Do Shipping Volumes Signal an End of the Recession?
The advance estimate of 3.5 percent for GDP growth in the third quarter of 2009 is welcome news, and it suggests that the longest recession since the Great Depression and the deepest since the 1950s is likely to have ended at some point around the middle of the year. This isn’t to say that the widespread pain experienced by households and firms is over, just that the economy has at least stopped contracting and is starting to grow. It is fairly widely agreed that it will take months if not years for output and employment to return to their former peak levels. Of course, because the NBER Business Cycle Dating Committee waits until the recovery is no longer in doubt, it could be 6 to 18 months before it assigns an official date for the end of the latest recession, in part because initial data are subject to revision.
While GDP is the best measure of overall economic activity, it is available only at a quarterly frequency (although the previous quarter’s estimate is updated every month as more data become available). This can make for a long wait to be certain about turning points in the economy. Employment estimates, another major series considered by the dating committee, come out monthly, but this series tends to be a lagging indicator. In light of this, economic observers try to glean evidence of turning points from data series that come out more frequently and with less of a lag. Transportation data are good candidates, as many series are published monthly, relatively soon after the close of the month. Equally important for this purpose is the fact that they should be highly correlated with economic activity.
First consider manufacturers’ shipments. Manufacturing constitutes the bulk of industrial production, and industrial production is an economic activity explicitly examined by the dating committee. Moreover, manufacturers’ shipments should be closely in sync with business cycles. A look at the most recent data for manufacturers’ shipments shows that the overall series and its components have all moved off their recent lows, a trend that began last May or June. However, they have yet to show robust growth. This pattern looks similar to the way the 2001 recession ended.
But while manufacturing accounts for most industrial production, it accounts for only about 13 percent of GDP, and its employment share is under 10 percent. For those reasons, it may be too narrow a measure to serve as a reliable indicator of overall economic activity. Some broader measures, based on transportation activity, are available from the Bureau of Transportation Statistics. It produces two data series of transportation services: one for freight and one for passengers.
The most recent values for freight and passenger transportation services’ 12-month percentage change are still declining, but since June they have moved up from their recent lows.
Unfortunately, the link between these series and overall economic activity may not be as tight as one would hope. In the 1991 recession, the growth rate for freight declined before the recession officially started, was roughly flat through the recession, and began to grow again at the start of the recovery. The passenger series was even more problematic in that recession. Growth in passenger travel slowed at the onset of the recession but then plunged after the 9/11 terrorist attacks. These shocks put this series out of sync with the overall economy at the end of the last recession.
A series that goes back further—and that seems more closely tuned to the overall business cycle—is the Federal Highway Administration’s estimate of vehicle miles of travel. This monthly series has the advantage of being dependent on a broader array of economic activity. People drive for a variety of reasons (to work, to shop, for vacations) in addition to delivering goods. While this measure does not always turn sharply negative during a recession, the growth rate does remain depressed relative to what it was just prior to a downturn. This occurs in part because there appears to be a modest long-run decline in the growth rate of vehicle miles, and so this trend must be accounted for in looking for turning points in the economy.
Finally, on a related note, we look at a transportation measure with an application to more local economic activity. If you have a view of the Great Lakes, it’s not your imagination, there really is less ore boat traffic out there. Water is the least expensive way to ship bulk items, and the Great Lake states have long benefited from this advantage. Iron ore, limestone, and coal, all key ingredients in the production of iron and steel, comprise almost 95 percent of Great Lakes cargo by tonnage. While Great Lakes shipping volumes have stopped shrinking as rapidly as they had been in the beginning of the year (when total shipments were down 80 percent year-over-year), they are still down nearly 40 percent. With iron and steel mill production down nearly 57 percent year-over-year, this should not be too surprising. Great Lakes shipping will recover only once these mills return to a higher production rate.