared that the the economy would fall back into depression, just as happened after the Napoleonic Wars, the US Civil War (Long Depression) and WWI (Great Depression). Companies like Xerox traded at 3 times earnings in 1948, you could have had the best of Corporate America, companies with decades ahead of them, for a song. Wall Street and Bay Street were that out of it.

The surprise was that the next Depression didn't happen (then) and we went onto have "Trente Glorieuses" or the "Long Boom" and decades of middle-class bliss. It all rested on the fact that Governments acknowledged that government spending could rescue an economy, private debts were kept low and the middle class would be reinforced by a policy choice of full employment, pensions, benefits (the shadow welfare state, epitomized by GM) and would thus be free to consume in a predictable way. Businesses prospered from this new stability. It was a virtuous reinvestment cycle.

And then we systematically attacked and dismantled every one of those tenets. A great deal of NDP rhetoric is actually about trying to bring the successful policy mix of the Great Compression (1940-1970) back.

Krugman and Jacob Hacker's The Great Risk Shift detail all of this quite well.

[duncan cameron]

Much of wartime finance relied on selling bonds to the Bank of Canada, quantitative easing in todays language. The Bank did not exist in World War I of course, but it was in the Regina Manifesto.

Because we have assigned ourselves a junior role as U.S. ally in defence, planned investment goes on today in military procurement (and seems to be done badly leading to scandals and massive over-spending) but is considered not on by Canadian mainstream economists for the private sector because of the inability to pick winners etc. I am not sure how well the FTA did in assigning us a role as bitumen exporter to the U.S. Things look to boom now, but the staples trap looms as well. It has seemed to me intelligent industrial planning would have worked better for employment, wages, and equality than the open door to foreign investment and the bitumen subsidies allowed under the FTA.

Sorry if I offended you Brian. That was me being mild by the way. How will you react when I get strident I wonder.

As part of the Canadian economic conversation you might look at the 1943 book Frank Scott and David Lewis did called Make this Your Canada where the CCF model is set out. Now neither were economists so that may rule them out as interlocutors but the model they lay out does include key elements used in wartime including the wartime prices and control board and planned investment.

[Simon van Norden]

"The debate about “Dutch Disease” is focused on the relationship between natural resource export booms, currency appreciation and the decline of Canadian manufacturing. I decided it was worth hunting up some long-term data on manufacturing’s share of Canada's economy given that my economic history background tells me that over the long-term, the share of the economy in goods production has declined while that of services has risen."

I guess this intro (and your title) really threw me off. I thought this post had something to do with Dutch Disease. But am I right in understanding that the two sentences in this paragraph have pretty much nothing to do with each other? As our exchange in the comments on your next make clearer, neither of us think that these long-term trends have anything to do with Dutch Disease, right?

[Simon van Norden]

Ooops.....on your next post....

[Livio Di Matteo]

I suppose the first sentence would have been better written if I had said: "The debate about “Dutch Disease” is focused on the short-term relationship between natural resource export booms, currency appreciation and the decline of Canadian manufacturing..." The post is about Dutch disease in the sense that the debate over Dutch disease inspired it. However, I am pointing out the long term decline in manufacturing's share - which is due to a wide range of factors (eg. productivity improvements in the sector) and put the "Dutch disease" rhetorical flourish in the title. I'm trying to have a little fun as well as contribute to debate.

[The Keystone Garter]

To me the Dutch's problem was they focused upon two-blade wind turbine models. The later coming Danes rode the successful three-blade variety. If we know demand for carbon-intensive industries, regardless of the workforce or industrial sector classifications, is limited, we should use our resources to plan (yes I said the forbidden word) an industry beyond petro. No, once you go to petro at $33/hr, you can't go back to manufacturing cars again. Maybe economists should be questioning the writers of their text-books about their models; the scientific brain-pans of whoever wrote the utility function?
Coyne was against stimulus until the crowd swayed him. Haprer was against the stimulus until the crowd swayed him. Don't be popular. Be correct. Be efficient.

[The Keystone Garter]

...in 1999, looked at stock option prices, and thought: I'll buy out-of-the-money LEAPs, using value investing. Wasn't allowed to: didn't know what Black-Scholes was; had to believe popular false (as economics is 1/2 politically pro-rich and only 1/2 a science) economic theories. Losing manufacturing was inevitable and very predictable. We subsidized (at the expense of future doctors and low interest rates) big car plants in Ontario. They were most efficient measured by production. But based on $8/barrel oil. Based on USA consumer demand. Under GWB, the USA middle class was wiped out (utility function ignores Gini and consumer demand thus ignores inflation thus ignores our future Greek stagflation). Under W, USA debt rose....
Manufacturing would've lasted longer if teeny cars. Producing oil diminishes industrial freshwater in many nations, and future global harvests. It kills meat for instance. It kills rover's grand-puppies. U r killing your child's future dog. Dog murderers!!

We got lucky resource prices rose: China went from 3% of global food production in 1985-ish to 10% in 1995-ish. Only reason can argue about what to do/advise with petro-revenue (AB suburban sprawl currently). Luck. Good luck when China drops 2/3s of harvests *and* freshwater.

[henryc]

It would be surprising indeed if Cdn mfg had grown as a share of GDP over the period in question. Worldwide, it's a declining share, for reasons already mentioned above. But some countries (Germany, Japan, for instance) have managed to retain a disproportionate share. Neither enjoyed a resources windfall income. My question is, amongst the countries that have enjoyed such windfalls (if any) which ones have maximized their mfg output, and how, and why? I think that's an interesting research question.

[Ivars Sayfy]

See the following link for a manufacturing to GDP ratio from 1970-2010 for the G7 and others

http://opinion.financialpost.com/2012/05/29/everybodys-dutch/

The "Dutch disease" is not just about the high dollar or mfg. to GDP ratio but how the tax income and royalties earned from the resource extraction are invested (eg. Heritage fund) by the different levels of Government.

The Norway solution for their North Sea oil seemed to avoid many of the problems that the Dutch encountered decades before. A critical requisite seems to involve the implementation of a strategic plan for sustainable exploitation over a longer period of time. A certain portion must be set aside for future infrastructure, education, income security programs (as well as clean-up) and not just serve to encourage more current program spending.

[Ryan]

I think to really analyze this, you can't just look at share of GDP. You have to look at share of those contributors to GDP that existed when the graph started, which still exist today, and you have to also control for government spending somehow. I'm not sure such an analysis is even possible (for lack of clean data), which is unfortunate, because it means pundits will continue to have plenty of latitude to spin the facts their own way.