I was curious how the the composition of durable and nondurable goods consumption changes during recessions. This graph shows the monthly variation in the ratio of nondurable to durable consumption since 1959 :
And, for a better comparison over time, the log of the ratio:
As you can see, there has been a fairly consistent decline in this ratio over time, from about 7 dollars of nondurables per dollar of durables in 1960 to close to 2 dollars in nondurables per dollar of durables today. But there is also variation around the declining trend. Most of the deviation around the trend appears to be due to recessions, but there are also time periods such as the late 70s and the late 80s where the series takes noticeable upward turn outside of recessionary conditions (and in other cases the increase in the ratio appears to lead — as opposed to being caused by — the recession, e.g. in the 69-70 and 73-74 downturns). In most recessions the ratio rises as consumers cut back on durables more than they cut back on nondurables, and the ratio falls once the recession ends (the 2000 recession is an obvious exception).
What has happened in this recession? In the earlier part of the downturn, the ratio rose abruptly (this is the unlogged series). It then fell sharply in August of this year, and increased again in September. The dip in the ratio this August appears to be due to the Cash for Clunkers program:
What will happen after the crisis? If the economy grows as before and as incomes grow along with it, there’s no reason to rule out the possibility that the ratio will continue to decline. So it will be interesting to see if the economy picks up this long-run downward trend again once things return to (the new?) normal. In the meantime, though I’m not quite sure what to make of this ratio, whatever good news might have been taken from its sharp decline in August was surely tempered in September.