Wednesday, January 5, 2011

Bad News from the Housing Sector?

Copyright, The New York Times Company

A few economists are contending that our housing market is now in a “double dip,” based in part on last week’s report of housing price indexes for September and October that were lower than they were during the summer. In my opinion, the data on housing prices and construction do not show any significant housing market change during the second half of 2010.

When connecting the housing sector with the wider economy, three different measures of housing prices are helpful: inflation-adjusted housing prices, inflation-unadjusted housing prices and cost-adjusted housing prices.

Inflation-adjusted housing prices tell us how much the prices of homes have changed relative to the prices of other consumer goods. If, for example, we want to know whether demand for housing these days is any different than it was before the housing bubble, it helps to check whether, from the 1990s through 2010, housing prices failed to increase as much as other prices have.

In this case I look at a housing price index that has been normalized by a consumer price index.

Inflation adjustments are not appropriate for the purposes of analyzing foreclosures – a big drag on our economy – because the mortgage principal that pulls homeowners “under water” is not adjusted for inflation either. If unadjusted housing prices increase, even if more slowly than other consumer prices, that helps homeowners swim out of the water.

For this purpose, I look at an index of the dollar value of housing properties, without any adjustment for inflation.

For the purposes of understanding construction activity, it helps to know whether housing prices have increased more than the costs of building materials. The more that housing prices increase beyond the cost of materials, the more value that can be created by home construction activity.

For this purpose, I look at an index of housing prices that has been normalized by an index of building costs.

It turns out that practice is messier than theory, because there are so many different houses in America and many different price trends. In practice, it matters which housing price index is used, regardless of which inflation or cost adjustment is used.

The Case-Shiller repeat sales index is one such index of existing homes. The Federal Housing Finance Agency has another index of existing homes (and there are others, as well). The Census Bureau has a quality-adjusted index of new home prices.

Chart 1 displays the three aforementioned home price indexes, measured quarterly without any inflation adjustment. The Case-Shiller index for the third quarter of 2010 (the first quarter without the government’s home buyer tax credit) was essentially the same as in the previous quarter. The other two indexes show slight decreases over the same time period, although well within the range of ups and downs over the previous six quarters.

By themselves, these data suggest that homeowners did not go significantly deeper under water in the third quarter and that the housing market trends were not dramatically different in the third quarter than in previous quarters.


Chart 2 displays the same three indexes, adjusted by the implicit price deflator for consumer spending. Because inflation has been low recently, it shows a similar pattern to Chart 1. By themselves, these series show no dramatic change in housing demand over the most recent quarter.


Chart 3 displays the same three indexes, adjusted by the producer price index for home building materials. Deflated this way, the Case-Shiller index actually shows a housing price increase from the second to the third quarter. That’s because building costs peaked in May and have been lower since then.

Without home prices falling by this measure, we do not expect construction activity to be lower than it was during 2009 (but, unsurprisingly, lower than it was during the short rush to sell homes before the tax credit expired).


You may notice that various housing price indexes disagree, and our most recent data is still three months old. Yet another approach is to look at home construction activity. Chart 4 displays monthly home construction activity through November 2010, measured as the number of housing permits, housing starts, homes under construction and homes completing construction.


Permits and starts are particularly interesting, because homes take time to build and we presume that many builders are looking ahead to the prices homes will command in the future, when the construction project is complete. Those series were actually higher in November 2010 than they were for several months before.

Predicting the future is difficult, but the price and construction data so far do not seem to suggest that home values will be significantly different this year than they were in 2010.

No comments: