Stephen Roach (New York)


The incoming calls were dripping with sarcasm today. The GDP rose instead of fell, rounding out a slew of data that smacked of economic recovery."Some double dip," a friend snorted. The e-mail traffic was even more pointed. My favorite came with the subject header,"What if they gave a recession party and no one came?"
Sure, the number surprised me. We, along with everyone else, were looking for a 4Q01 decline of a little more than 1%. While an anemic +0.2% increase is hardly much to write home about, it’s an example, nevertheless, of one of the most painful misses of any forecaster -- getting the sign wrong. The burden of proof had clearly shifted to the few of us who remain in the recession camp. Later in the day, the Fed added insult to injury. While there was no surprise on the policy front -- especially after Chairman Greenspan’s recent mea culpa -- it was more what the central bank said. There was none of the classic ambiguity of Fedspeak in the 30 January policy statement. Instead, the FOMC proclaimed that"the outlook for economic recovery has become more promising." The central bank was urging financial markets to look beyond recession.
I continue to believe that any such optimism is premature. Forecast miss, or not, I remain firmly in the double-dip camp. Indeed, as I pore over my spreadsheets, I see double-dip written all over the fourth quarter data. I believe the fractional increase reflects the confluence of several one-off developments that are unlikely to be repeated in the months ahead -- in essence, a classic set-up for the demand relapse that has long held the key to the second dip.
The American consumer stands out as the most obvious case in point. The 5.4% annualized surge in real personal consumption expenditures in 4Q01 says it all: Never before have consumers spent with such a vengeance in the depths of recession. In the 28 quarters of the past six recessions, real consumption growth averaged a scant +0.5%. In only two of those quarters did consumption growth come in at 3.5% or greater -- 2Q60 (+5.1%) and 3Q70 (+3.5%). And those spending bursts borrowed from the immediate future; they were both followed by declines in the subsequent quarter that averaged -1.4%. The lesson is clear: With jobs and income under pressure, paybacks are the norm in the aftermath of mid-recession consumption spurts.
There’s no dark secret behind last quarter’s consumption surge. Fully 94% of the gain was concentrated in durable goods, and over 70% of that increase is attributable to a spike in motor vehicles and parts. Even in recession, consumers love a bargain; zero-interest-rate vehicle financing apparently was too tempting to resist. This surge in durable goods consumption is without precedent in the annals of the past six recessions. Over the 28 quarters of those recessions, durable goods spending actually declined by 5%, on average. In only 11 of those 28 quarters did durable goods consumption increase, and the largest of those gains was 14.6% in 1Q82 -- less than half the increase just recorded. The"lumpiness" of durable goods spending patterns is one of the more enduring features of consumer behavior. Big-ticket items don’t get bought every day, especially the family car. A surge of the magnitude just recorded has undoubtedly borrowed from sales that would have otherwise occurred in the first half of 2002, in my view. That points to a fairly prompt payback -- an outcome that could well be decisive in satisfying the demand-relapse condition of the fabled double dip.
There were other one-off factors in the 4Q01 GDP report that seem likely to go the other way in the first half of this year. A 9.2% annualized increase in total government spending -- federal plus state and local, combined -- is particularly suspicious in that regard. Special factors associated with the 11 September terrorist attacks were clearly at work, especially heightened defense spending traceable to the war in Afghanistan. But there was also an apparent weather-related pop in state and local construction activity, which surged at an unsustainable 35% annual rate. Unseasonably warm weather probably also played a role in propping up private construction activity in 4Q01; although such activity fell at a 17.5%% annual rate last quarter -- residential and nonresidential, combined -- there is good reason to believe that it would have been even weaker were it not for the unseasonably warm weather. As the weather reverts to seasonal norms, a payback effect should also be apparent in the seasonally adjusted construction data.
The double-dip call, however, is not premised solely on a statistical analysis of the national income accounts. There is, in fact, a much deeper meaning to all this. This recession -- especially if it’s now over, as most suspect -- has done next to nothing to purge America of the excesses that built up in the Roaring 1990s. The American consumer remains overly indebted and saving-short; the personal saving rate fell back to 0.5% in 4Q01, and there is good reason to believe it was even lower than that at the end of the quarter. Corporate America continues to be plagued by bloated costs -- from both capacity and workers; the capital spending share of GDP has now fallen from a cycle high of 13.2% in 4Q00 to 11.6% in 4Q91, completing only about half the adjustment that has occurred in past secular downturns in business fixed investment. And America’s current-account deficit remains at about 4% of GDP -- a near-record shortfall and in sharp contrast to the near balance that typically occurs in recession.
Such profound and persistent excesses are simply not conducive to sustained economic recovery, in my view. They leave the US economy bucking powerful headwinds in the aftermath of any inventory-related pop to activity that may be occurring in the current period. In short, I have a hard time believing that a sustained cyclical lift-off can occur with a zero saving rate, a 4% current-account deficit, and a lingering overhang of excess capacity. Given the likelihood of a demand relapse, these lingering structural excesses provide a seemingly classic set-up for a double dip.
How this plays out on a quarter-to-quarter basis is anyone’s guess. With business expectations now shifting into a recovery mindset, a lifting of production to reduce the rate of inventory burn seems quite likely in the current quarter. That would mark two consecutive quarters of positive GDP growth, assuming that 4Q01 is not revised downward. Such an outcome is not without precedent in the midst of recession. The downturn of 1969-70 had an interlude of two consecutive quarters of positive GDP growth in the middle quarters of 1970 before the second dip hit at the end of that same year. I still wouldn’t be surprised to see a similar outcome in the spring quarter of 2002. The double-dip call is now totally out of favor again. But as I see it, the case remains just as compelling as ever -- notwithstanding last quarter’s GDP surprise.



<center>

<HR>

</center>