Industrial Production and Housing Starts

Industrial production and housing starts were reported today. It has been well covered elsewhere, but I though I would make a couple of comments.

Compared to other cycles this recovery in industrial production continues to be moderate.

It is stronger than in the weak recoveries, but compared to the depth of the downturns the rebound is quite weak. The good point is that in the early stage of a recovery industrial production is driven largely by inventory rebuilding. But we have probably passed that point in the cycle as the economy shifts from the recovery stage to the expansion stage. This means that we are now seeing quite strong industrial production that is driven by changes in final demand rather than by inventory restocking. This implies that good growth in industrial production is likely to continue in contrast to previous expansions when industrial production growth flattened out after inventory restocking ended.

The other point in the report was that capacity utilization was rising. Normally rising capacity utilization is an important driver of business capital spending and is a good omen for continued growth. But that optimistic premise should be tempered by the point that one of the reasons capacity utilization is rising is that industrial capacity is actually contracting.

It is down -1.3% from a year ago, the largest contraction on record.

In contrast to growth in industrial production housing starts actually fell from 659,000 to 593,000. This reflects the major difference between the two economic sectors. Industrial production is being driven by a rebound in final demand while final demand for housing is weak. Moreover, this weakness in housing demand reflects the over a decade of over-production in housing that built excess supply that still has to be worked off.

The fundamental driving force behind housing demand is household formation. In the short run other factors enter the picture. But no matter how many bells and whistles you add to the model, household formation will remain the most important factor. It is not a good determine of month to month fluctuations in the housing market but it is the key factor driving long term demand. Household formation is essentially a function of young people leaving home and setting up independent house keeping. That is one reason it is surprisingly cyclical as in hard times young people either remain or return home. But the long run trend is clear. After the baby boomers leaving home and setting up housekeeping lifted household formation to over 2 million annually in the 1970s and 1980s household formation has fallen to under 1.5 million annually. Moreover, it should remain at that low level unless we have some fantastic rebound in immigration.

This is easy to see if you look at a smoothed series of household formation and housing starts. In the 1970s and 1980s annual housing starts averaged over 2 million. But this was accompanied by household formations of over 2 million annually. So over that period supply and demand were in rough balance and despite the highly cyclical nature of both economic series, significant long term imbalances never developed. But look at what has happened since the late 1990s. Housing starts have significantly outpaced household formation creating a large supply of excess housing that will have to be worked out of the system. But with household formation now expected to remain well below 1.5 million this implies an extended period of housing starts remaining at or near their current low levels rather than the historic pattern of strong rebounds. This cycle the pent-up demand for housing is negative.