Surprises in Store for Economists

The commerce department reported that retail sales in May were down by 1.2% from April. This surprised most economists who had expected a modest increase. The media were filled with accounts of economists trying to explain why consumers were still reluctant to open up their wallets and spend in a big way. It would have been much more interesting to hear accounts of why economists were surprised.

There is always a large random element in month-to-month movements in retail sales or any other economic variable. Therefore no one is ever going to be able to explain these changes with any precision. (The data are also subject to large revisions, so it is entirely possible that revised data will look very different from the report released last week (pdf).)

Nonetheless, there is little basis for the surprise shown by so many economic analysts. With few exceptions these analysts failed to see the $8tn housing bubble, the collapse of which sank the economy. Remarkably, even now they apparently cannot understand its importance.

To put it as simply as possible (so even an economist can understand it), the housing bubble was driving the economy in the period prior to its collapse, beginning in 2007. It drove the economy in two ways. The run up in house prices led to a building boom. Residential construction, which is typically less than 4% of GDP, rose to more than 6%, creating more than $300bn in additional annual demand. A bubble in non-residential real estate added perhaps another $150bn to annual demand.

The bubble also drove the economy through the effect of housing wealth on consumption. Economists usually estimate that $1 of additional housing wealth increases annual consumption by between 5-7 cents. This implies that the $8tn of housing bubble wealth would lead increase consumption by $400bn to $560bn a year.

With most of the bubble wealth eliminated by the collapse of house prices over the last three years, we should expect a sharp drop in consumption. Furthermore, stock prices have lost a bit less than a third of their value (around $6tn), which we should expect to cause a further decline in consumption. With the stock wealth effect estimated at 3-4 cents on the dollar, the decline in stock prices should have reduced annual consumption by $180bn to $240bn. In total we should expect to see annual consumption have dropped by between $600bn and $900bn as a result of the loss of housing and stock wealth.

This is all very simple arithmetic and basic economics. Consumption had been driven by the housing bubble prior to the recession. Now that the bubble has collapsed and trillions of dollars of wealth has disappeared we should expect much lower levels of consumption. To flip this around, the savings rate, which had averaged more than 8% in the decades prior to the 90s, fell to near zero in the years leading up to the recession. Now that the bubble has collapsed, we should expect consumption to fall and saving to return back to its normal level. In fact, it might even go higher since the huge cohort of baby boomers is now mostly in their 50s and most have almost nothing saved for retirement. This might lead savings rates to go above their long-term average.

There seems very little room for argument in this story. The existence and housing and stock wealth effects are among the least questioned propositions in economics. Nor is there too much dispute about their size. How could any economist see the collapse of an $8tn housing bubble and the destruction of more than $6tn in stock wealth and not expect to see a substantial decline in consumption?

Yet, we have dozens of economists being cited in newspapers and broadcast news, all saying that they are surprised by weak consumption. If anything the surprise should be that consumption is still as strong as it is. The saving rate is still near 4%, far below its historic average. Why on earth would any economist expect it to go still lower?

The reason that consumers are not spending more money has nothing to do with attitudes. The reason that most consumers aren't spending is the same reason that homeless people don't spend much money: they don't have any.

Economists used to be able to understand basic economic concepts. Apparently, most have lost this ability. As a result we are likely to see many more surprised economists and much proposed in the way of very bad economic policy.