The National Bureau of Economic Research, the non-profit organization that has long been responsible for marking the beginning and end of America's recessions, has finally declared that the recession that it said began in March 2001 is over. In fact, the NBER says that it ended almost two years ago, in November 2001.

Non-economists find it hard to understand why it took the NBER so long to decide that the US recession was over. Economics is not supposed to be an exact science, but isn't this tardiness taking things a bit too far?

More importantly, if the recession is over--and has been for so long--why don't Americans feel any better? Why is the economic mood so negative? The answer lies in the definition of a recession. In fact, the questions posed by the NBER may not be the right questions, or at least the most relevant.

A recession is normally defined as two quarters (six months) of decline in GDP. By carefully poring over the data, the NBER determined that output most probably started to grow again in December 2001. Having determined that, it was easy to date the end of the recession. Indeed, today output is 3.3% above its pre-recession peak of late 2000.

But that does not mean that the economy is performing well. Not at all. In economics, one has to run to stand still. Every year, more than one million additional workers enter America's labor force, and a functioning economy must find jobs for them.

One of the marvels of a market economy is how, every year, year after year, productivity--the amount of output produced by each worker in an hour--increases. During the 1990's, Americans' productivity soared. The new economy may have been hyped, but behind all the hype was something real: the rate of productivity growth increased.

In the 1960's, productivity increased by more than 3% per year, but in the early 1970's, productivity growth suddenly declined, by almost two thirds. Economists puzzle over why. More entrants to the labor force, higher oil prices, and lower investment played a role, but most mysteriously, the pace of technological progress seemed to slow dramatically. For all the vaunted supply-side economics, productivity gains remained weak in the 1980's.

The turnaround came during the Clinton administration, although some investments in technology made in earlier years almost surely played a role. But with productivity now back up to the levels of the 1960's, the economy's annual growth potential increased from around 2.5% to around 3.5%.

The important point is that unless the economy is growing at a fairly strong pace, say around 3%, no jobs will be created. That is what has happened during the Bush years. The economy has grown, but not enough to create new jobs, let alone to create enough new jobs for all the new entrants to the labor force.

Today, unemployment is the highest it has been since 1994, at 6.2% (far higher than the 5.3% of Japan's supposedly moribund economy). But even this number masks the true extent of the problem.

Jobs are so scarce that many Americans have given up looking--the so-called discouraged worker effect. Had labor force participation remained at the level of the late 1990's, the unemployment rate would be far higher. More and more workers have joined the ranks of the long-term unemployed--out of a job so long that they are no longer eligible for unemployment benefits.

Ironically, one early argument used by the Bush administration to resist increases in unemployment benefits was that it would discourage workers from looking for a job. They seemed callously to suggest that the problem of unemployment was with workers too lazy to pound the pavement to find work. But if there are no jobs, looking does no good.

So the NBER's efforts at dating recessions--defined the way they do--are of only limited relevance. After all, there is nothing magic about the number zero. Indeed, the focus on zero can be misleading. The gap between an economy's potential and its actual performance is what matters most.

Assume that America's annual growth potential is 3.5%. In a $10 trillion economy, that means America lost something like $320 billion in 2001, $440 billion in 2002, and, if growth turns out to be even 3% in 2003, $500 billion in 2003. The Bush administration has been setting all kinds of records, such as the largest adverse turnaround in America's fiscal position in the space of just over two years--more than $700 billion per year--from a robust surplus to a yawning deficit. But this waste, in excess of a trillion dollars over three years, sets another record.

So Americans are right not to let declarations about the end of the US recession make them feel good. Weak economic performance--whether it is called a recession or not--smells just as bad. Those responsible for such economic mismanagement should be held accountable.