Far Too Low for Far Too Long

by JW Mason

So, what caused the economic and financial crisis of 2007-whenever? It’s an open question. Presumably we’ll continue to argue about it until the next crisis (or until the Moon Men invade.) But right now let’s talk about one possible story, or really family of stories: that the root cause of the crisis is that interest rates were too low for too long.

This is often taken to be a conservative view; one of its more prominent exponents has been John Taylor, who complained last week in in the Wall Street Journal that

the Fed has returned to its discretionary, unpredictable ways, and the results are not good. Starting in 2003-05, it held interest rates too low for too long and thereby encouraged excessive risk-taking and the housing boom.

Rortybomb’s favorite conservative Fed president Thomas Hoenig made a similar argument when he left the FOMC last year:

We as a nation have consumed more than we produced now for well over a decade. Having very low rates for an extended period of time encourages us to continue focusing on consumption, but to correct our imbalances, we have to focus on production. … If I thought zero rates would bring jobs, I’d want it forever. But it distorts the economy. In 2003, when we lowered rates and kept them there because unemployment was 6.5 percent — look at the consequences.

More broadly, this view is associated with so-called Austrian Business cycle theory, which holds that macroeconomic instability is fundamentally due to departures of the market interest rate from the natural rate. (As Don Patinkin pointed out, this would better be called Swedish business cycle theory, since it really originated with Wicksell but came to London “via Austria.”) In this view, an artificially low interest rate encourages investment in assets whose returns are lower than the true social rate of discount; when interest rates return to their natural level, investment will be depressed until this excess stock of physical capital is worked off.

But it’s not only conservatives who argue this. One hears something similar from the global imbalances crowd – most famously, of course Bernanke’s “global savings glut” hypothesis that excessively high savings in Asian countries drove world interest rates excessively low. (I am deliberately leaving unanswered the question of what “excessively” means here.) You can find a whole host of prominent economists making similar arguments, including Martin Feldstein, Obstfeld and Rogoff, and Chinn and Frieden. When flows of capital from Asia to the United States (or from northern to southern Europe) pushed down interest rates in the recipient countries, they argue, it was inevitable that the projects funded as a result would be speculative and ultimately wasteful.

You can even find this type of argument among Keynes scholars, like the brilliant Axel Leijonhufvud:

Operating an interest targeting regime keying on the consumer price index (CPI), the Fed was lured into keeping interest rates far too low for far too long. The result was inflation of asset prices combined with a general deterioration of credit quality. This, of course, does not make a Keynesian story. Rather, it is a variation on the Austrian overinvestment (or malinvestment) theme.

In the current crisis, he adds, “The General Theory is not particularly helpful.” (Heresy!) Well, it is certainly true that Keynes wasn’t writing about a situation where the market interest rate was in some sense too low. Indeed he believed that an excessively high rate of interest explained not only crises in rich countries but what would we would today call the underdevelopment of poor ones: “The history of India at all times has provided an example of a country impoverished by a preference for liquidity amounting to so strong a passion that even an enormous and chronic influx of the precious metals has been insufficient to bring down the rate of interest to a level which was compatible with the growth of real wealth.”

One is loath to argue with Leijonhufvud, a far smarter and deeper thinker about the economy than most of us (me certainly included!) can ever hope to be. And yet I can’t shake a nagging feeling that the “far too low for far too long” story doesn’t quite add up.

The first problem is that the different stories, while agreeing that interest rates were in some sense too low, don’t line up with each other. The blame-the-Fed and blame-the-Chinese arguments, while seemingly similar – and even sometimes made by the same people – are actually in tension, if not outright contradictory. After all, if interest rates are set by the Fed, they are not set by international capital flows, and conversely. More precisely, in a world of floating exchange rates and mobile capital – the world we’re normally supposed to live in – expansionary monetary policy should be associated with lower capital inflows, currency depreciation, and a more favorable trade balance. Any macro textbook will tell you that the high interest rates of the early 1980s (supposedly in part due to high federal deficits) were a big factor in the growth of the US trade deficit. So insofar as global imbalances are the issue, looseness at the Fed looks like part of the solution, not part of the problem. There’s a reason why Brazil has called the ultra-loose Fed policy of recent years a form of “currency war.”

There’s also a tension between the Austrians’ version of “far too low for far too long” and Hoenig’s. In the first, after all, the problem with low interest rates is that they result in investment being too high; for the latter, it’s been too low. Since high interest rates are a cost for businesses undertaking investment projects, it’s not at all clear (to me at least) how Hoenig’s version is supposed to work.

Empirically, there are reasons for doubt too. Here is Ed Glaeser, for instance, suggesting that low interest rates can only account for a quarter of the runup of housing prices in the 2000s. More broadly, it is obvious that there have been many times and places where interest rates have been low for prolonged periods, and only some of them have experienced asset bubbles.

But there’s a deeper set of problems here, which it’s not clear that any of the “far too low for far too long” proponents have really grappled with. It’s an article of faith that the private financial system channels society’s investable funds to their best use. That’s why we have a financial system! So if a big increase in funds available for investment – either due to an inflow of foreign savings, or the creation of liquidity by the banking system abetted by the Federal Reserve – flow to activities that are socially useless or worse, what does that tell us?

It seems to me there are only two possible answers. Either we have already invested in everything worth investing in, or the financial system is not doing its job. I cannot see how higher interest rates are an appropriate response in either case.

Let me spell out the thought.

Many of the global imbalancers combine their proposals for less saving in the Asian countries with calls for higher savings in the US. That sounds very reasonable. But on closer examination, there’s something very odd about it. Because if it was excess saving that caused the crisis, why on earth would we want more of it?

Obstfeld and Rogoff, for instance, argue both that the underlying cause of the crisis was the capital inflows from Asia, and that the best outcome would be for the US government to reduce its borrowing. I’m sorry, but this makes no sense. From a macroeconomic standpoint, a reduction in US government borrowing by $100 billion, and a Chinese purchase of $100 billion in Treasury bonds, are identical. Both leave the US private sector holding $100 billion less of treasuries, and both – if you believe the crowding-out stories that are the whole basis, normally, of supporting balanced government budgets – should increase its appetite for private debt by exactly the same amount. In general, the benefit of a more favorable fiscal balance is supposed to be a greater supply of savings available to the private sector, resulting in lower interest rates and higher investment. But if lower interest rates would only lead to asset bubbles, what is the argument for reducing the fiscal deficit? There is no reason to expect the financial system to be any more successful in channeling the savings made available via lower public deficits into productive investment, than it is supposed to have been in channeling the inflow of foreign savings.

I don’t think you can argue that a “savings glut” or low interest rates caused the crisis, and then go on arguing in other, longer-run contexts that higher public and/or private savings are desirable.

People making the argument that, in modern conditions, very low interest rates can only call forth speculative or wasteful investment, have implicitly accepted Keynes’ view that there is finite limit on the capital society requires, so that returns on private investment will fall toward zero as this limit is approached. But they haven’t followed him to the logical conclusion that in such a savings-abundant world, consumption and public expenditure should rise secularly until private investment disappears entirely.

Of course there’s an alternative view, which is that socially useful investment opportunities are far from exhausted, and savings are still scarce, but that the private financial system is no longer adequate for matching the latter with the former. If we were going to go this route, the first step would be to take seriously the idea that there is no “the interest rate,” there are various interest rates, and various degrees of access to credit, and they often don’t move together. And the relationship between the policy rate controlled by the central bank, and those various market rates, depends on the specific institutional and regulatory context.

Has everyone read Ben Bernanke’s classic article, Inside the Black Box? One of the key points he makes there is that in a world where reserve requirements don’t bind, one of the main channels by which monetary policy affects the real economy is via asset prices. Descriptively, I think that’s clearly right; but it’s a pretty crude instrument for steering aggregate activity, especially when you consider the positive feedback that tends to happen in asset markets. What if it turns out that the interest rate compatible with full employment is systematically lower than the interest rate compatible with stable asset prices? Does that mean we have to accept mass unemployment forever? That would seem to be the implication of “far too low for far too long.”

The alternative, of course, means more active management of credit. It means accepting that if we think government has a responsibility to maintain macroeconomic stability, a single policy instrument will not suffice. It means abandoning the conventional wisdom on macropolicy of the past three decades: Leave it to the central bank. It means that it’s not enough to set “the” interest rate, some policymaker will have to decide on the different interest rates for different kinds of borrowers, just like in the old days of European social democracy. (And just like in the old days, we may find that making this effective requires rather stringent limits on cross border financial flows.)

I don’t say this is true; we can debate whether industrial policy is a necessity or not. But I do say that this, or the even less palatable previous alternative, is a logical consequence of the view that the crisis was caused by excessively low interest rates.