Wednesday, September 23, 2009

The 1949 Recession, and why it suggests this recovery won't be like the post 2001 recovery

- by New Deal democrat

Just about everyone seems to think that we live in an era of ever-lengthening "jobless recoveries." If you have followed me long enough, you know that I am usually skeptical of conventional wisdom, and frequently it turns out to be wrong. In that vein, I have become more and more convinced that the incipient Recovery will be unlike the recovery from the 2001 recession. My recent/ongoing examination of leading indicators for job growth as caused me to take a long look at another, similar recession to 2001: the first post-war recession of 1949.

Here is what more typical recessions and recoveries look like graphically:

Notice that both industrial production and real retail sales recover strongly after 3 of the 4 1970s era recessions, and employment follows almost immediately. In the weakest recovery, 1970, industrial production grew just slightly over 2.5% and jobs expanded, but slowly.

Yesterday I noted that in 1949, even though Real Retail Sales grew at a very slight rate throughout the recession, Industrial Production declined. Once both started to grow again, so did Jobs within several months thereafter:

And in 2002, at first Real Retail Sales grew, but Industrial Production did not. Then Industrial Production grew, but Real Retail Sales did not. As soon as both started to grow together, Jobs finally started to increase:

Although not shown on the two graphs above, there is another important series that behaved oddly during both recessions: housing. Housing starts continued to increase throughout the 2001 recession. In 1949, they declined for a few months into the recession, and then began to grow exclusively throughout the rest of the recession and beyond.

It's as if both recessions left ordinary consumers untouched. In fact, that's exactly what happened, both times.

The 1949 recession was discussed extensively in a Federal Reserve paper in 1956 (pdf) by Benjamin Caplan. Here are a few choice excerpts:
The commonly accepted view is that it was an inventory recession. What this means is that the forces which initiated the downturn had their major impact on the accumulation of inventories.
The postwar abnormal expansion in consumers' demand began to level off..... [but] Private fixed investment failed to maintain its rate of expansion and, on the contrary, developed a slight declining tendency. The 1948 drop in the rate of growth of demand at the same time that supply capacity continued to expand was only temporary. It was only a brief pause in the postwar boom. It did not take long to complete the relatively small market readjustments called for, such as price and inventory declines. Because of that fact, the recession turned out to be very mild. (p. 29)
....Preceding the decline in total output, there was thus a weakening in the rate of growth of final sales. When the downturn got under way, final sales did not decline but tended to level out for a while before resuming their upward progress. Consequently, the fluctuations in GNP were primarily in that part of output which does not go to final sales, viz, inventory. (p. 31)
over-all business sales reached their peak in August 1948, for durable manufacturers in December 1948, for nondurable manufacturers in June 1948, and for both types of retailers in December 1948;... [while] total business inventories (book value) increased steadily through 1948, reaching their peak in February.

On the labor front, unemployment in 1948 averaged slightly less than in 1947, about 2.0 million or 3.4 per cent of the labor force compared with 2.1 million or 3.6 per cent of the labor force. At the same time the civilian labor force expanded by over 1.3 million in 1948 compared with 1947. However, in the fourth quarter of 1948, layoffs began to appear in nondurable industries and in some consumer durable industries. (p. 34)

Just as in 1949 traditional producers built too much, and had to reduce inventories, even though consumer demand remained strong, so in 2001 the drastic overbuilding of fiberoptic and other infrastructure to take advantage of the New Internet Era, i.e., excessive capital expenditure by tech businesses, burst dramatically as such businesses as and, among other business "plans", blew up. If you were in one of the affected industries, you suffered terribly. If not, you sailed right through (I am not including in this discussion the exogenous shock of 9/11 which drastically curtailed all business and consumption activity for a brief period thereafter).

In short, consumption remained almost completely unaffected by both downturns. As a result, both were shallow and, in the case of 2001, it was barely a recession at all, as there were only two unconnected quarters of declining economic activity. Real retail sales stagnated (2001) or slowly rose throughout the recession (1949). Unemployment rose only slightly, and almost entirely due to the downturn in the specific productive industries. As the inventory correction ran its course, rising unemployment slowly abated.

Thus it is no surprise that the low point for Real retail sales actually occurred before the recession in 1949, and at the very beginning of the recession in 2001, and rising so slowly during the course of both economic downturns that that it gave rise to the two big outliers of its relationship with payrolls -- a 3 month lag for the onset and 9 month lead over payrolls in the 1949 downturn/recovery, and a 23 month lead over payrolls in the 2002-3 recovery. In essence the point at which employment began to decline and grow was lost within the "noise-to-signal" band of normal monthly variation in Real retail sales.

The contrast with the recent "Great Recession" could not be more stark. This was a consumer retrenchment with a vengeance! Production and employment reacted immediately to the downturn, dropping precipitously. Conversely, the Index of Leading Economic Indicators has similarly rocketed upward in the last few months, with private forecasting group ECRI just reporting that their index hit an all time high -- even higher than the rebound from the 1982 recession that featured 20% interest rates and 10%+ unemployment.

While I have already noted that employers seem to be "hoarding" jobs, waiting to see if the economic growth is sustainable, it is unlikely to go on long if the consumer retrenchment of September 2008 even partly reverses. In short, there is every reason to believe that this recovery will not behave at all like the recovery after the 2001 recession. If it is to approximate a "jobless" recovery, then the model is much more likely to be 1992 in which jobs did start to grow shortly after the recession ended, but not robustly enough to overcome rising unemployment.