Correcting Brad DeLong on the Housing Bubble by Dean Baker

I see that Brad has a post saying that the economy was adjusting nicely to the bursting of the housing bubble until the financial crisis set in. He notes that housing construction fell by 2.5 percentage points of GDP between 2005 and 2008. This was replaced by an increase in gross exports of 2.0 pp of GDP and increase in equipment investment of 0.5 pp. Everything was moving along nicely until the financial crisis in 2008.

I see things a bit differently. First, gross exports don't create jobs, net exports do. When we move an auto assembly plant from Ohio to Mexico, we are not creating additional jobs with the car parts exported to Mexico. That's intro textbook stuff. If we look at the net export picture, the gain is only about 1 pp of GDP. Furthermore it is hard to see the improvement in the trade picture having gone very much further without a further decline in the dollar. (That was a possibility, but far from a certainty -- depends on policy decisions elsewhere.)

The rest of the gap was made up by a surge in non-residential construction (can you say bubble?), which rose by more than 33 percent as a share of GDP, or more than 1 pp of GDP. This boom led to considerable overbuilding in retail, office space and most other categories of non-residential construction. Assuming the burst of spending in non-residential construction was another bubble, then the portion of the demand gap filled through this channel was destined to be temporary. It was inevitable that this bubble would also burst and we would need something else to make up the hole in demand.

The other factor in the mix is the drop off in consumption. Savings rates had been driven to nearly zero by the wealth created by the housing bubble. It seems to me inevitable that consumption would fall in response to the disappearance of this wealth. The financial crisis gave us a Wily E. Coyote moment where everyone stopped spending at the same time, but I would argue that this just brought the decline in spending forward in time.

The savings rate remains much higher today than at the peak of the bubble, although still low by historic standards. (It's currently around 4.0 percent, the pre-bubble average was over 8.0 percent.) We have two alternative hypotheses here. I gather Brad would say that people are spending at a lower rate because they are still freaked out by the financial crisis. I would argue that they are spending at a lower rate for the same reason that homeless people don't spend, they don't have the money.

Homeowners are down $8 trillion in housing equity as a result of the crash. I would expect that loss of wealth to have a substantial impact on their spending. I gather Brad does not.

This is what bothers me about his graph. It leaves out the loss of the wealth effect. The popping of the stock market bubble in 2000-2001 wasn't as damaging because it didn't have the same loss of demand related to the wealth effect.