America's broke recovery
ONE of the ideas for why America's recovery has been relatively slow and jobless refers to the fact that the country experienced a balance-sheet recession. In the years prior to the crisis, households accumulated a lot of debt, which was offset by rising asset prices. Those asset prices then collapsed and many households are now desperately attempting to pay down their debts. Because they're heavily indebted, efforts to spark a recovery by encouraging household spending or residential investments are likely to go nowhere; people are simply too broke.In a new San Francisco Fed Economic Letter, Atif Mian and Amir Sufi investigate this hypothesis. They divide counties into deciles based on the pre-crisis increase in indebtedness and then analyse differences in behaviour in counties in the top and bottom deciles. Their results are striking. Residential investment never fell much in low-debt counties while it plummeted during the crisis in high-debt counties. Auto sales, as you can see at right, sprang back in low-debt counties but remain low in high-debt counties. Employment has followed a similar path.The implication is that there is a direct connection between leveraged households and slow recovery. Indebted individuals aren't spending or investing, and so firms aren't investing or hiring, and so banks aren't lending.To take this further, one could say that the problem is a shortage of demand, and it's therefore necessary for fiscal authorities to step in with a spending boost:If the main problems facing businesses relate to depressed consumer demand due to a household sector weighed down by debt, investment tax subsidies and lower interest rates may have a limited effect on business investment and employment growth.Firms need customers and the government must become the customer of last resort. I think this is a reasonable hypothesis, but I also want to return to the Robert Hall view of the recovery that I mentioned two weeks ago:Mr Hall constructed a model, some of which he presented in the session and some of which came out later in his presidential lecture, in which the crisis gives rise to "financial frictions". Lenders must then be induced to provide additional credit through reductions in the real interest rate. But, he pointed out, interest rates are constrained by the zero lower bound. In his model, it might take a real interest rate of something like -2.5% to clear the economy. But obviously the Fed is constrained once nominal rates hit zero, and so the economy returns to its trend growth rate but never recovers the ground lost during the financial shock.Keep that in mind as you read the concluding paragraph from the Fed letter;The evidence is more consistent with the view that problems related to household balance sheets and house prices are the primary culprits of the weak economic recovery. King (1994) provides a detailed discussion of how differences in the marginal propensity to consume between borrowing and lending households can generate an aggregate downturn in an economy with high household leverage. This idea goes back to at least Irving Fisher’s debt deflation hypothesis (1933) and has found empirical support in several studies (Mishkin 1978, King 1994, Olney 1999, Eichengreen and Mitchener 2003, Glick and Lansing 2010, and Mian and Sufi 2010). Our view is that the depth and length of the current recession relative to previous recessions is closely linked to the tremendous rise in household debt that preceded it.Is there not a real interest rate that would induce lending households to become borrowing households? At a sufficiently negative real interest rate, wouldn't some households decide that now is a good time to buy a new home or invest in new appliances or get that new computer? Isn't there some real interest rate that would convince firms to update their plant and equipment? Mr Hall says that there is such an interest rate, but that it's negative. The way to get there, then, is to raise inflation, which obviously addresses the debt deflation concerns.A criticism of this view is that it sees the solution to indebtedness as more indebtedness, but that's not quite right. Debt is debt, and for every borrower there is a lender. Those lenders were savers. Now it seems that those who borrowed heavily want to become savers, and if the savers don't become what economists oh-so-elegantly call dissavers then the economy develops a demand hole. The trick is to find the real interest rate at which those with perfectly healthy balance sheets decide to put their savings to use.A lot of people are uncomfortable with this rather mechanistic view of expansionary policy, but I don't know that it's not telling us something important.