Why export growth isn't a necessary condition for a sustained expansion

The Globe and Mail's Kevin Carmichael called me yesterday to talk about the idea that Canada's recovery won't be fully underway until exports started to increase at a significant rate. My comments didn't make the cut for the print edition, but they are in the online version of the story:

Not all economists are convinced Canada's economy is so reliant on exports.

Stephen Gordon, an economics professor at Laval University in Quebec City, noted that the economic growth of the boom years before the crisis was mostly the result of higher commodity prices, which generated income that fed a growing services industry.

Export volumes were little changed, and manufacturing production has been declining for years, he said.

“If people are willing to give us a lot of money for our raw materials, we should be cashing the cheques and saying, ‘Thank you,' ” Prof. Gordon said. “There's nothing wrong with a services-based economy.”

Nevertheless, the more common view is that Canada will need a thriving export industry to maintain stronger growth rates.

I'm now going to explain why the more common view is wrong.

We'll start with the expression for aggregate demand:

AD = C + I + G + NX

where

• C is domestic private expenditures on consumption goods and services,
• I is domestic private expenditure on capital goods,
• G is public expenditures on goods and services, and
• NX is net exports: exports minus imports.

Clearly, if net exports increase and everything else stays the same, then aggregate demand increases. But this isn't a necessary condition. In fact, the contribution of net exports to total GDP growth has been negative in two of the last three expansions:

Let's see what happens when you break net exports down into its components:

Not every expansion is a variation of the one we had in the 1990s. So why the obsession with exports?

Part of the problem is the way net exports enter into that expression for aggregate demand: exports have a positive sign, and imports have a negative sign. But that's exactly opposite to how we should be thinking about what the costs and benefits of international trade are:

• Imports are benefits: We engage in international trade to get things more cheaply than it would cost to make it ourselves.
• Exports are costs: We send exports to foreigners so that they will give us the imports we want.

Suppose we live in an economy where our international trade consists of exchanging truckloads of trees for truckloads of bananas. Some workers spend their time cutting down trees to send to foreigners, and the rest work on producing goods and services for domestic consumption.

Now let's suppose that foreigners are no longer satisfied with trees in exchange for bananas: they will now only exchange a truckload of bananas for a truckload of furniture. Since we must have bananas, some workers must shift out of the domestic production sector and start transforming trees into furniture for export. Even though the national accounts would record an increase in exports (that is, an increase in the value-added), this development is clearly bad news. The shift of workers to the export sector and away from domestic production means that fewer are available to provide the same level of goods and services we used to enjoy.

Now suppose things shift back to what they were before: foreigners are now willing to accept trees for bananas. Since the people who cut down trees are able to get bananas on their own, the workers who had been making furniture for export can go back to producing goods and services for local consumption. Although the national accounts would show a decrease in the value-added of exports, we are clearly better off.

These two stories are pretty much what happened in the last two expansions. In the early 1990's, commodity prices fell, and the only way for us to obtain the imports we wanted was to shift workers to the manufacturing sector, and to increase the value-added of exports. But devoting more of our productive capacity to making things that are to be consumed by foreigners isn't a path to prosperity, and workers' real buying power stagnated.

In 2002, commodity prices rose, and we were able to get the imports we wanted with fewer productive resources allocated to the export sector. The expansion of 2002-2008 was characterised by a shift out of export-oriented manufacturing, and these workers were able to produce more for domestic consumption. Exports stagnated, but real incomes increased.

In this context, the idea that "Canada will need a thriving export industry to maintain stronger growth rates" seems rather odd. I don't see how paying more for imports is going to make us better off.

Comments

[Curmudgeon]

If anything, zero growth in exports is a good idea because it protects against Dutch Disease.

An export-dominated economy is a bad thing.

[westslope]

"More" exports are nice when in previous months the value of exports has sharply declined. Any economic restructuring will at least temporarily lower living standards.

Darn our open economy anyways. If we were only obsessive savers--we're not--then, oh well, forget it..... here comes the roller coaster. Wheee!!!!!!

[Ramanan]

If there is only one country in the world, "exports" has no meaning. IMO, the IMF has forced countries to go into that path and a better way to grow is domestically. "G" always has the same sign as "X" when they occur in expressions together and it is better to increase G than increase X.

[Adam P]

Stephen, seems to me the facination with export growth can be traced to an obsession with high employment. Higher AD coming from X doesn't directly increase domestic aggregate consumption and thus doesn't increase domestic living standards. However, higher incomes coming from exports will feed through to domestic demand for services and the like.

The point though is that higher AD leads to higher employment regardless the source and their is a distributional effect in play. Extracting commodities or providing services might employ unskilled labour that doesn't have the skills to work (productively) in the manufacturing sector. After all, the statement "workers' real buying power stagnated" implies that they didn't have a comparative advantage in producing manufactures relative to the external competition.

I'm not quite claiming this is actually the case (I don't really know), only that some might believe it to be the case. China makes a similar claim about their export sector, I don't actually believe the argument in the case of China, I think for them it's about rents for the ruling class but western China defenders do reasonably make this sort of argument. (What makes me think the argument silly about China is the implicit assumption that there is no way to employ unskilled Chinese in producing services to be consumed by other Chinese, there is less external competition in domestic services.)

[Mike Moffatt]

In this context, the idea that "Canada will need a thriving export industry to maintain stronger growth rates" seems rather odd.

Agreed. But what if you change the focus, to, say, Southwestern Ontario?

Overall I agree with your analysis - but there are important distributional effects that need to be considered. All I keep thinking about is the man with a foot in a bucket of ice water and the other foot in a bucket of boiling water...

[Adam P]

Just to be clear the China analogy is meant to be taken loosely. In China the (silly) claim is something like: "We must employ people in the export sector because we just can't employ them to produce for domestic consumption".

In Canada the claim could (reasonably) be: High commodities prices drive the growth rate of wages for unskilled labour higher than it would be in the manufacturing sector. Thus high commodities prices might be necessary "to maintain stronger growth rates", emphasis on "strong".

[Stephen Gordon]

Mike (and others): Certainly there is an implicit assumption that workers can move from one sector to the other, and from one region to the other. These transitions may not be instantaneous or costless, but they do happen. And given the usual churn in the labour market (10%-15% of workers change jobs in a given year), they can happen faster than you might think.

[Adam P]

the point is not that they can move, it's a matter of where is their marginal product higher. Double the price of a comodity and you double the marginal product of the worker extracting it, this then drives up the wages of all unskilled labour, (assuming extracting the commodity is an unskilled job).

[Mike Moffatt]

Mike (and others): Certainly there is an implicit assumption that workers can move from one sector to the other, and from one region to the other. These transitions may not be instantaneous or costless, but they do happen. And given the usual churn in the labour market (10%-15% of workers change jobs in a given year), they can happen faster than you might think.

Agreed. For instance, there was a job fair in Windsor, ON a couple years ago (2007?) where there were more employers from Sask and Alberta than there were from Ontario.

[Joseph Heath]

I think you're giving Carmichael too much credit. I actually clipped this paragraph from my paper when I saw it on Wednesday morning:

"Right now, Canada's recovery is being almost entirely fuelled by consumers, raising questions about how long a relatively small population of 33 million people can prop up a $1.3-trillion economy."

This is just a weird, weird thing to say -- unless you're someone who simply fails to realize that consumption is the flip side of production, exports the flip side of imports, etc.

[Nick Rowe]

Joseph: "This is just a weird, weird thing to say -- unless you're someone who simply fails to realize that consumption is the flip side of production,..."

Yep. How can people who produce and sell $1.3 trillion worth of goods and services possibly have the income needed to afford to buy $1.3 trillion worth of goods and services?

But I'm not sure that all economists understand this point either, unfortunately.

Though it might be possible to re-state the views of those who worry about net exports, in a way that might make a bit more sense. Net exports are the flip side of national savings minus investment. If there really is a global savings glut, then countries that have investment higher than national savings are, in a sense, doing "more than their fair share" of spending.

[himaginary]

Very intriguing post for us Japanese, too. Here is my application of this post's analysis to Japan:

In the early 1990's, commodity prices fell, and we were able to get the imports we wanted with fewer productive resources allocated to the export sector. Workers were able to produce more for domestic consumption.
Unfortunately, we didn't have enough domestic consumption to match that redundant supply power, and slid into liquidity trap.

In 2002, commodity prices rose, and the only way for us to obtain the imports we wanted was to shift workers to the manufacturing sector, and to increase the value-added of exports. Yes, our terms of trade were clearly worse off, but that shift seemed to help us to get out of liquidity trap -- until Lehman shock arrived.

[westslope]

Labour market churn is good but momentarily painful. Perhaps that's why so many Canadians devote massive political resources to minimizing churn.

Recessions can be viewed as larger than usual labour market churning events.

Curious how many pundits shrug off a high valued currency because it allows businesses to update machinery and equipment. Clearly the M&E capital in this story is being imported, not bought from domestic sources. The enhanced productivity should ideally improve market penetration in both domestic and foreign markets. But I suspect that the pundits have in mind performance in foreign markets.

[Matthew]

An export-dominated economy is a bad thing.

Germany is another good example.

[Rogue]

This just shows that current conventional thinking commonly inter-changes national economies with individual firms. Individual firms do not consume their own production, they have to sell to earn income. National economies on the other hand can grow simply by having their domestic participants produce more and trade more with one another.

We need to go beyond looking at trade as Canada Inc. vs China Inc vs Japan Inc. There can still be growth in more production even if more of it is consumed locally. The real problem only starts if you are importing more (buying from outside) than you are exporting (selling to outside).

[Stephen Gordon]

This just shows that current conventional thinking commonly inter-changes national economies with individual firms.

That's a great point. Whenever the advantages of free trade are being discussed in the media, they almost always talk about export opportunities - which is the perspective of an individual company. But the real benefits are the new opportunities for importing.

[Matthew]

Stephen and Nick:

This is unrelated, but the BOC's latest financial system review seems to highlight household debt as a significant problem. Would you agree? In previous posts you were skeptical about this.

[Nick Rowe]

Matthew: I will have to take a look. I am skeptical about aggregate debt. The distribution of debt, though, is another matter. Last time I looked, the BoC analysis was v. good, because it looked at distribution.

[Too Much Fed]

Matthew, here is how I like to look at it.

(S-I) of the rich [I think there is a way to get both domestic and foreign in there] equals G-T minus (S-I) of the lower and middle class.

When the lower and middle class stopped going into currency denominated debt to the rich and some even tried to default, the rich got their gov'ts to go into currency denominated debt for them and tried to stop the defaults too.

[Too Much Fed]

Imports are benefits: We engage in international trade to get things more cheaply than it would cost to make it ourselves.

Sounds like economistspeak (or something greenspan would say) for we like to exploit cheap labor. It also sounds to me that a supply-constrained economy is also being assumed.

Exports are costs: We send exports to foreigners so that they will give us the imports we want.

What happens if the foreigners don't want or need the exports or the exporting country does not get high enough prices or quantities for its exports? Rev up the currency denominated debt machine? Sell other financial assets? Start lowering wages?

Will Australia and Canada be a little different than the USA (except maybe with food exports), the UK, and Japan?

[himaginary]

More than fifteen years ago, Krugman wrote:
"One of the most popular, enduring misconceptions of practical men is that countries are in competition with each other in the same way that companies in the same business are in competition. Ricardo already knew better in 1817. An introductory economics course should drive home to students the point that international trade is not about competition, it is about mutually beneficial exchange. Even more fundamentally, we should be able to teach students that imports, not exports, are the purpose of trade. That is, what a country gains from trade is the ability to import things it wants. Exports are not an objective in and of themselves: the need to export is a burden that a country must bear because its import suppliers are crass enough to demand payment."
(David Henderson fully endorses this essay. )

It seems that old misconceptions never die.

[Too Much Fed]

Nick's post said: "Matthew: I will have to take a look. I am skeptical about aggregate debt. The distribution of debt, though, is another matter. Last time I looked, the BoC analysis was v. good, because it looked at distribution."

Should people be skeptical about high aggregate currency denominated debt levels when price inflation is low?

Does that usually mean some entity is suffering negative "real earnings" growth somewhere (meaning the distribution of currency denominated debt is skewed somewhere also)?

[Alex Plante]

Canada really has two economies - an industrial core in Southern Ontario and Quebec, and a resource-based periphery in the rest of the country. When commodity prices are high (such as in the 1970s or 2000s), we have a strong dollar which hurts manufacturing in the core, but the periphery booms because of the improvement in the terms of trade. When commodity prices are low (as in the 1980s and 90s), we have a cheap dollar, manufacturing exports from the core expands, causing the economy of the core to boom, but the resource-dependant periphery suffers.

In both cases we have economic growth, but in a high-commodity-price world the growth is concentrated in the periphery and comes about from improved terms of trade, whereas in a cheap-commodity world, growth is concentrated in the core and comes from increased manufacturing exports because of a cheap Canadian dollar.

[Jean Cooper]

Thanks for the charts, very interesting to see what has contributed to growth over the last few decades.

"Right now, Canada's recovery is being almost entirely fuelled by consumers, raising questions about how long a relatively

small population of 33 million people can prop up a $1.3-trillion economy."

Is this all that weird? Canadians now have a debt to income ratio of 1.42, which is the highest it's ever been. Could some of

this 90% contribution to GDP (2001-2008) be as a result of taking on more debt to fuel consumption? What happens when more

debt can't be taken on and consumption drops? What will take up that very big place in our GDP? I don't know, I'm just

throwing out some questions because I think that the debt load of Canadians is rising,and they won't be able to consume as

they did in the past 8 years (also, Canadians took a big hit in their investments in the past year, so are also saving more

for their retirement)

I also don't get why making a value added product like furniture would be less beneficial for the economy than simply

exporting logs. Hopefully, you would receive more money from your export of furniture than you do for just exporting logs,

and this money would allow you to buy more imports. It's true than people would be taken from the service and goods area, but

as Mish says "we don't need any more nail salons, burger joints, and strip malls". I don't think it would be a big loss if

less people were in the service and good producing sector, and more were in the value added export sector.

Where do you see growth coming from if the consumer end of things drops off?

[Nick Rowe]

Jean: "Canadians now have a debt to income ratio of 1.42, which is the highest it's ever been. Could some of this 90% contribution to GDP (2001-2008) be as a result of taking on more debt to fuel consumption?"

No. It wasn't. This is an example of a fallacy of composition. It is true that any individual can go into debt to consume more. But it's not true that every individual can go into debt to consume more. Somebody must be lending them the money. And, in Canada's case, it wasn't foreigners. Canada was reducing/eliminating its net foreign debt over the same period.

For every financial liability there's a financial asset. Some Canadians were borrowing to spend. But other Canadians were saving to lend.

If, under one definition of "debt" Canadians have a 1.42 ratio of debt to income, Canadians also have a 1.42 ratio of assets to income, if we define "debt" and "assets" the same way.

[Jean Cooper]

Hi Nick, thanks for your response. I'm wondering though if people are really saving, or are banks leveraging themselves more? I understood that Canadian banks were leveraged about 20 to 1. I recently read an article from Sprott Asset Management where they say that Canadian banks are leveraged about 30 to 1. (http://www.sprott.com/Docs/MarketsataGlance/11_09%20Dont%20Bank%20on%20the%20Banks.pdf) I don't know which is the correct number, but let's say banks are leveraged 20 to 1 and gradually increase their leverage to 30 to 1. They are increasing their lending by 50%, but there are no savings backing up this lending. The bank is just taking more risk by increasing their leverage. So they've basically created money by leverage, and people are spending this money, but there is no corresponding increase in savings. Isn't this what caused the financial crisis in the U.S.? Too many bad loans, and banks were too heavily leveraged.

[Nick Rowe]

Hi Jean:

I don't know if the Sprott figures are accurate. But if they are, that means a shortage of bank capital, which is a very specific type of asset.