Who benefits from the oil sands? Who benefits from saying that only Alberta does?

This didn't pass my sniff test:

The economic benefits of oil sands development, while considerable, are unevenly distributed across the country, making interprovincial tensions understandable. While provinces other than Alberta are projected to benefit, modelling by the Canadian Energy Research Institute projects that 94 per cent of the GDP impact of oil sands development will occur within Alberta. With so much benefit concentrated in one province, one can hardly call fast-tracking oil sands expansion a nation-building project. Little wonder that the promise of benefits from oil sands development is cold comfort for Ontarians and Quebeckers as the once-dominant manufacturing sector struggles to reinvent and revitalize itself.

Did you read "94 per cent of the GDP impact of oil sands development will occur in Alberta" and interpret it as "94 per cent of the economic benefits of oil sands development will occur in Alberta"? I'm convinced that that the vast majority of the people who read that passage on the Globe's op-ed page interpreted it that way. And I'm only slightly less convinced that the author meant his readers to interpret it that way. Of course, that would be the wrong interpretation.

For reasons I'll get to later, there seems to be a concerted effort to convince Canadians that almost no-one outside Alberta is seeing any economic benefits from high oil prices. For the most part, these efforts appear to be enjoying some measure of success. But the fact of the matter is that the oil sands have increased incomes across Canada to an extent much greater than that paragraph implies.

Firstly, income is what matters, not production. It's remarkable how so many Canadian pundits can seize terrier-like on an indirect cost of an exchange rate increase and completely ignore the direct benefit: how cheaper imports increase Canadians' purchasing power and real incomes. Roughly a third of the increase in Gross Domestic Income since 2002 - see here for an explanation of what it measures and its relationship with GDP - came from the improvements in Canada's terms of trade.

This breakdown of the income growth over 2000-2007 suggests that half the income growth came from the terms of trade effect.

Moreover, the regional distribution of those benefits has been much more broad than what that "94 per cent of the GDP impact" is meant to suggest. The most obvious mechanism for redistributing oil sands revenue is the federal government. This involves more than just the equalization payments made by the federal government to the provincial governments. The data here break down federal government revenues and expenditure by province. Here are expenditures by region/province as a percentage of revenues:

And this is how the federal government balance breaks down by province/region:

Between 2003 and 2008, the difference between what Albertans paid to the federal government and what they received increased by about $12b - roughly 0.75% of Canadian GDP, or $3,000 per Albertan. I don't know how much people were expecting Albertans to share with their fellow Canadians, but that's not nothing.

But the most important mechanism for redistributing the income generated by the oil sands is its effects on wages across sectors and regions.

Here is the undergrad micro version of what happens to employment and wages when there's an increase in labour demand in one sector. Suppose we start with an equilibrium where wages are equalised across sectors, and there's an increase in the demand for labour in sector A:

Unsurprisingly, this increases wages sector A. I suspect this summarises the thinking behind the "only Alberta benefits from the oil sands" meme, but the story doesn't end there. Workers in sector B will see the higher wage w^* in sector A and some will shift sectors in order to take advantage of those higher wages. This shift has the effect of shifting out the labour supply curve in A (bringing wages down there) and shifting in the labour supply curve in B (bringing wages up there):

This isn't a fully-worked-out treatment, but it does illustrate the point that the decline in employment in sector B occurs because workers leave in search of higher wages. This mechanism is implicit in all Dutch disease stories, but almost no-one seems to notice that a wage increase is what makes it work.   

Notice also that this model doesn't predict that all workers move from B to A; only enough to bring wages in line across sectors. And not all of this movement required people to physically move - although net interprovincial migration to Alberta between 2002 and 2008 was some 160,000 persons. An improvement in the terms of trade benefits sectors serving the domestic market: cheaper imports means more extra income to spend on domestically-produced goods and services. Workers in Ontario - centrally-located and with almost 40% of the population - are not necessarily the worst-placed to take advantage of this shift.

So much for theory - what do the data for wages say? I'm going to use two sources: the annual weekly earnings data from the Labour Force Survey - which goes back to 1997 - and those from the Survey of Employment, Payrolls and Hours, which go back to 1991. The LFS data have a shorter sample, but they also provide estimate for median earnings. The provincial CPI data are used to transform them to constant 2002 dollar weekly earnings.

No-one will be surprised to see what happened to wages in Alberta: 

But look at what happened here in Quebec:

I don't see how Quebecers could look at those numbers and conclude that the maladie hollandaise was something to be feared.

What about Ontario?

 

Wages in Ontario may not have grown as strongly as they did in Quebec, but I don't see bloodshed there.

And here are the Canadian averages:

After years of stagnation, the 2002-2008 period saw strong growth in both average and median real weekly earnings across Canada.

The theory of asymmetric labour demand shocks doesn't say much about what will happen to total employment. If you eyeball the graphs above, it looks as though total employment increases, but I chose the slopes of those curves at random.

So here are the employment rates:

and the unemployment rates:

Again, I see good news for Quebec and not-bad-news for Ontario here. It may hurt Ontarian pride to no longer perform better than the Canadian average, but this was clearly a case of the ROC catching up, and not a deterioration of conditions in Ontario.

There are two, overlapping classes of people for whom it is convenient to downplay the economic benefits of the oil sands outside Alberta.

1) Politicians. Mark Carney's recommendation to Canadian manufacturers is, essentially, to face the realities of life with a Canadian dollar that trades at levels higher than it did in the 1990s. This is a much less seductive line than "Blame Alberta and the oil it rode in on", and Tom Mulcair cannot be disappointed with the bump in the polls that the NDP has received since he started talking about Dutch disease.

2) Climate hawks. As Spencer Keys and UBC's George Hoberg note,

If the choice comes down to oil sands expansion with demonstrably inadequate environmental checks and balances, or a sketchy economic argument with powerful narrative potential, the choice seems pretty clear for climate hawks.

Mr Mulcair wouldn't be the first politician to reach the top of the greasy pole by exploiting interprovincial rivalries. But it seems to me that this is a remarkably irresponsible strategy for those concerned with climate change. The biggest asset that environmentalists have is their intellectual integrity, and they simply cannot afford to fritter that credibility away by cheerfully adopting an economic analysis they know to be wrong in order to advance their agenda.

Comments

[Kevin Andrew]

It would be interesting to look at some of those graphs disaggregated. For instance, if the gains have been concentrated in the top end of the income distribution then using provincial and national averages could miss part of the story. The reason that the narrative is successful is because a lot of people perceive there to be less "good" jobs in Ontario and Quebec. While no doubt there are winners in the ROC from oil sands production, Mulcair's argument would fall on deaf ears if there wasn't already a perception that a lot of people are hurt by the exchange rate appreciation (although you are correct to outline the advantage of terms of trade improvements).

[Stephen Gordon]

That's why I made it a point to graph median real wages along with the averages.

[Kevin Andrew]

Oops, that was silly of me. Nice post.

[Gregory Sokoloff]

Excellent post. Many thanks.

Now Albertans should use your analysis to pose the following question to ROC: are you willing to give up the $X benefits you get from the oil sands in order to eliminate its harm to the environmental? Of course the answer for many of us would still be yes, but it would be interesting to model the answer for different values of X.

[Nick Rowe]

Yep.

Interprovincial migration is also (presumably) always bad for the GDP of the provinces who lose population. Should we ban interprovincial migration? GDP is a bad measure of welfare. At a minimum we should switch to GDP per capita.

[Robert McClelland]

Your post appears to be little more than a variation of the broken windows fallacy.

[J.V. Dubois]

Excellent post. But anyways, even if (as I see it) all benefits (such as real income) would occur in Alberta, would it be sufficient condition to ban it? I cannot see any moral grounds where one would deliberately sabotage raise in living standards of fellow countrymen just because they can. They are still Canadians, aren't they? And even if they would be of different nationality the case is still very weak.

At best these accusations can be interpreted as calls for higher taxation or more "solidarity" in between provinces. But I do not think that could be a something those any groups you mentioned at the end would aim for. It is just about creating negative PR that can be accumulated and then used up in some not related campaign.

[YM]

I lived in Quebec and now I live in New York state. It is much easier to find hormone free milk in NY than Qc. The quality argument is BS to me. If Quebec or Canadian milk producers were competing on quality, there would much more brands around than we see. Instead we have all producer shipping there milk in the same tank. Try to find out where your milk came from exactly, how easy is that? What does that tell you about quality?

[Simon van Norden]

Stephen, please help me understand how I should think about the economic impact of the Oil Sands (or other non-renewable natural resources.) I'm forever wondering whether we economists are thinking correctly about this (and I have to visit my neo-con pro-logging pro-oil brother-in-law soon, so I want to get my thinking straight on this.)

My problem is that the economic debate is largely focused on the impact on GDP. But with a stock of non-renewable resources, GDP will be influenced by how quickly we choose to run those stocks down. I would have thought that we should care more about wealth; maximizing the present value of the rents we can generate from those resource.

I'm no resource economist (resource economics is one of my "known unknowns"), but I know there's a literature on the optimal extraction rates of natural resources. I also know that serious proposals have been made to overhaul our system of national accounts to better account for the depreciation of natural resource assets.

At the same time, when I try to think of how to put a price tag on our resource wealth, my mind boggles. I'll need a very-long-term discount rate; should that reflect the rate at which oil companies can borrow? or the federal government? or Alberta? I'll need very long-term price projections for the price of oil. Is it going up due to peak oil and Asian growth? or are we seeing a temporary bubble before the long-term elasticity of demand kicks in? I'm need projections of productivity growth in extraction business as well as guesses about total recoverable reserves. And to deal with the uncertainty in all these projections, I guess I'll need to take a stand on our degree of risk aversion.

I suspect that the problems in estimating the wealth associated with a natural resource has kept that approach from catching on. But I can't shake the feeling that focusing on GDP impacts is the wrong way to go.

What's your take on this?

[Jciconsult]

I do NOT subscribe to any policies to manipulate the exchange rate. However, the magnitude of an exchange rate impact probably exceeds the value of any incremental demand from A;berta. let's look the relative size of the imports from Alberta and Ontario exports. International and interprovincial imports are roughly the same size for Alberta according to the interprovincial trade statistics (2008 is last year unfortunately). However, for ontario, the domestic supply to Alberta is roughly 1/10 of the size of Ontario's international exports. Exports are far more important to Ontario than sales to Alberta. To associate much resilience in Ontario employment with the incremental effects of the oil sands is a bit of a stretch.

[Neil]

@Simon - don't worry, the hard work is already done. The oil companies themselves already have a method for determining the present value of their in-ground reserves to put on their balance sheets. All we need to do to our public accounts is start thinking like a business does - the present value of future royalties should be recorded as an asset. Prove more reserves and you can increase that asset, extract oil and they decrease.

This way, we'll stop thinking about royalties as revenue. We can correctly recognise that collecting royalties is simply collecting on a debt that was already owed to the public purse. And then maybe Alberta will actually set a sustainable tax rate, instead of living off oil royalties like they'll be here forever.

The biggest problem in government accounting generally is that it focuses on cash flow instead of actual profitability.

[K]

Simon: Really well put. Amen!

Neil: "The oil companies themselves already have a method for determining the present value of their in-ground reserves to put on their balance sheets."

Yeah, but like Simon says, their discount rate is quite a lot higher than ours which means that they are vastly more incentivized to extract sooner rather than later.

What we should do is first decide on a schedule on which we want to consume our resources. *Then* contract for the cheapest bidder to do nothing more than extract it for us. There is no reason for private balance sheets to bear commodity price risk when the public is likely more able to bear long term price risk. In case of price collapse, we can always just leave it in the ground for the grand children.

Unfortunately politicians discount rates are even higher than the industry's, so the grand children are basically screwed.

[Simon van Norden]

Neil:
Thanks for the info....I didn't know that....and it is clearly a step in the right direction. As K noted, we can argue about discount rates (and private benefits vs social costs)....but it helps us move away from an extraction-rate focus.

K:
"Unfortunately politicians discount rates are even higher than the industry's, so the grand children are basically screwed."

Actually, I've heard some people use this as a argument for monarchy; since they should care about the state of the nation that they leave to your kids, they should get the discount rate right!

(It's the royal jubilee.....I couldn't resist.)

[Simon van Norden]

sorry....should have been "....leave to THEIR kids....."

[Patrick]

I have a question. Sorry if it's stupid: is oil induced Dutch disease even plausible? FX is a big-ass market. We consume quite a lot of the oil we produce. The oil we export is traded in $US - much of which I assume firms leave in $US rather than converting to $CDN. Also, if I look here:

http://www.bankofcanada.ca/rates/exchange/monthly-average-lookup/

Eye-balling the numbers it seems to me that the run-up in the $CDN v. $US occurred between 2002 and 2007, and has since plateaued. Seems to me that it coincides with China's boom and developed world building bubble. People wanted raw materials and we have them. It wasn't just oil.

To my mind, the big mystery is why the $CDN has tayed high relative to the $US post-crisis/crash. What the heck is generating all that demand for $CDN? (Ghawar, I'm looking at you).

[henryc]

A note on some GDI implications over the longer term. From Statscan:

"Over the whole period between 1961 and 2008, the gap between Canadian and US GDI per capita declined. Higher energy prices raised Canadian income relative to American income in both the 1970s and after 2000. Incomes in Canada reached their highest relative level at 98% of American incomes in the early 1980s and have recently returned to levels over 90%. However, whether this increase is the result of an emerging trend or is just the result of a short-run commodity cycle depends on the future course of world commodity prices." [http://www.statcan.gc.ca/pub/11-010-x/2009012/part-partie3-eng.htm]

Now, the Conference Board of Cda [http://www.conferenceboard.ca/hcp/details/economy/income-per-capita.aspx#Falling_oil] had (2007) an interesting comparison of Canada and Norway v.a.v. their GDI's and the relative sizes of their resource sectors (Norway's is relatively larger than ours). How each country has managed its resource sector spells the difference in the results. Norway has a $300 bi sovereign fund; Alberta $15 billion. For the past three decades, in terms of improving its GDI, Conference Board rates Norway A,A,A, while Canada earns a B,C,C, for each consecutive decade. (But, heck, they also thought Ireland was a showcase for GDI improvement, at the time...)

And if we go to Statscan again [http://www.statcan.gc.ca/pub/75-202-x/2009000/analysis-analyses-eng.htm], for relative median income mobility, there is little difference between the 1993-1997 and 2005-2009 overall results. The real story is the relationship between the unemployment rate trend and the income trend over time, particularly, of course, for the 25-54 year-olds.

So, yes, I see the point of looking at GDI, at terms of trade, at cross-sector wage changes as a function of demand changes in one sector, etc. What is not clear to me -- I'm no expert -- is whether or not we are becoming more reliant on a resource-based economy, and if so, whether that's adequate, and if not, whether we really have any choice in the matter.

[beowulf]

Don't discount the diplomatic benefits of the oil sands. Because US politicians of all stripes strongly endorse the goal of "North American energy independence", Uncle Sam won't do anything to hamper exploitation of Albertan oil (which is why that pipeline is going to be built regardless of Obama's personal wishes). The implications of this are that when the US eventually gets around to reducing its trade deficit (a 4% of GDP demand leakage year in, year out), its because of the oil sands that Canada will almost certainly be exempt from any trade barriers.

[Stephen Kessels]

The right-wing spin is oil wealth is shared. Reality is that it is not. 40 percent of Ontario's tax revenues is still being siphoned out to fund programs. Why isn't Alberta paying more when it can afford it? Harper has arranged for Alberta to receive another $1 for healthcare transfer when Ontario is already underpaid in healthcare transfers.

Let's face it. Ontario is getting the short end of the stick.

[Stephen Kessels]

There won't be a east vs west situation if oil wealth is truly shared. Since it is not, the other alternative is for Ontario to take a hiatus from the federation. Ontario has supported the country for decades; it is time for Ontario to clean house. That means, Ontario's tax revenues be used solely for the benefit of the province. Let the rich provinces do the heavy lifting for the country.