The most important development in the Canadian economy over the past 5 years has been the 40% appreciation of the CAD-USD exchange rate since January 2002, largely due to the increase in the price of oil and other commodities. The effects have been predictable: booms in the mining and oil sectors, and hard times for manufacturing. That sounds like a recipe for 'Dutch disease', but it didn't happen: unemployment is down to levels we haven't seen since the 1970's, and the employment-population ratio is at an all-time high. Moreover, the effect on wages has been (marginally) positive: if the distribution of employment across sectors had stayed at what it was in 2001, then average wages would be a fraction of a percentage point lower than what they are now.
The key seems to have been a labour market that has proven to be flexible enough to handle the sectoral and geographic shifts produced by the CAD's appreciation.
It's important to understand what 'labour market flexibility' actually is. For example, many seem to think of it as a synonym for reducing wages - an interpretation that is (as we've seen) exactly wrong. What it means is the ability to re-allocate workers to sectors where labour has become relatively more productive. This increases labour productivity for the entire economy, and will increase wages.
An interesting feature of the current re-adjustment is geography: employment growth is largely concentrated in Alberta. And workers are moving to where the work is. To illustrate the point, I've calculated the gap between employment growth would have been in each region if it had followed the national average since 2002, and actual employment growth in that period. Here is a plot of that gap graphed against net interprovincial migration over 2002-2006:
The gap between actual and trend employment corresponds almost exactly to net interprovincial migration for Ontario, Alberta and for Atlantic Canada. BC's net migration is lower than what the 45-degree line would predict, but it receives lots of migrants from outside Canada. Physical proximity to Alberta might explain why Manitoba and Saskatchewan's net migration is larger than what the employment gap would predict, and it seems reasonable to conjecture that net migration out of Quebec is attenuated by the language barrier.
Of course, people aren't willing to haul up stakes and move to Alberta for just any job. Here is net interprovincial migration (this time expressed as a percentage of 2001 population levels) against the per cent deviation of weekly earnings from the national average:
Not only is Alberta the primary source of new jobs, those jobs are generally better-paying ones than elsewhere. And because workers from outside the province have been willing and able to move to Alberta, employment and wage growth have managed to stay relatively steady over the past five years.