They've run out of superlatives to describe the performance of Canada's job market in the last year. Literally. "I don't think jaw-dropping and earth-shattering do these numbers any justice any more," said Doug Porter at BMO Nesbitt Burns, after the release of figures showing another 58,000 new jobs were created in December. "A shocker, a barnburner, another head-spinning gain -- call it what you wish," said TD Bank's Marc Levesque. Well, they're economists; they're not paid for their literary skills.

As earth-shattering etc. as the employment numbers were, I have to say we buried the lead. The real shocker in the figures was not the nearly 560,000 jobs created in 2002, the largest absolute increase since figures were first compiled. Nor was it the 3.7% rate of growth in employment for the year, the best in 15 years. Nor was it that 60% of those new jobs were full-time. Nor was it that all this occurred in a year of falling employment in the United States, on whose economy we so famously rely.

No, the most extraordinary number in Statistics Canada's latest data dump is this: there are more Canadians at work today, as a percentage of the population, then at any time in our history. The employment rate, as it is called, reached a record 62.4% in December, up from just 60% a year ago. (Or to put it another way, the economy is making use of all but 38% of its potential supply of labour.)

You don't see this figure reported much. Canadians who get their news from the headlines or the TV would have probably heard only that the unemployment rate, the most popular gauge of how well the labour market is doing, had remained stuck at 7.5%, barely lower than the 8% at which it began the year. How is it that the unemployment rate could fall only half a percentage point, even as the employment rate was rising by two and a half? Because the unemployment rate is expressed as a percentage, not of the working-age population, but of the labour force: those who tell StatsCan's surveyers they are available for work.

So many people entered the labour force last year, more than 500,000 of them, that it was all the market could do just to absorb them all. Had the participation rate -- the proportion of the population who are in the labour force -- remained the same at the end of the year as it was at the beginning, the unemployment rate would have fallen to less than 6%, not far off U.S. levels.

But again, you had to dig deep into the coverage to find that morsel. When the unemployment rate is rising, commentators are always available to point out that the "real" unemployment rate is much higher, because of all those discouraged workers who have given up looking. But when the rate is falling, and workers start flooding back into the market, you rarely hear anyone suggest that the "real" rate is lower than the headline figure.

Mind you, it's a mug's game either way. The only significance of any particular unemployment rate is in comparison to another: what it was last year, or last month, or what it is somewhere else. For those comparisons to be meaningful, they have to be based on a common definition. (We know that 10% unemployment is "high," but only because it's usually much less than that. If unemployment is actually 7%, as it usually defined, but is redefined in such a way as to produce a figure of 10%, it may appear shockingly high -- but only because we are still subconsciously using the previous definition to assess its significance).

What we need, rather, is a measure of the health of the labour market that takes account of swings in the participation rate. Fortunately, that figure is already at hand: the employment rate. So why don't we use it?

I can think of a few reasons. One, the unemployment rate puts the accent on the negative, which the media are well-known to prefer: the glass is not 92.5% full, but 7.5% empty. Two, because it is a smaller number, it tends to produce more dramatic swings. Unemployment rises two percentage points, from 6% to 8%, it's shot up by a third! Employment rate falls by two percentage points, from 62% to 60%, it's gone down by, er, um, carry the three, what was the question again?

And three, the unemployment rate focuses our attention on the misery of the unemployed: those available for work, but unable to find it. While that's a legitimate object of concern, the more salient point in economic terms is how much of a society's total potential supply of labour is being employed. A society in which everyone who wants a job has one, but only half the population is working, will be poorer (other things being equal) than a society with 5% unemployment but a 65% employment rate.

In any event, it's not always clear what it means to be "available for work." The seasonal workers who show up on Atlantic Canada's rolls as "unemployed" are not looking for work in any meaningful sense. They are on hiatus, not as a result of some temporary misfortune, but as a planned, annual event. From society's point of view, the issue is not whether they want to work, in some hypothetical sense, but whether they are working. Switching to the employment rate as our yardstick of the state of the labour market would do more than iron out a statistical kink: it would change the terms of debate.