The Agenda

Why the Slow Recovery?

Hello all — I’m Arpit Gupta. I’m an entering graduate student in Finance at Columbia’s Graduate School of Business this fall. I do some writing for Hello all — I’m Arpit Gupta. I’m an entering graduate student in Finance at Columbia’s Graduate School of Business this fall. I do some writing for Economics 21, and also blog at Calculated Exuberance.

One of the puzzles I’m interested in is the slowness of the economic recovery after the recession in 2008. This was the central question in this controversial set of slides by Robert Lucas. Unlike every recession since World War 2, economic outcome has remained below trend for a prolonged period of time, with no recovery in sight:

We are frequently told that this is just what one can expect after a financial crisis, perhaps by pundits drawing on the empirical work of Carmen Reinhart and Kenneth Rogoff, who have found that recoveries after financial recoveries tend to be weak. This strain of thinking has grown prominent among economic policymakers in the Obama White House, much to the dismay of progressives. For instance, here is President Obama:

Now, the economy is now growing. It’s not growing quite as fast as we would like, because after a financial crisis, typically there’s a bigger drag on the economy for a longer period of time. But it is growing. And over the last year and a half we’ve seen almost 2 million jobs created in the private sector.

Set aside for a moment the question of why recoveries are slower after financial crises. Arguably, such recoveries are slow due to self-induced paralysis on the part of policymakers. Is the basic empirical relationship true at least? Rogoff and Reinhart’s figures were based on estimates taken globally. Are they relevant to the United States?

Economists David Lopez-Salido and Edward Nelson argue against this idea. They begin by noting that while Rogoff and Reinhardt included one example of an American financial crisis in their sample — in 1984 — an alternate chronology would include three separate financial crises; the periods from 1973-1975, 1982-1984, and 1988-1991.

These were not minor crisis periods. In certain respects they reflected a deeper financial crisis than the problems in 2008. For instance, in 1982, concerns over the default risk of Latin American sovereign debt led to a period of financial stress followed by an elevated rate of bank failures. The government intervened in several banks, including Continental Illinois Bank (which faced a bank run). Likewise, the savings and loan failures in the late eighties and early nineties led to the failure of many savings institutions and a large drop in new home construction. The mid-seventies saw bank equity/capital ratios reach post-war lows.

Yet the recovery from the first two crises was brisk. The recovery was somewhat slower in the case of the 1991 recession — the first of the modern “jobless” recessions — yet it is difficult to associate that slowness with financial frictions.

So why is that our modern financial crisis has led to problems that are so much deeper? Again, in terms of the nuts and bolts of lending, in certain ways we are in better shape now than before. As the financial crisis primarily affected the shadow banking system, banks have been left free (at least in theory) to lend as usual. Where the private sector has stepped back, agencies like Fannie/Freddie and the Fed have provided credit for the whole economy. Consumers and firms who require credit do not appear to be facing historically unprecedented difficulties in acquiring financing.  It is difficult to argue, years after the crisis, that financial frictions alone can account for the slow pace of the recovery.

It is true that the financial sector has grown dramatically in size, so that a simple idiosyncratic shock to that part of the economy may lead to oversize effects. Yet here again, many financial firms have regained their previous profits, even as many other parts of the economy remain depressed.

I don’t have an easy explanation for this. The answer to which I am increasingly drawn is that this crisis is much more about unprecedented debt burdens on households rather than problems with banks.

For instance, check out this graph by Justin Wolfers: 

Stagnation in income happened quite a while before the actual “recession” took place. If you look at the spending-based measure, you see large drops roughly around the period in 2008 when you had Lehman Brothers and other financial problems. Yet the income-based measure shows greater stagnation over a longer period. If all you have is spending, you might think that the financial crisis is the real problem. Once you have both measures, you think that we have had serious problems for some time, and have been papering over these problems with debt. Part of what changed in 2008 is that one credit spigot shut down.

This was in fact the basic story for the whole decade. During the early ’00s, consumption grew at about 2.5% annually. This was lower than previous decades, which saw growth rates in consumption around 3-4%. What was really new and different though is that those previous decades saw income grow in proportion to consumption. This decade, income grew almost a full percentage point below consumption. That balance can be accounted for by a drawing down in assets and a rise in debt. Either way, that left saving rates plunging:

Through credit cards, student loans, auto loans, and especially mortgages (and associated equity cash outs) households were able to maintain their standard of living in the face of plunging incomes. It’s not as if the housing bubble years led to an excess of consumption in general — though undoubtedly that happened here and there. The bigger problem was that maintaing our basic trend in consumption growth simply became impossible, possibly because so much of our compensation has been taken in the form of fringe benefits like healthcare, or otherwise spent on other commitment goods like education. More and more of our economy consists of commanding heights service areas like education and healthcare. In these sectors, we face durable problems in generating meaningful productivity advances.

We had to pile on debt to keep up. In the case of mortgages, that debt was backed by housing. Falling house prices and the general economic downturn led to two dramatic changes — not only did households stop using debt to finance consumption they could not afford; but they actually went back in reverse and started de-leveraging (ie, the spike in the savings rate in recent years).

The underlying driver here is a fundamental difficulty in generating productive growth, paired with a difficulty in translating that growth into growth in household income. It’s taken a decade for households to translate that understanding into their savings and consumption behavior, and the economic recovery has been dicey as a result. We really do face a dramatic drop in purchasing power, relative to trend.

I’m not fully convinced by this idea — but I am convinced that Rogoff and Reinhardt are not the last word here. 


The Latest