73rd International Atlantic Economic Conference

March 28 - 31, 2012 | Istanbul, Turkey

Why are U.S. light vehicle sales still so low?

Saturday, 31 March 2012: 5:25 PM
Maria Otoo, Ph.D. , Industrial Output, Federal Reserve Board, Washington, DC

By the third quarter of 2011, consumer spending on most durable goods had more than recovered from the recent recession-induced slump.  Real personal consumption expenditures (PCE) excluding motor vehicles were about 11 percent above the level in the third quarter of 2007 before the onset of the last recession.  In contrast, real PCE for new motor vehicles was 20 percent below its 2007:Q3 level.  Similarly, unit sales of light vehicles in the third quarter of this year (at an annual rate of 12-1/2 million units) were well off the 16 million unit average sales pace just prior to the recession.  Thus, regardless of whether one looks at units or expenditures, demand for light vehicles remains relatively low more than two years from the end of the last recession.

This paper examines one possible contributor to the current sluggish pace of recovery in the motor vehicle sector.  As occurred in the housing market, it is possible that light vehicle purchases, partly fueled by generous price incentives from auto and truck manufacturers as well as booming real estate values, overshot the desired or target stock of motor vehicles toward the end of the last decade.  This may be particularly true for light trucks.  Over time, the increased durability of vehicles has caused the median age of the vehicle stock to drift higher.  From 2000 through 2004, the medium age of the stock of automobiles rose 6 months.  In contrast, the median age of the stock of light trucks fell about 3 months as consumers snapped up new light trucks at an extraordinary pace.


I use an estimate of the user cost of capital to generate separate estimates of the desired or target stock of automobiles and light trucks.  The difference between the actual and target stock of vehicles is then used in an error correction model of sales.  This will determine if differences between the actual and target stocks play a role in sales and how quickly the two converge.  Data from R. L. Polk on annual vehicle registrations are used to estimate the vehicle stock.  Sales data are from Ward’s Automotive.  In addition data on prices, income, and employment are from the Bureau of Labor Statistics and the Bureau of Economic Analysis.  


Preliminary results suggest that the actual stock overshot the target stock for light trucks but not automobiles and has contributed to weak sales for several years.  Initial estimates using a simple model of light vehicle sales suggests that it may take several additional years to converge.