UH-OH. February's inflation number - a 1.8 per cent rise in prices from the same month a year before - may not seem like anything to get too excited about. But year-on-year figures tell us where inflation has been, not where it is now. The more alarming fact is that in the last four months, prices have been rising at an annualized rate of more than 4 per cent. That's unlikely to ease, as the economy approaches the limits of its capacity. As if financial markets didn't have enough reasons already to dump Canadian bonds.

Which means we are about to revisit the inflation battles of the last decade. The Bank of Canada will have to act before long to rein inflation back within its 1 to 3 per cent target range. At which point, it will meet the same chorus of critics as before. Why risk the recovery, they will say, over 4 per cent inflation? Nobody wants double digits, but a little inflation isn't going to hurt the economy, is it? It might even help - as a "lubricant."

The "lubricant" argument is based on two faulty assumptions. The first is extreme downside "stickiness": the notion that the economy has a structural bias against wage and price decreases, at least in nominal terms. A market economy depends on the prices of different things rising or falling as a signal of their relative scarcity. Stickiness means that, at zero inflation, these relative price shifts are flattened out: If no price can ever fall, then no price can ever rise, either. Inflation, on the other hand, allows wages and prices to fall in real terms, even if they are unchanged in current dollars.

Yet the past few years have furnished plenty of evidence that nominal prices are not so sticky as all that. The price of oil, which could never fall, according to the authors of the National Energy Program, because OPEC "has demonstrated its capacity to raise prices at will," is less than half of what it was a decade ago. The prices of many other goods and services, from computers to air travel, have fallen sharply in the face of heightened competition or declining demand.

The case for an inflationary lube job rests upon a second, even shakier assumption: extreme upside myopia. The implication is that workers and firms will somehow look the other way while inflation eats into their purchasing power. So long as their nominal wages and prices aren't falling, they won't mind or won't see that their real wages and prices are. After 30 years' bitter experience with inflation, this "money illusion" scenario is a little hard to sustain. You just can't fool people like that any more.

That the public tends now to expect inflation helps explain any observed downward stickiness in prices. Take our present situation. Having allowed inflation to bump up to 4 per cent, the Bank might reasonably be expected to let it stay there. So either the Bank confirms expectations, and inflation ratchets higher, or it confounds them - and we go through a recession. For if wages rise, and prices don't, then real wages jump, and with them unemployment.

Naturally, those in power are particularly leery of the latter prospect. Which, paradoxically, only makes it more likely. If wage and price setters suspect the authorities cannot stomach another recession, they will have little incentive to moderate their demands, still less accept any reductions, since (they believe) the Bank can always be counted on to engineer enough inflation to keep pace. At some point, the merry-go- round would have to be brought to a halt, with much squealing and grinding.

If, on the other hand, economic agents are convinced the authorities are just crazy enough to go through a recession rather than let wages and prices rise, they get a lot less sticky in a hurry. Monetary policy, in short, works on much the same principle as deterrence theory: the credible threat. And the Bank's credibility is never greater than when inflation is flat on the floor.

"A little" inflation does enough damage all on its own. Suppose an investor requires a return of 4 per cent, after inflation and taxes, to go ahead with a project. At a 50 per cent marginal tax rate and zero inflation he will need an 8 per cent nominal pre-tax return to justify his investment. But at 4 per cent inflation, he will need a return of 16 per cent (equals 8 per cent after tax, equals 4 per cent after inflation) - twice the original threshold. A lot of investments that would otherwise take place never get off the ground.

Four per cent inflation is not so little: at that rate, the purchasing power of the dollar is cut in half in just 17 years. But even a little inflation is too much. For "a little" inflation has a nasty way of becoming "a lot."