Stephen Roach (New York)



Forgive me if I throw cold water on the holiest of the sacred cows. Much has been made of the celebrated 3.5% productivity surge in the fourth quarter of 2001. So much that most economists, policy makers, politicians, and journalists have all seized on this seemingly spectacular increase as the smoking gun of the New Economy. The serious productivity scholars have long argued caution in assessing this critical characteristic of economic performance during a boom. A more decisive verdict would have to withstand the test of a downturn. And now it supposedly has. Just when the US economy was widely presumed to be down and out, it has apparently come through with flying colors. If that’s as bad as it gets in a recession, goes the logic, just wait until recovery begins in earnest. Inasmuch as the latest productivity surge occurred during a recession, the New Economy is finally thought to have passed the acid test.

Context is, of course, key -- especially since the latest productivity results haven’t exactly occurred in a vacuum. The productivity metrics of record -- notwithstanding serious measurement shortcomings -- do, indeed, reveal a discernible pick-up in the underlying trend. Average gains in output-per-hour in the nonfarm business sector are estimated to have doubled from 1% over most of the 1970s and 1980s to 2% in the 1990s. Never mind the recent sharp downward revision to gains in the later half of the 1990s. The footprints of the New Economy were unmistakable. That makes the seemingly stellar results of 4Q01 all the more impressive -- yet another testament to America’s new prowess on the productivity front. But before licking your chops in gleeful anticipation of what lies ahead, consider an alternative explanation: The productivity results from the fourth quarter of 2001 may tell us next to nothing. That’s because the US economy failed to conform to the recessionary mindset that most thought was unfolding at the time. It may well be that the recent productivity breakthrough has more to do with this disconnect between expectations and reality than anything else.

This latest twist in the great productivity saga all started with the terrorist attacks of 11 September. As the US economy ground to a virtual standstill in the weeks immediately following this devastating shock, Corporate America hunkered down for the worst. As well it should have. Any doubts over the possibility of recession -- and there were still plenty of them at the time -- were quickly cast aside. Fearing a swift and ever-deepening downturn, businesses cut labor inputs and other costs with a vengeance. Already in the midst of the worst compression in operating earnings for the S&P 500 in some 50 years, corporate managers could ill afford to do anything else in the context of what they perceived to be a shock-battered US economy.

Not surprisingly, the cost cutting that ensued in this climate was aggressive by any standards. Hours worked in the nonfarm business sector were slashed at a 3.7% annual rate in 4Q01 (relative to the third quarter), well in excess of the 2.9% average annualized contraction that occurred in the second and third quarters of 2001. In fact, the reduction in labor input in the final period of last year was the sharpest such plunge for any quarter since the last recession in early 1991. Moreover, reflecting a moderation in worker compensation, unit labor costs -- the broadest measure of the labor context of business production expenses -- fell at a 1.1% annual rate in the final period of 2001. Reflecting this intense wave of cost-cutting, nonfarm payrolls were slashed by 1 million workers in the final three months of the year and the jobless rate rose from 5.0% to 5.8% over that same period. Recession was at hand and the US economy had responded accordingly.

But then something strange happened along the way. Instead of falling sharply in the fourth quarter of 2001 as widely expected, real GDP actually rose. To be sure, the increase was fractional (+0.2% and likely to be revised upward a bit) but it was an increase, nevertheless. The result was an astounding positive output surprise for a recessionary US economy that had just been hit by one of the most lethal shocks in memory. With labor inputs set for a much weaker economy, the positive output shock showed up in the form of an astounding productivity windfall. Yes, that’s exactly the way the math works: Productivity is calculated as output per unit of labor input. The denominator (labor input) was driven down by the expectations of ongoing and ever-deepening recession. The numerator (output) failed to validate that scenario. And presto -- a productivity miracle was borne.

In my humble opinion, this proves next to nothing insofar as trend productivity is concerned. The productivity test that awaited full-blown recession -- essential for the validation of any new trendline -- has yet to occur. So far during this recession, real GDP has recorded a fractional drop of just 0.3% -- all of it concentrated in the third quarter of 2001. That’s, by far, the shortest and mildest recession on record for the post-World War II US economy. By contrast, the typical recession lasts 11 quarters and entails about a 1.5% drop in real GDP from the peak of the business cycle to its subsequent trough. If this recession is now over as most suspect, it side-stepped about 80% of the typical output decline. That’s hardly a stress test of America’s newfound productivity prowess.

What to make of all this? The terrorist attacks of 11 September were a shock to all of us. But they were a far greater shock to the statistical system that generates the productivity results. The business sector’s response was in keeping with widespread expectations of a precipitous decline in the economy. Yet the output side of the economy failed to conform to those expectations -- rising instead of falling. If there’s one lesson from the long and rich history of US productivity experience that bears underscoring in this climate it’s that quarterly fluctuations in output-per-hour are all but meaningless. Production technologies are almost always flexible enough to finesse transitory disparities between inputs and outputs. But that begs the more basic and important question: Has there been a fundamental shift in trend productivity?

I think it is irresponsible to render a verdict either way at this point in time. I would go even further and argue that it is ludicrous to draw great comfort from any productivity resilience so far in this recession. The counter to such a consensus claim is, What recession? The long-awaited stress-test has yet to occur. That’s not to say there can’t or won’t be such a test. The now forgotten double-dip could be just such a learning opportunity. So, too would an anemic recovery (see my 15 February Global Economic Forum dispatch,"The Swoosh"). But until the US economy is truly stress-tested on the downside, I dare say the jury is still out on America’s productivity renaissance.



<center>

<HR>

</center>