Politicians, pundits, and think- tank wonks have been lining up lately to proclaim the Great Recession of 2008-09 is over. Earlier in the summer, it was the Bank of Canada reading the eulogy of the recession; in July, the Conference Board of Canada joined the chorus. While a few politi- cians have advised caution in calling it over too quickly, most everyone is singing ”œDing, dong the (recession) witch is dead!”

So, are we now officially in a peri- od of economic recovery?

The problem with trying to pin- point the ”œend of the recession” and the ”œbeginning of the recovery” is that there is a definition for the former, but not for the latter.

In Canada, a recession is common- ly defined as at least two consecutive quarters of contraction in the GDP. (The National Bureau of Economic Research in the US defines recessions somewhat differently, using a wider range of indicators such as employ- ment and housing). So, by Canada’s definition, we have without question witnessed a true recession: we’ve seen at least two quarters of contraction (Q4 2008 and Q1 2009), and will almost certainly see contraction in Q2 2009, when all of the numbers are available.

But the third quarter of this year (July to September) is expected to eke out some modest GDP growth. That would formally mark an end to the recession.

An economic recovery, on the other hand, has no handy definition. Is a recovery simply the absence of a recession?

For example, if growth in the third quarter is weak (say, less than 1 percent growth on an annualized basis), and that is followed by several more quar- ters of very weak growth, and if we see the occasional quarter slipping back into contraction, would that still be considered a ”œrecovery”? It sure wouldn’t feel like it.

What constitutes a period of recovery is really up to individual interpretation " and that could pres- ent some very tricky problems for politicians and central bankers, who are trying to time the recovery with the much-talked about ”œexit strategy” whereby the fiscal and monetary stim- ulus come to an end.

Both Ben Bernanke at the US Federal Reserve and Mark Carney at the Bank of Canada have a lot on the line in timing their exit strategies. If they withdraw the stimulus too soon, it would be a repeat of the mistake made in 1937 when it appeared the Great Depression was over. The sudden withdrawal of stimulus hurled the economy back into more years of pun- ishing contraction.

On the other hand, if the stimulus is kept in place for too long, the risk is runaway inflation, a severely devalued currency, and an even larger mountain of public debt.

Adding to the problem is the time lag in the data. Expansions or contrac- tions in the economy are never really discovered with much certainty until months after the fact. It’s an unavoidable problem with the collection of financial statistics. Reporting move- ments in GDP too soon often results in large revisions to the data later on. But waiting until there is complete certainty means reporting on quarters that are so long in the past it would be nearly irrelevant.

The timing of the exit strategy will be more art than science, no mat- ter what. But if we can define a reces- sion, perhaps it would be a good idea to define a recovery as well.

Maybe we should define a recov- ery as at least two consecutive quar- ters of growth that meet or exceed the 10-year average level of GDP growth. By this definition, we could confi- dently call a recovery this year or next when we see two quarters of growth around 3 percent higher. (The settings of the exact parameters are obviously arbitrary, but so is the definition of recession).

Now that we are past the end of July 2009, we are more confident in saying that this nasty recession is over. But we can’t properly call it a recovery. The coming quarters will in all likeli- hood be sluggish, even if the economy is technically no longer in contraction. Calling it a recovery would just be set- ting us up for disappointment.

The recession may in fact be dead. But we’re still waiting for the recovery to fully be born.