So, here we are a year into the deflation of the housing bubble. So far so good. Prices are only down marginally. Transactions are down, but not vertiginously. Unsold inventory is up, but not horrendously. Builders are reacting suitably by reducing construction. So, that's all right then...

Perhaps. At any rate, so far, so good. But that is not to imply that we are back to business as usual.

My too extensive experience suggests that there are three kinds of burst bubbles in housing.

The first, unsurprisingly, happens when, whatever else is going on, local/regional employment falls. As housing prices are ultimately supported by wages, they fall too. But it takes some time; usually on the order of a couple years. An example in the SF Bay area is the Peninsula after 1989. Prices in Redwood Shores dropped about 20-25% by 1992-3. They regained their 1989 levels in about1995.

The second type of bubble is where there are problems, but the local market is supported by a stable employment base. Sales volume falls, but prices do not. Unsold properties proliferate, but are eventually cleared from the market as the economy recovers. Berkeley, California exemplified this pattern in all of the national recessions of the 1970's. One of the interesting footnotes from that market, however, was that many sales in that era were predicated on assumed mortgages, often topped up with "seller financing". There was much debate in the era as to whether such prices were representative of market value, as defined for most legal purposes. Analysis of the data suggested that prices were more predictable at the gross level, including assumed financing. There was no discounting technique for seller financing that produced as uniform a result. That is theoretically unsurprising. The market was best informed about the gross price.

The third type of bubble was exemplified by Lake Tahoe after 1979. In spring of 1979 prices in some areas were going up ten to twenty percent per month, but only based on one sale per month. People started comparing Tahoe with Aspen in order to justify the prices. Tahoe is not Aspen. The national economy stumbled and suddenly second homes were found to be discretionary purchases. There were few or no sales in the next two years. The situation wasn't improved by mortgage interest rates rivalling credit cards. That market was not supported by local employment and the homes on the market were not usually purchased as primary residences.

So, what do we learn? Employment and wages are key. As long as the national economy keeps ticking along, with high employment, low unemployment and wage growth, the magnitude of the house price problem will remain fairly small: Berkeley.

Falling employment, rising unemployment, and, absent inflation, static or declining wages: Redwood Shores. Unless you are worried that the US economy is totally hollowed out, that all we have left to sell are "The Emperor's New Clothes", we can probably not worry about the third scenario: Tahoe

Recession and inflation are the wild cards. Recession is coincident with falling employment. But this economic expansion is getting pretty mature. My imagination can conjure a number of ways we could enter the next recession.

Inflation is even more pernicious. Phillips just got his Nobel for his study of the inverse relationship of inflation and unemployment. And because foreclosure almost never happens when the value of the property exceeds the loan amount, don't expect Fannie and its friends to be too hawkish on inflation. But, of course, long term rates, with a greater or lesser lag, reflect inflationary expectations. As housing affordability is immediately hurt by rising mortgage rates, just a little more inflation could protract the housing correction by years.