This month’s release of new home sales numbers (for April) once again brought hope for the housing market, but that hope faded a day later with the release of existing home sales numbers. It might at first seem surprising that new and existing home numbers would move in different directions, since one might expect new and existing homes to be nearly perfect substitutes for one another. But monthly numbers can be pretty volatile, and it might not be that unusual for the two series to move in opposite directions in any given month. However, if we smooth the two series by creating three-month moving averages for each, we might expect them to be more strongly correlated.
But it turns out that smoothing monthly fluctuations does not increase the correlation between the two series. Existing single-family home sales were actually fairly stable over the eight months ending in April, while median prices were falling in the period’s final months. Throughout the same period, new single-family home sales continued to fall, but their median prices rose in the period's final months. Moreover, new home sales were down about 32 percent from their peak in mid-2005, while existing home sales dropped only about 12.5 percent from their peak around the same time.
These data may lead us to question whether new and existing homes really are close substitutes for one another. In several aspects, they are quite dissimilar. For instance, there are many quality issues: New homes have modern plumbing, electrical wiring, and insulation, whereas existing homes are likely to have more dated systems. There is also a location factor: New homes generally are built much farther from city centers, while existing homes are closer in.
Furthermore, as an investment, one would expect new home sales to be more strongly affected by the business cycle than existing home sales. The construction and sale of a new home represents an investment at the macro level and so has an important impact on GDP. Over the past 20 years, residential investment has contributed only 0.1 percent on average to real GDP growth; but over the past four quarters, it subtracted 1.0 percent from real GDP growth. In contrast, selling an existing home, which simply transfers ownership of existing capital, has a very small impact on GDP (consisting of the realtor’s commission). For all these reasons, existing and new home sales might not be highly correlated in the medium term.
The housing market can also have an indirect influence on GDP. Since the typical American household holds a very large proportion of its wealth in the form of housing capital, a change in home prices creates a wealth effect, which can in turn affect households’ consumption decisions. However, we have not yet seen any sign that the decrease in wealth caused by the weak housing market is spilling over into reduced consumption (that is, into a lower growth rate of consumption). Of course, there is no guarantee that continued weakness in the housing market will not affect consumption spending in the future.