Peter Van Loan’s Made-Up Trade Statistics

Here’s an amusing epilogue to my recent CCPA report on the possible economic and employment impacts of a free trade agreement between Canada and the EU.  (Here is the report; here is my blog on it.)

The day after the report came out, reporters asked Trade Minister Peter Van Loan to comment on its predictions that Canadian industries (mostly in manufacturing) would lose jobs (as many as 150,000 in total) as a result of free trade.  His response was curious for several reasons.

[Here is a Globe and Mail web story with Mr. Van Loan’s comments; similar stories appeared in the Montreal Gazette and Toronto Sun.]

First off, Mr. Van Loan suggested he “had no difficulty dismissing” the report (he didn’t mention actually reading it!) because its author and publisher are “ideologically opposed” to free trade.  This strikes me as a rather “ideological” way to live one’s life: dismissing research based on the political outlook of the researcher (as opposed to the validity or lack thereof of the research).

Never mind, more interesting was Mr. Van Loan’s claim that contrary to the CAW’s past warnings about the impact of free trade on the auto industry and other key sectors, free trade has clearly benefited Canada’s auto industry and Canadian workers more generally.

This will come as a pleasant surprise to tens of thousands of laid-off Canadian autoworkers.  They must be partying in Windsor tonight.

Here’s the one piece of evidence the Minister offered to support his case: Canada accounts for 9% of the North American market for vehicles, he said, yet is the source of 21% of North American vehicle production.  This implies that Canada possesses more than twice “its share” of the continental auto industry — contrary to the CAW’s past warnings that our role in this industry would decline.

Only problem is, neither statistic is valid.  The following table reports Canadian auto sales and production, in absolute terms and relative to North American totals, for 2009:

 

 

 

 

 

 

 

Canada accounted for 17% of all North American vehicle assembly last year, and 12% of all North American sales (not 21% and 9%).  This still makes it seem as if we are assembling more than we are buying (by about half), but this is not the case.  We can’t forget that close to one-third of North American vehicle sales come from offshore (Japan, Korea, and Europe), whereas North America exports virtually zero vehicles offshore.  In fact Canada’s production in 2009 was actually almost perfectly equal to its sales (at just under 1.5 million units).  That’s half the production levels of the late 1990s, and the lowest ratio of production to sales (1 to 1) that our auto industry has recorded since the 1960s (when we signed the Auto Pact, a managed trade agreement, with the U.S.).  As recently as the mid- and late-1990s, we assembled 2 vehicles for every 1 sold here.

The most appropriate comparison to make (in order to judge Canada’s relative production position within the North American market) is between Canadian production and sales both as a share of total North American sales.  That shows whether we’re getting a proportional share of the continental market.  (Given the flood of offshore imports, a constant share of continental production is no guarantee of a steady share of continental sales.)  Compare the shaded cell on line 5 of the table, with the shaded cell on line 7.  They are almost exactly equal — not surprisingly, since Canadian production and sales were also almost exactly equal.

Even this comparison overstates Canada’s relative position, however, since it considers only sales and production of final vehicles.  Canada has a relatively strong position in vehicle assembly, but a weaker position in auto parts (of which we produce much less than “our share”).  In overall auto production (considering both assembly and parts), Canada incurred a $15 billion trade deficit last year — our worst in history.  That compares badly to the annual auto trade surpluses we used to rack up prior to NAFTA.  Within NAFTA, the collapse of our automotive trade balance has been even more dramatic: moving from annual surpluses of $10 billion before NAFTA, to a $3 billion deficit last year.

 

 

 

 

 

 

 

In terms of employment, the trend under NAFTA is also clear: over 20,000 assembly jobs gone, and over 10,000 in parts.  (Measured against the higher peak levels that industry attained in the late 1990s, the losses have been more severe: over 60,000 auto jobs lost since the turn of the century.)  In no way can it be credibly claimed that NAFTA has secured Canada’s proportional share of auto production or jobs; our industry has declined dramatically by any measure.

Where did Mr. Van Loan’s 9% number come from?  Some years ago, before the U.S. economy melted down, Canada’s sales equaled 9% of the continental market.  (Now, because U.S. sales fell so dramatically, our market accounts for a larger share of the North American total than it used to.)  So perhaps this error could be attributed, generously, to out-of-date statistics.

But where did the 21% figure come from?  This is more interesting.  Canada has never accounted for that much of North American aut0 production. Our share of continental assembly has been relatively stable (at around 15-17%) since the late 1980s, despite layoffs and plant closures.  This reflects the fact that we’ve done better at hanging onto production than the Americans  (whose auto industry has shrunk more dramatically), but we’ve still lost proportionately to Mexico (whose auto production has doubled under NAFTA, while ours shrank).

I have contacted Mr. Van Loan’s office three times to ask his media relations person for background to the 21% number.  As a courtesy I wanted to give them the chance to explain the number, but after more than a week I have yet to receive a response.

Here is where I think it came from: Canada’s assembly divided by the sum of Canadian and U.S. assembly, equaled 21% last year.  (See line 6 of the first table above, right column.)  Likely an anonymous DFAIT minion calculated what he thought was a handy factoid to support his Minister’s denunciation of my report; unfortunately, he or she forgot to include Mexico in the NAFTA total!

Frankly, I wouldn’t object to NAFTA so much if Mexico were not part of it.  Not that I have anything against Mexico.  It’s just that under NAFTA, Canada’s bilateral trade deficit with Mexico swelled to $12 billion last year — almost half of that due to autos.  That deficit translates into hundreds of thousands of lost jobs here in Canada.

This error is so silly, I am not surprised that Mr. Van Loan’s office won’t respond.  But I think we deserve a little bit more statistical accuracy, and a little bit more integrity (being willing to acknowledge and correct a public factual mistake), from our Trade Minister.  After all, right now Mr. Van Loan is negotiating a free trade agreement with a major auto export powerhouse: the EU.  It would be helpful if our negotiator knew a little bit more about the key export industries he is responsible for defending.

Our automotive trade with Europe today is precariously unbalanced.  Last year we imported $3.7 billion in automotive products from the EU (mostly from Germany), and exported just $174 million (less than one-twentieth as much).  Bilateral tariff and NTB elimination can only make that deficit even wider — a truth that even the government’s own economic study acknowledges (see Table 2.13 of the joint Canada-EU economic study).

At any rate, I think I now  better understand why the Conservative cabinet supported abolishing the long-form census.  It’s much cheaper to just make up your own statistics, than to go out and gather real ones!