Since the Second World War, real GDP in the United States has grown, on average, at a yearly rate of 3.2 percent. This is what economists call trend growth. Whenever the U.S. economy is faced with a recession and grows below trend for a while, a recovery period typically follows in which growth is above trend. In a previous Trends article I pointed out that the current recovery and the previous one are weak in the context of past recessions. As the figure below illustrates, in these two instances, unlike in previous recoveries, GDP grew either at or below trend for the year following the trough.
The latest numbers from the National Income and Product Accounts suggest that the state of the recovery is not as bad as one might think at first glance. Looking at the behavior of the different GDP components reveals some short-term effects that are likely to go away in the third quarter. While overall GDP grew at a rate of only 1.6 percent, gross domestic purchases, a series which subtracts exports from GDP and adds imports, grew at the healthy pace of 4.9 percent. This means net exports robbed GDP of 3.3 percentage growth points. In fact, imports alone grew at a yearly equivalent rate of 32.4 percent, a clearly unsustainable rate that no doubt owes much to the broad appreciation in the U.S. dollar vis-à-vis the currencies of major U.S. trading partners.
Even if things do improve slightly in the near future, we would still be growing along with the trend and not above it as in most recoveries. The reasons for the sluggish pace of the two latest recoveries are to be found in the differences between these two recessions and previous ones. While many factors may qualify, I will focus on the effect of the downturns households balance sheets.
The chart below shows the behavior of households (and nonprofit organizations) net worth in the last six recessions. It is apparent that in the last two the damage to households balance sheets was both deeper with and more protracted than in the previous episodes. What was behind the drop in the latest recession? During this period, liabilities were roughly constant, so the drop happened because of declines in asset values caused by the real-estate collapse and the subsequent depreciation in financial assets. In the 2000 recession the drop was due to the stock market collapse. In contrast, in the twin recessions of the early 1980s, net worth never decreased, and in the early 1990s it dropped only about 2 percent.
The drops in household net worth help explain the protracted recoveries after the last two recessions. Personal consumption expenditures are the single biggest component of GDP at around 70 percent. If there is to be a solid recovery, consumption needs to increase at a substantially higher rate than the 1.7 percent it has averaged over the last year. But households are not going to start consuming at substantially higher rates until they have fixed their balance sheet problems. This is why the savings rate has been so high lately: Households are working hard at improving their wealth to income ratios at the expense of consumption. In previous recessions, since net worth did not fall by a substantial amount, this was not a problem. As incomes started growing again, consumption followed suit. Right now, an important part of that income growth is being channeled to savings. As the chart above illustrates, net worth is still well below prerecession levels and, barring an increase in asset prices (real-estate prices or stock market prices), the only way to increase it is by saving more and consuming less, further delaying the recovery.
Finally, note that this figure hides a lot of heterogeneity in terms of asset holdings across households. At the peak that preceded the most recent recession, real estate represented roughly a third of total household assets, while most of the remainder was in the form of other financial assets (stocks, bonds and related derivatives). Households at the very top of the income scale hold a disproportionate amount of wealth in the form of these financial assets, which in turn means that the vast majority of households have most of their wealth in the form of housing. Since real-estate-related assets declined by 30 percent from peak to trough (compared to a 22 percent decline in other financial assets), the decline shown in the graph, as large as it seems, actually underestimates the losses most house