Thirteen Ways of Looking at Money

I taught a class last semester on alternative theories of money, drawing heavily on Money and Things, the book I am working on with Arjun Jayadev. It was one of the best classes I’ve ever taught in terms of the quality of the discussions. John Jay MA students are always great, but this group was really exceptional. It was a a privilege to have such  thoughtful and wide-ranging conversations, with such an enthusiastic and engaged group of (mostly) young people. 

The class syllabus is here. A number of the readings were draft chapters from the book. I am not posting these publicly, but if you are interested you can contact me and I’ll be happy to share.

In this post, I want to sketch out some of the puzzles and questions around money — my own version of what makes money difficult. Many of these were explicit topics during the semester, others were in the background. I wouldn’t claim this is a comprehensive list, but I think most debates around money fall somewhere on here.

The first problem is defining the topic. When we talk about “money” as a distinct set of questions in economics, what are we distinguishing from what? In particular, are finance, credit and interest on the money side of the line?  Given that aggregate demand is, presumably, defined in terms of desired  monetary expenditure, are demand and its effects a subset of questions around money? The main classification codes for economics articles include a category for “Macroeconomics and Monetary Economics”; this suggests an affirmative answer, at least in the mainstream imagination. Do we agree?

Put another way, a focus on money in economic analysis means something quite different if the implied alternative is an imagined world of barter, versus if it’s a a broader range of financial arrangements. In the first case, talking about money involves a broadening of perspective, in the second case a narrowing of it. If someone says, “we have to think about the business cycle in terms of money” are they rejecting Real Business Cycle approaches (a good thing, in my book) or are they telling us to focus on M2 (not so good)?

In principle one would like to delineate the field designated by “money” before asking questions within it. But in practice what concepts we group with money depends on our views about it. So let’s move on to some more substantive questions. 

*

1. First, is money better imagined as a (physical) token, or as a unit of measurement? Or perhaps better, does our analysis of money start with exchange, or with accounting? Do we start by asking what is the thing that is exchanged with commodities, and then build an account of its use as a standard of value and in debt contracts on top of that? Or do we start with he idea of money as a unit like the meter or second, which is used to denominate obligations? In which case the debts incurred in the circulation of commodities appear as one particular case of the more general category, and the question of what exactly is accepted in settlement of an obligation is secondary.

I think of this as the difference between an exchange-first and and an accounting-first approach; Schumpeter makes a similar distinction between money theories of credit and credit theories of money. You’ll find most economists from Adam Smith to modern textbook writers on the first side, along with Marx (arguably) and most Marxists (definitely). On the second side you’ll find Keynes (in the Treatise if not the General Theory) along with Schumpeter, various chartalists, and sociologists like Geoffrey Ingham. 

This is a question about logical priority, about where we should start analytically. But the same question can be, and often is, posed as a historical one. Did money originate out of barter, or out of a system of public record-keeping? In principle, the origin of money is separate from the question of how we should best think go if it today. But in practice, almost everyone writing about the origin of money is interested in it because they think it is informative about, or a parable for, how money works in the present.

Another dimension of this question is how we think of the central bank. Do we think of it as — in some more or less metaphorical sense — issuing the nation’s currency? Or do we think of it as the peak institution of the banking system? 

2. A second question, related to the first one, is, where do we draw the line between money and credit? Is there a sharp divide, or a continuum? Or does money just describe particular kinds of credit, or credit as it is used in certain settings? To the extent there is a distinction, which is primary and which is derivative? Is money any promise, or any promise that can be transferred to a third party, or anything that can be used to settle an obligation? Almost any statement about money can have a different meaning depending on what parts of credit and finance are implicitly being included with it.

Similarly, is there a sharp line between money and other assets, or does money describe some function(s) that can be performed to different degrees by many assets? An important corollary of this is, is there a meaningful quantity of money? If there is a sharp line between money and other assets then at any moment there should be a definite quantity of money in existence. If  “moneyness” is a property which all kinds of assets possess in different degrees, then there isn’t. This is a more important question than it might seem, because many older debates about money are were framed in terms of the quantity of it, and it’s not always obvious how to translate them for a world where liquidity exists across the balance sheet, on both the asset and liability sides. 

This is a question where the conventional wisdom has shifted quite sharply over the past generation. Into the 1990s, both mainstream and heterodox writers used the money stock M as a basic part of the theoretical toolkit. But now it has almost entirely disappeared from the conversation in both academic and policy worlds.1

3. Third: To the extent that it is meaningful to talk about a quantity of money, is the quantity fixed independently of demand for it, or does it vary endogenously with demand? (And if so, does this happen within the banking system, or through the actions of the monetary authority? — the old horizontalists versus verticalists debate.) When I was first studying economics, this question was a central line of conflict between (Post) Keynesians and the mainstream, but its valence has shifted since then. “Banks create money” used to be a touchstone for heterodox views; now it’s something that everyone knows. There is still the question of how much this matters, i.e. how much bank lending is constrained by the supply of reserves or monetary policy more broadly. Victoria Chick has a fascinating piece on shifting views on this question over the 20th century.

In general, talking about how M varies with demand for it now feels a bit conservative and old-fashioned, since it assumes that the money stock is an economically meaningful quantity.  After teaching some of the same articles on these questions that I read in graduate school, I feel like the question is now: How can the debate over endogenous money be reformulated for a world without a distinct money stock? Another possible reframing: Is endogeneity inherent in the nature of money, or is it a contingent, institutional fact that evolves over time? At one point, bitcoin looked like an effort to re-exogenize the money supply; but I don’t think anybody talks about it that way anymore.

The flip side of the question of endogenous money — or maybe an alternative formulation of it — is, is the supply of money ever a constraint (on credit creation, and/or on real activity)? A negative answer is stronger than simply saying that the money supply is endogenous, since it further implies it can be expanded costlessly. 

4. This leads to the fourth question: What role does money play in the determination of the interest rate? Is interest, as Bagehot got put it, the price of money? Or is it the price of savings, or of future relative to present  consumption, which just happens to be expressed (like other prices) in terms of money? This is another long-standing frontline between orthodox economics and its Keynesian challengers, which remains an active site of conflict.

In the General Theory, Keynes developed his claims about money and interest in terms of  demand for an exogenously fixed stock of money. This was a serious wrong turn, in my view; chapter 17 (“The Essential Properties of Interest and Money”) is in my opinion the worst chapter of the book, the one most likely to confuse and mislead modern readers.2 But unlike endogeneity, this is a Keynesian theme that is easily transposed to an accounting-first key. We simply have to think of interest as the price of liquidity, rather than of one particular asset. This view of interest — as opposed to one that starts from savings — remains arguably the most important dividing line between orthodoxy and followers of Keynes. In general, if you want to work within Keynes’ system, you shouldn’t be talking about saving at all. 

5. The role of money in the determination of the interest rate leads to a fifth, broader question: Is money neutral? If so, with respect to what? And over what time horizon? In other words, do changes in the supply (or availability) of money affect “real” variables such as employment, or do they affect the price level? Or do they affect both, or neither?

From a political-policy perspective, neutrality is the question. Can increasing the availability of money (in general, or to some people in particular) solve coordination problems, mobilize unutilized resources, or otherwise increase the real wealth of the community? Or will it only bid up the price of the stuff that already exists? When, let’s say, late-19th century Populists demanded a more elastic currency, were they expressing the real interests of their farmer and artisan constituency, or were they victims (or peddlers) of economic snake oil? And if the former, what were the specific conditions that made more abundant money a meaningful political demand?

Another way of looking at this: Does money just facilitate trades that would have happened anyway? (What does it mean to facilitate, in that case?) On the other hand, if we think of money as a technology for making promises, for substituting a general obligation for a particular one, then it may do so to a greater or lesser extent. Increasing the availability of money, or broadening the range of ways it can be used, should make new forms of cooperation possible. If money is useful, shouldn’t it follow that more money is more useful?

Turning to the present, is the availability of money an important constraint on decarbonization?  The content of this question is contingent on some of the earlier ones; is the terms no which credit is available to green projects a question of money? But even if you say yes, it’s not clear how important this dimension of the problem is. There’s a plausible case, to me at least, that there is a vast universe of decarbonization projects with positive private returns at any reasonable discount rate, which nonetheless aren’t undertaken because of a lack of financing. But it’s also possible that credit constraints are not all that important, at least not directly; that what’s scarce is the relevant skilled labor and organizational capacity, not financing.3 

Though it lies a bit downstream from some of the more fundamental theoretical issues, money’s neutrality is probably the highest-stake question in these debates.

To what extent, and under what conditions, can increasing access to money and credit develop the real productive capacities of a community? To what extent are shorter-term fluctuations and crises the result of interruptions in the supply of money and credit? One reason, it seems to me, that debates on these questions can be so murky and acrimonious is that while economic orthodoxy makes a strong claim that money is neutral, there is no well-defined pole on the other side. Rejecting the textbook view, in itself, doesn’t tell us much about when and how money does matter. 

6. The other side of this is the sixth question: What is the relationship between money and inflation? If money is neutral with respect to the “real” economy (bracketing what exactly this means) then what it does affect must be the price level.  If you pick up, let’s say, Paul Krugman’s international economics textbook, you will find the thoroughly Friedmanesque claim that the central banks its the supply of money (M), that in the short run an increase in the supply of money may raise output and employment, but over periods beyond a few years, changes in the money supply simply translate one for one into changes the price level, with output and other “real” variables following the same path regardless of what the central bank does.

The claim that the price level varies directly with an exogenously fixed money supply is the quantity theory of money, arguably the oldest theory in economics. This can be derived on first principles only under a set of stringent assumptions that clearly done’t describe real economies. So is there some broader metaphorical sense in which it is sort of true, at least in some times and places? Inflation is only defined with respect a unit of account, but it’s not clear that there is any necessary link with money in its concrete existence. 

Here, unlike the previous question, there are (at least) two well-defined poles. Anyone who has read anything on these issues has encountered Friedman’s koan that inflation is everywhere and always a monetary phenomenon. Against this there is a vocal group of economists (both Post Keynesian and more mainstream) who counter that “inflation is always and everywhere a conflict phenomenon.” Personally, I am not convinced that inflation is always and everywhere any one particular thing. But that is a topic for another time.

7. More broadly, whether reimagine “the money supply” as a fixed quantity or in terms of more or less elastic credit, we can ask, are changes in money supply  linked to changes in prices, in incomes, in the interest rate, or some combination of them? This leads to the seventh question: Is the money supply, or the terms on which money is provided or created, an appropriate object of policy? This is partly question about what social objectives can be advanced by changes in the availability of money. But it is also a question about whether there is something inherently public about money as a social ledger, which means that it should be (or in some sense always is) the responsibility of the state.

8. Which brings us to question eight: Is there a fundamental relationship between money and the state, and with the authority to collect taxes? Georg Simmel famously described money as “a claim against society”. Who represents society, in this case? Is it — necessarily or in practice — the government? If we think of money as a ledger recording all kinds of obligations as commensurable quantities and allowing them to be netted out, is the use of such a shared ledger necessarily imposed by a sovereign authority, or can we think of it as arising organically? A bit more concretely: Is the value of money backed, in some sense, by the authority to tax? This view is strongly associated with chartalism. But you can also get a version of it from someone like Duncan Foley, working within the Marxist tradition.

9. Turning to money as a unit of measurement, our ninth question is: Do money values refer to some objective underlying quantity? And if so, what is it? What does it mean to speak of “real” values underlying the monetary ones? Obviously money values have objective content within a given pay community. For an individual within the community, the fact that two objects – or more precisely, two distinct property rights – have an equal price, implies the possibility of a choice between them. Ownership of this stuff and ownership of that stuff are equivalent in the sense that one can have one more of one by giving up an equal value of the other. For the community as a whole, we can, on some not too unreasonable assumptions, interpret price as reflecting the possibilities of producing more of one thing as opposed to something else. 

But what about when comparisons are made outside of an exchange community? If there is no possibility of substitution either in the purchase or production of things – where there is no market in which they exchange – is there a sense in which we can nonetheless compare their value? Do the quantities of money describe some underlying “real” quantity? When we compare “real income” ver time or between different countries, what is it exactly that we are comparing?

The textbook answer is that we are thinking of the economy in terms of a single representative consumer whose preferences are the same in all times and places (and at all levels of income), and asking how much income in one setting it would take to buy a basket of goods that this representative consumer would willingly swap for the average basket of goods consumed somewhere else. When stated like this, it sounds absurd. Yet this is literally the basis for widely used price level measures like Purchasing Power Parity indexes used to compare real incomes across countries. The problem is actually even worse than this, since even on the most heroic assumptions there is no way to consistently measure price levels both across countries and over time.4 But it’s very hard for people — certainly for economists — to give up the idea that there exists something called “real GDP” or “real income” that behaves like a physical quantity. 

If the neutrality of money is the question with the most immediate real-world implications, this one, I think, is where there is the biggest gap between what people assume or think they know, and what holds up on closer examination.

10. Related to this, question ten: Are relative prices prior to, or independent of, money prices? In his review of David Graeber’s Debt, Mike Beggs insisted that “States print the money, but not the price lists.”  This is the orthodox view — if one of commodity A trades for two of commodity B, that is an intrinsic fact about the commodities themselves, reflecting their costs of production and/or their ability to satisfy human needs.It doesn’t depend on the fact that  the prices are expressed in terms of money, or that the commodities are bought and sold for money rather than directly exchanged for each other.

But as I pointed out in my reply to Mike, not all economists agree with this. Hyman Minsky’s two-price model (much more interesting, in my mind, than the financial fragility hypothesis) is precisely an argument that money matters for the price of long-lived assets in a way that it does not for current output. The price of a building, say, cannot be derived from just the cost of producing it and the rent people will pay for it; it depend fundamentally on the terms on which it can be financed.

More broadly, we can think of some activities — those that lock in payment commitments while promising distant or uncertain income — as being more demanding of liquidity. Changes in the availability of money will change the price of these activities relative to those that are less liquidity-demanding.  From a Minskyan perspective, money is not neutral; the price lists depend fundamentally on how much (and on what terms) money is being printed.

11. Finally, some questions about the international dimension of money. First, various questions related to exchange rates — how they are, and should be, determined, and what effects they have on real activity. This is one area— perhaps the only one on the whole list — where, it seems to me, there is a very clear difference between today’s textbook views and pre-Keynesian orthodoxy. Today, floating exchange rates are treated as normal, and government interventions in the foreign exchange market are viewed with suspicion. Whereas the older orthodoxy assumed that currencies should, and apart from exceptional cases would, be permanently fixed in terms of gold. 

12. Twelfth: If we think of money as a ledger, does it matter where the ledger is kept? That the dollar is the global currency is true in obvious, observable ways — its unrivaled dominance in reserve holdings, foreign-exchange transactions, and trade pricing. (And despite constant predictions to the contrary, this shows no signs of changing.) But what constraints does this fact impose on the rest of the world, both in terms of international positions and domestic finance? And what, advantages (or disadvantages) does it have for the United States? 

One argument (made powerfully by Jörg Bibow, and also in this old working paper by me) is that in a world of unmanaged cross-border trade and financial flows, the United States current account deficit plays an essential role as a source of dollar liquidity for the rest of the world — that efforts to balance US trade will only lead to slower growth elsewhere. The assumption here, which may or may not be reasonable, is that there is something like of an exogenous stock of global money, even if not at the national level.

A related issue is how the financial and current account sides of the balance of payments balance. If we think of money as a token or substance, then any given transaction involves a certain amount of it either flowing into or out of a country, and the need for these flows to equal out evidently calls for some kind of market mechanism. On the other hand, if we think of money as a ledger entry, then the mere fact that a transaction takes place automatically creates an offsetting entry on the financial account. There may well be ways in which, say, foreign demand for a country’s assets causes its trade balance to shift toward deficit. But the argument has to be made in behavioral terms, it is not necessarily true.

13. Finally, thirteen: What does it mean to possess monetary sovereignty? Is having control over your own money a binary, yes or no question, or does it exist on a continuum? A more concrete aspect of this question is under what conditions countries can set their own interest rates. The older view was that a floating exchange rate was sufficient; the newer view — among established as well as heterodox economists — is that autonomous monetary policy is only possible with limits on financial flows, i.e. capital controls. Otherwise, what happens to your interest rate depends on the Fed’s choices, not yours.

*

I have my own opinions about what are more and less convincing answers to these questions. But my goal is not to convince you, or my students, of the answers. My goal is to convince you that these are real questions.

One reason that arguments about money-related questions are so often so painful an inconclusive, it seems to me, is that people start out from strong commitments to particular answers to various of these questions, or questions like these, without even realizing that they are questions — that it is possible to take a view on the other side.  Almost nobody who talks about “real GDP” pauses to ask what exactly this number refers to. That the interest rate is the price of liquidity — of money — is the pivot of Keynes’ whole argument in The General Theory. But it’s constantly ignored or forgotten by people who think of themselves as Keynesians. In general, it seems to me, debates connected with money are less often about disagreements on substantive issues than about different premises, which are seldom recognized or acknowledged. Before denouncing each other, before accusing people of some basic error of fact, let’s at least try to map out the intellectual terrain we are fighting over. 

A second purpose of this list is to show how these are not just academic questions, but have important implications for our efforts to, in Haavelmo’s phrase, become masters the happenings of real life. To be sure, this post doesn’t do this. But it was a goal of the class. And it is very much a goal of the book.

 

Revisiting the Euro Crisis

The euro crisis of the 2010s is well in the past now, but it remains one of the central macroeconomic events of our time.  But the nature of the crisis remains widely misunderstood, not only by the mainstream but also — and more importantly from my point of view — by economists in the heterodox Keynesian tradition. In this post, I want to lay out what I think is the right way of thinking about the crisis. I am not offering much in the way of supporting evidence. For the moment, I just want to state my views as clearly as possible. You can accept them or not, as you choose. 

During the first 15 years of the euro, a group of peripheral European countries experienced an economic boom followed by a crash, with GDP, employment and asset prices rising and then falling even more rapidly. As far as I can tell, there are four broad sets of explanations on offer for the crises in Greece, Ireland, Italy, Portugal and Spain starting in 2008. 

(While the timing is the same as the US housing bubble and crash, that doesn’t mean they are directly linked — however different they are in other respects, most of the common explanations for the European crisis I’m aware of locate its causes primarily within Europe.s)

The four common stories are:

1. External imbalances. The fixed exchange rate created by the euro, plus some mix of slow productivity growth in periphery and weak demand growth in core led to large trade imbalances within Europe. The financial expansion in the periphery was the flip side of a causally prior current account deficit.

2. Monetary policy. Both financial instability and external imbalances were result of Europe being far from an optimal currency area. Trying to carry out monetary policy for the whole euro area inevitably produced a mix of stagnation in the core and unsustainable credit expansion in the periphery, since a monetary stance that was too expansionary for Greece, Spain etc. was too tight for Germany.

3. Fiscal irresponsibility. The root of the crisis in peripheral countries was the excessive debt incurred by their own governments. The euro was a contributing factor since it led to an excessive convergence of interest rates across Europe, as markets incorrectly believed that peripheral debt was now as safe as debt of core countries.

4. Banking crises. The booms and busts in peripheral Europe were driven by rapid expansions and then contractions of credit from the domestic banking systems, with dynamics similar to that in credit booms in other times and places. The specific features of the euro system did not play any significant role in the development of the crisis, though they did importantly shape its resolution. 

In my view, the fourth story is correct, and the other three are wrong. In particular, trade imbalances within Europe played no role in the crisis. In this post, I am going to focus on why I think the external balances story is wrong, since that’s the one that people who are on my side intellectually seem most inclined toward.

As I see it, there were two distinct causal chains at work, both starting with a credit boom in the peripheral countries.

easy credit —> increased aggregate spending —> increased output and income —> increased imports —> growing trade deficit —> net financial inflows

easy credit —> rising asset prices —> bubble and/or fraud —> asset price crash —> insolvent banks —> financial crisis

That the two outcomes — external imbalances and banking crisis — went together is not a coincidence. But there is no causal link from the first to the second. Both rather are results of the same underlying cause. 

Yes, in the specific conditions of the late-2000s euro area, a credit boom led to an external deficit. But in principle it is perfectly possible to have a a credit-financed asset bubble and ensuing crisis in a country with a current account surplus, or one with current account balance, or in a closed economy. What was specific to the euro system was not the crisis itself, but the response to it. The reason the euro made the crisis worse because it prevented national governments from taking appropriate action to rescue their banking systems and stabilize demand. 

This understanding is, I think, natural if we take a “money view” of the crisis, thinking in terms of balance sheets and the relationship between income and expenditure. Here is the story I would like to tell.

Following the introduction of the euro in 1998, there were large credit expansions in a number of European countries. In Spain, for example, bank credit to the non-financial economy increased from 80 percent of GDP in 1997 to 220 percent of GDP in 2010. Banks were more willing to make loans, at lower rates, on more favorable terms, with less stringent collateral requirements and other lending standards. Borrowers were more willing to incur debt. The proximate causes of this credit boom may well have been connected to the euro in various ways. European integration offered a plausible story for why assets in Spain might be valued more highly. The ECB might have followed a less restrictive policy than independent central banks would have (or not — this is just speculation). But the euro was in no way essential to the credit boom. Similar booms have happened in many other times and places in the absence of currency unions — including, of course, in the US at roughly the same time.

In most of these countries, the bulk of the new credit went toward speculative real estate development. (In Greece there was also a big increase in public-sector borrowing, but not elsewhere.) The specifics of this lending don’t matter too much. 

Now for the key point. What happens when a a Spanish bank makes a loan? In the first step the bank creates two new assets – a deposit for the borrower, and the loan for itself. Notice that this does not require any prior “saving” by a third party. Expansion of bank credit in Spain does not require any inflow of “capital” from Germany or anywhere else.

Failure to grasp is an important source of confusion. Many people with a Keynesian background talk about endogenous money, but fail to apply it consistently. Most of us still have a commodity money or loanable-funds intuition lodged in the back of our brains, especially in international contexts. Terms like “capital flows” and “capital flight” are, in this respect, unhelpful relics of a gold standard world, and should probably be retired.

Back to the story. After the deposits are created, they are spent, i.e. transferred to someone else in return, in return for title to an asset or possession of a commodity or use of a factor of production. If the other party to this transaction is also Spanish, as would usually be the case, the deposits remain in the Spanish banking system. At the aggregate level, we see an increase in bank credit, plus an increase in asset prices and/or output, depending on what the loan finances, amplified by any ensuing wealth effect or multiplier.

To the extent that the loans finance production – of beach houses in Galicia say — they generate incomes. Some fraction of new income is spent on imported consumption goods. Probably more important, production requires imported intermediate and capital goods. By both these channels, an increase in Spanish output results in higher imports. If the credit boom leads Spain to grow faster relative to its trade partners — which it will, unless they are experiencing similar booms — then its trade balance will move toward deficit.

(That changes in trade flows are primarily a function of income growth, and not of relative prices, is an important item in the Keynesian catechism.)

Now let’s turn to the financial counterpart of this deficit. A purchase of a German good by a Spanish firm requires a bank deposit to be transferred from the Spanish firm to the German firm. Since the German firm presumably doesn’t hold deposits in a Spanish bank, we’ll see a reduction in deposits in the Spanish banking system and an equal increase in deposits in the German banking system. The Spanish banks must now replace those deposits with some other funding, which they will seek in the interbank market. So in the aggregate the trade deficit will generate an equal financial inflow — or, better said, a new external liability for the Spanish banking system. 

The critical thing to notice here is that these new financial positions are generated mechanically by the imports themselves. It is simply replacing the deposit funding the Spanish banks lost via payment for the imports. The financial inflow must take place for the purchase to happen — otherwise, literally, the importer’s check won’t clear. 

But what if there is an autonomous inflow – what if German wealth owners really want to hold more assets in Spain? Certainly that can happen. These kinds of cross-border flows may well have contributed to the credit boom in the periphery. But they have nothing to do with the trade balance. By definition, autonomous financial flows involve offsetting financial transactions, with no implications for the current account. 

Suppose you are a German pension fund that would like to lend money to a Spanish firm, to take advantage of the higher interest rates in Spain. Then you purchase, let’s say, a bond issued by Spanish construction company. That shows up as a new liability for Spain in the international investment position. But the Spanish firm now holds a deposit in a German bank, and that is an equal new asset for Spain. (If the Spanish firm transfers the deposit to a Spanish bank in return for a deposit there, as I suppose it probably would, then we get an asset for Spain in the interbank market instead.) The overall financial balance has not changed, so there is no reason for the current account to change either. Or as this recent BIS paper puts it, “the high correlations between gross capital inflows and outflows are overwhelmingly the result of double-entry bookkeeping”. (The importance of gross rather than net financial positions for crises is a pint the bIS has emphasized for many years.)

It may well happen that the effect of these offsetting financial transactions is to raise incomes in Spain (the contractor got better terms than it would at home) and/or banking-system liquidity (thanks to the fact that the Spanish banking system gets the deposits without the illiquid loan). This may well contribute to a rise in incomes in Spain and thus to a rise in the trade deficit. But this seems to me to be a second-order factor. And in any case we need to be clear about the direction of causality here — even if the financial inflows did indirectly cause the higher deficit, they did not in any sense finance it. The trade balances of Germany and Spain in no way affect the ability of German institutions to buy Spanish debt, any more than a New Yorker’s ability to buy a house in California depends on the trade balance between those states.

At this point it’s important to bring in the TARGET2 system. 

Under normal conditions, when someone wants to take a cross-border position within the euro systems the other side will be passively accommodated somewhere in the banking system. But if a net position develops for whatever reason, central banks can accommodate it via TARGET2 balances. Concretely, let’s say soon in Spain wants to make a payment to someone in Germany, as above. This normally involves the reduction of a Spanish bank’s liability to the Spanish entity and the increase in a German bank’s liability to the German entity. To balance this, the Spanish bank needs to issue some other liability (or give up an asset) while the German bank needs to acquire some asset. Normally, this happens by the Spanish bank issuing some new interbank liability (commercial paper or whatever) which ends up, perhaps via various intermediaries, as an asset for a German bank. But if foreign banks are unwilling to hold the liabilities of Spanish banks (as happened during the crisis) the Spanish bank can instead borrow from its own central bank, which in turn can create two offsetting positions through TARGET2 — a liability to the euro system, and a reserve asset (a deposit at the ECB). Conceptually, rather than the transfer of the despot being offset by a liability fro the Spanish to the German bank the interbank market, it’s now offset by a debt owed by the Spanish bank to its own national central bank, a debt between the central banks in the TARGET2 system, and a claim by the German bank against its own national central bank.

In this sense, within the euro system TARGET2 balances stand at the top of the hierarchy of money. Just as non financial actors settle their accounts by transfers of deposits at commercial banks, and banks settle their balances by transfers of deposits at the central bank, central banks settle any outstanding balances via TARGET2. It plays the same role as gold in the old gold standard system. Indeed, I sometimes think it would be better to describe the euro system as the “TARGET2 system.” 

There is however a critical difference between these balances and gold. Gold is an asset for central bank; TARGET2 balances are a liability. When a payment is made from country X to country Y in the euro area, with no offsetting private payment, the effect on central bank balance sheets is NOT a decrease in the assets of the central bank of X (and increase in the assets of the central bank of Y) but an increase in the liabilities of the central bank of X. This distinction is critical because assets are finite and can be exhausted, but new liabilities can be issued indefinitely. The automatic financing of payments imbalances through the TARGET2 system seems like an obscure technical detail but it transforms the functioning of the system. Every national central bank in the euro area is in effect in the situation of the Fed. It can never be financially constrained because all its obligations can be satisfied with its own liabilities. 

People are sometimes uncomfortable with this aspect of the euro system and suggest that there must be some limit on TARGET2 balances. But to me, this fundamentally misunderstands the nature of a single currency. What makes “the euro” a single currency is not that it has the same name, or that the bills look the same in the various countries, or even that it trades at a fixed ratio of one for one. What makes it a single currency is that a bank deposit in any euro-area country will settle a debt in any other euro-area country, at par. TARGET2 balances have to be unlimited to guarantee the this will be the case — in other words, for there to be a single currency at all.

(In this sense, we should not have been so fixed on the question of being “in” versus “out” of the euro. The relevant question is the terms on which payments can be made from one bank account another, for settlement of which obligations.)

The view of the euro crisis in which trade imbalances finance or somehow enable credit expansion is dependent on a loanable-funds perspective in which incomes are fixed, money is exogenous and saving is a binding constraint. It’s implicitly based on a model of the gold standard in which increased lending impossible without inflow of reserves — something that was not really true in practice even in the high gold standard era and isn’t true even in principle today. What’s strange is that many people who accept this view would reject those premises – if they realized they were applying them.

Meanwhile, on the domestic side, abundant credit was bidding up asset prices and encouraging investment that was, ex post, unwise (and in some case fraudulent, though I have no idea how important this was quantitatively). When asset prices collapsed and the failure of investment projects to generate the expected returns became clear, many banks faced insolvency. There was a collapse in activity in the real-estate development and construction activity that had driven the boom and, as banks tightened credit standards across the board, in other credit-dependent activity; falling asset values further reduced private spending; all these effects were amplified by the usual multiplier. The result was a steep fall in output and employment.

I don’t believe there’s any sense in which a sudden stop of cross-border lending precipitated the crisis. Rather, the “nationalization” of finance came after. Banks tried to limit their cross-border positions came only once the crisis was underway, as it became clear that there would be no systematic euro-wide response to insolvent banks, so that any rescues or bailouts would be by national governments for their own banks.

Credit-fueled asset booms and crashes have happened in many times and places. There was nothing specific to the euro system about the property booms of the 2000s. What was specific to the euro system was what happened next. Thanks to the euro, the affected governments could not respond as developed country governments have always responded to financial crises since World War II — by recapitalizing insolvent banks and shifting public budgets toward deficit until private demand recovers. 

The constraints on euro area governments were not an inevitable feature of system, in this view. Rather, they were deliberately imposed through discretionary choices by the authorities in order to use the crisis to advance a substantive political agenda.

 

At Barron’s: Are Low Rates to Blame for Bubbles?

(I write a monthly opinion piece for Barron’s. These sometimes run in the print edition, which I appreciate — it’s a vote of confidence from the editors, and means more readers. It does impose a tighter word count limit, though. The text below is the longer version I originally submitted. The version that was published is here. All of my previous Barron’s pieces are here.)

The past year has seen a parade of financial failures and asset crashes. Silicon Valley bank was the first bank failure since 2020, and the biggest since 2008. Before that came the collapse of FTX, and of much of the larger crypto ecosystem. Corporate bankruptcies are coming faster than at any time since 2011.  Even luxury watches are in freefall. 

The proximate cause of much of this turmoil is the rise in interest rates. So it’s natural to ask if the converse is true. Is the overvaluing of so many worthless assets  – whether through bubbles or fraud – the fault of a decade-plus of low rates? For those who believe this, the long period of low rates following the global financial crisis fueled an “everything bubble”, just as the earlier period of low rates fueled the housing boom of the 2000s. The rise of fragile or fraudulent institutions, which float up on easy credit before inevitably crashing back to earth, is a sign that monetary policy should never have been so loose. As journalist Rana Foorohar put it in a much-discussed article, “Keeping rates too low for too long encourages speculation and debt bubbles.”

You can find versions of this argument being made by  prominent Keynesians, as well as by economists of a more conservative bent. At the Bank for International Settlements “too low for too long” is practically a mantra. But, does the story make sense?

Yes, low interest rates are associated mean high asset prices. But that’s not the same as a bubble.To the extent an asset represents a stream of future payments, a low discount rate should raise its value. 

On the other hand, asset prices are not just about discounted future income streams; they also incorporate a bet on the future price of the asset itself. If a fall in interest rates leads to a rise in asset prices, market participants may mistakenly expect that rise to continue. That could lead to assets being overvalued even relative to the current low rates.

Another argument one sometimes hears for why low rates lead to bubbles is that when income from safe assets is low, investors will “reach for yield” by taking on more risk, bidding up the price of more speculative assets. Investors’ own liabilities also matter. When it’s cheap and easy to borrow, an asset may be attractive that wouldn’t be if financing were harder to come by.

But if low interest rates make acquiring risky assets more attractive, is that a problem? After all, that’s how monetary policy is supposed to work. The goal of rate cuts is precisely to encourage investment spending that wouldn’t happen if rates were higher.  As I argued recently, it’s not clear that most business investment is very responsive to interest rates. But whether the effect on the economy is strong or weak,  “low interest rates cause people to buy assets they otherwise wouldn’t” is just monetary policy working as intended.

Still, intended results may have unintended consequences. When people are reaching for yield, the argument goes, they are more likely to buy into projects that turn out to be driven by fraud, hype or fantasy.

Arguments for the dangers of low rates tend to take this last step for granted. But it’s not obvious why an environment of low yields should be more favorable to frauds. Projects with modest expected returns are, after all, much more common than projects with very high ones; when risk-free returns are very low, there should be more legitimate higher-yielding alternatives, and less need for risky long shots. Conversely, it is the projects that promise very high returns that are most likely to be frauds  — and that are viable at very high rates.

Certainly this was Adam Smith’s view. For him, the danger of speculation and fraud was not an argument for high interest rates, but the opposite. If legal interest rates were “so high as 8 or 10 percent,” he believed, then “the greater part of the money which was to be lent would be lent to prodigals and projectors, who alone would be willing to give this high interest. Sober people … would not venture into the competition.” 

The FTX saga is an excellent example. At one point, Sam Bankman-Fried—a projector and prodigal if ever there was one—offered as much as 20% on new loans to his hedge fund, Alameda, according to The Wall Street Journal. It wouldn’t take low rates to make that attractive — if he was good for it. But, of course, he was not. And that is the crux of the problem. Someone like Bankman-Fried is not offering a product with low but positive returns, that would be attractive only when rates are low but not when they were high. He was offering a product with an expected return that, in retrospect, was in the vicinity of -100 percent. Giving  him your money to him would be a bad idea at any interest rate. 

We can debate what it would take to prevent fraud-fueled bubbles in assets like cryptocurrency. Perhaps it calls for tighter restrictions on the kinds of products that can be offered for sale, or more stringent rules on the choices of retail investors. Or perhaps, given crypto’s isolation from the broader financial system, this is a case where it’s ok to just let the buyer beware. In any case, the problem was not that crypto offered higher returns than the alternative. The problem was that people believed the returns in crypto were much higher than they actually were. Is this a problem that interest rates can solve?

Let’s suppose for the sake of argument that it is. Suppose that without the option of risk-free returns of 3 or 4 or 5 percent, people will throw their money away on crazy longshots and obvious frauds. If you take this idea seriously, it has some funny implications. Normally, when we ask why asset owners are entitled to their income in the first place, the answer is that it’s an incentive to pick out the projects with the highest returns. (Hopefully these are also the most socially useful ones.) The “too low for too long” argument turns this logic on its head. It says that asset owners need to be guaranteed high returns because they can’t tell a good project from a bad one.

That said, there is one convincing version of this story. For all the reasons above, it does not make sense to think of ordinary investors being driven toward dangerous speculation by low interest rates. Institutions like insurance and pension funds are a different matter. They have long-term liabilities that are more or less fixed and, critically, independent of interest rates. Their long investment horizons mean their loss of income from lower rates will normally outweigh their capital gains when they fall. (This is one thing the BIS surely gets right.) When the alternative is insolvency, it can make sense to choose a project where the expected return is negative, if it offers a chance of getting out of the hole. That’s a common explanation for the seemingly irresponsible loans made by many Savings & Loans in the 1980s—faced with bankruptcy, they “gambled for resurrection.” One can imagine other institutions making a similar choice.

What broke the S&Ls in was high rates, not low ones. But there is a common thread. A structure set up when interest rates are in a certain range may not work when they move outside of it. A balance sheet set up on the basis of interest rates in some range will have problems if they move outside it. 

Modern economies depend on a vast web of payment expectations and commitments stretching far into the future. Changes in interest rates modify many change of those future payments; whether upward or downward, this means disappointed expectations and broken commitments. 

If the recent period of low rates was financially destabilizing,  then, the problem wasn’t the not low rates in themselves. It was that they weren’t what was planned on. If the Fed is going to draw general lessons from the bubbles that are now popping, it should not be about the dangers of low rates, but that of drastic and unexpected moves in either direction. 

At Barron’s: Americans Owe Less Than They Used To. Will the Fed Change That?

(I write a monthly opinion piece for Barron’s. This one was published there in September.)

Almost everyone, it seems, now agrees that higher interest rates mean economic pain. This pain is usually thought of in terms of lost jobs and shuttered businesses. Those costs are very real. But there’s another cost of rate increases that is less discussed: their effect on balance sheets.

Economists tend to frame the effects of interest rates in terms of incentives for new borrowing. As with (almost) anything else, if loans cost more, people will take less of them. But interest rates don’t matter only for new borrowers, they also affect people who borrowed in the past. As debt rolls over, higher or lower current rates get passed on to the servicing costs of existing debt. The effect of interest rate changes on the burden of existing debt can dwarf their effect on new borrowing—especially when debt is already high.

Let’s step back for a moment from current debates. One of the central macroeconomic stories of recent decades is the rise in household debt. In 1984, it was a bit over 60% of disposable income, a ratio that had hardly changed since 1960. But over the next quarter-century, debt-income ratios would double, reaching 130%. This rise in household debt was the background of the worldwide financial crisis of 2007-2008, and made household debt a live political question for the first time in modern American history.

Household debt peaked in 2008; it has since fallen almost as quickly as it rose. On the eve of the pandemic, the aggregate household debt-income ratio stood at 92%—still high, by historical standards, but far lower than a decade before.

These dramatic swings are often explained in terms of household behavior. For some on the political right, rising debt in the 1984-2008 period was the result of misguided government programs that encouraged excessive borrowing, and perhaps also a symptom of cultural shifts that undermined responsible financial management. On the political left, it was more likely to be seen as the result of financial deregulation that encouraged irresponsible lending, along with income inequality that pushed those lower down the income ladder to spend beyond their means.

Perhaps the one thing these two sides would agree on is that a higher debt burden is the result of more borrowing.

But as economist Arjun Jayadev and I have shown in a series of papers, this isn’t necessarily so. During much of the period of rising debt, households borrowed less on average than during the 1960s and 1970s. Not more. So what changed? In the earlier period, low interest rates and faster nominal income growth meant that a higher level of debt-financed expenditure was consistent with stable debt-income ratios.

The rise in debt ratios between 1984 and 2008, we found, was not mainly a story of people borrowing more. Rather, it was a shift in macroeconomic conditions that meant that the same level of borrowing that had been sustainable in a high-growth, low-interest era was unsustainable in the higher-interest environment that followed the steep rate hikes under Federal Reserve Chair Paul Volcker. With higher rates, a level of spending on houses, cars, education and other debt-financed assets that would previously have been consistent with a constant debt-income ratio, now led to a rising one.

(Yes, there would later be a big rise in borrowing during the housing boom of the 2000s. But this is not the whole story, or even the biggest part of it.)

Similarly, the fall in debt after 2008 in part reflects sharply reduced borrowing in the wake of the crisis—but only in part. Defaults, which resulted in the writing-off of about 10% of household debt over 2008-2012, also played a role. More important were the low interest rates of these years. Thanks to low rates, the overall debt burden continued to fall even as households began to borrow again.

In effect, low rates mean that the same fraction of income devoted to debt service leads to a larger fall in principal—a dynamic any homeowner can understand.

The figure nearby illustrates the relative contributions of low rates and reduced borrowing to the fall in debt ratios after 2008. The heavy black line is the actual path of the aggregate household debt-income ratio. The red line shows the path it would have followed if households had not reduced their borrowing after 2008, but instead had continued to take on the same amount of new debt (as a share of their income) as they did on average during the previous 25 years of rising debt. The blue line shows what would have happened to the debt ratio if households had borrowed as much as they actually did, but had faced the average effective interest rate of that earlier period.

As you can see, both reduced borrowing and lower rates were necessary for household debt to fall. Hold either one constant at its earlier level, and household debt would today be approaching 150% of disposable income. Note also that households were paying down debt mainly during the crisis itself and its immediate aftermath—that’s where the red and black lines diverge sharply. Since 2014, as household spending has picked up again, it’s only thanks to low rates that debt burdens have continued to fall.

(Yes, most household debt is in the form of fixed-rate mortgages. But over time, as families move homes or refinance, the effective interest rate on their debt tends to follow the rate set by the Fed.)

The rebuilding of household finances is an important but seldom-acknowledged benefit of the decade of ultra-low rates after 2007. It’s a big reason why the U.S. economy weathered the pandemic with relatively little damage, and why it’s growing so resiliently today.

And that brings us back to the present. If low rates relieved the burden of debt on American families, will rate hikes put them back on an unsustainable path?

The danger is certainly real. While almost all the discussion of rate hikes focuses on their effects on new borrowing, their effects on the burden of existing debt are arguably more important. The 1980s—often seen as an inflation-control success story—are a cautionary tale in this respect. Even though household borrowing fell in the 1980s, debt burdens still rose. The developing world—where foreign borrowing had soared in response to the oil shock—fared much worse.

Yes, with higher rates people will borrow less. But it’s unlikely they will borrow enough less to offset the increased burden of the debt they already have. The main assets financed by credit—houses, cars, and college degrees—are deeply woven into American life, and can’t be easily foregone. It’s a safe bet that a prolonged period of high rates will result in families carrying more debt, not less.

That said, there are reasons for optimism. Interest rates are still low by historical standards. The improvement in household finances during the post-2008 decade was reinforced by the substantial income-support programs in the relief packages Congress passed in response to the pandemic; this will not be reversed quickly. Continued strong growth in employment means rising household incomes, which, mechanically, pushes down the debt-income ratio.

Student debt cancellation is also well-timed in this respect. Despite the fears of some, debt forgiveness will not boost  current demand—no interest has been paid on this debt since March 2020, so the immediate effect on spending will be minimal. But forgiveness will improve household balance sheets, offsetting some of the effect of interest rate hikes and encouraging spending in the future, when the economy may be struggling with too little demand rather than (arguably) too much.

Reducing the burden of debt is also one of the few silver linings of inflation. It’s often assumed that if people’s incomes are rising at the same pace as the prices of the things they buy, they are no better off. But strictly speaking, this isn’t true—income is used for servicing debt as well as for buying things. Even if real incomes are stagnant or falling, rising nominal incomes reduce the burden of existing debt. This is not an argument that high inflation is a good thing. But even bad things can have benefits as well as costs.

Will we look back on this moment as the beginning of a new era of financial instability, as families, businesses, and governments find themselves unable to keep up with the rising costs of servicing their debt? Or will the Fed be able to declare victory before it has done too much damage? At this point, it’s hard to say.

Either way we should focus less on how monetary policy affects incentives, and more time on how it affects the existing structure of assets and liabilities. The Fed’s ability to steer real variables like GDP and employment in real time has, I think, been greatly exaggerated. Its long-run influence over the financial system is a different story entirely.

Fisher Dynamics Revisited

Back in the 2010s, Arjun Jayadev and I wrote a pair of papers (one, two) on the evolution of debt-income ratios for US households. This post updates a couple key findings from those papers. (The new stuff begins at the table below.)

Rather than econometric exercises, the papers were based on a historical accounting decomposition —  an approach that I think could be used much more widely. We separated changes in the debt-income ratio into six components — the primary deficit (borrowing net of debt service payments); interest payments; real income growth; inflation; and write downs of debt through default — and calculated the contribution of each to the change in debt ratios over various periods. This is something that is sometimes done for sovereign debt but, as far as I know, we were the first to do it for private debt-income ratios.

We referred to the contributions of the non-borrowing components as “Fisher dynamics,” in honor of Irving Fisher’s seminal paper on depressions as “debt deflations.” A key aspect of the debt-deflation story was that when nominal incomes fell, the burden of debt could rise even as debtors sharply reduced new borrowing and devoted a greater share of their income to paying down existing debt. In Fisher’s view, this was one of the central dynamics of the Great Depression. Our argument was that something like a slow-motion version of this took place in the US (and perhaps elsewhere) in recent decades.

The logic here is that the change in debt-income ratios is a function not only of new borrowing but also of the effects of interest, inflation and (real) income growth on the existing debt ratio, as well as of charge offs due to defaults.

Imagine you have a mortgage equal to double your annual income. That ratio can go down if your current spending is less than your income, so that you can devote part of your income to paying off the principal. Or it can go down if your income rises, i.e. by raising the denominator rather than lowering the numerator. It can also go down if you refinance at a lower interest rate; then the same fraction of your income devoted to debt service will pay down the principal faster. Our of course it can go down if some or all of it is written off in bankruptcy.

It is possible to decompose actual historical changes in debt-income ratios for any economic unit or sector into these various factors. The details are in either of the papers linked above. One critical point to note: The contributions of debt and income growth are proportional to the existing debt ratio, so the higher it already is, the more important these factors are relative to the current surplus or deficit.

Breaking out changes in debt ratios into these components was what we did in the two papers. (The second paper also explored alternative decompositions to look at the relationship been debt ratio changes and new demand from the household sector.) The thing we wanted to explain was why some periods saw rising debt-income ratios while others saw stable or falling ones.

While debt–income ratios were roughly stable for the household sector in the 1960s and 1970s, they rose sharply starting in the early 1980s. The rise in household leverage after 1980 is normally explained in terms of higher household borrowing. But increased household borrowing cannot explain the rise in household debt after 1980, as the net flow of funds to households through credit markets was substantially lower in this period than in earlier postwar decades. During the housing boom period of 2000–2007, there was indeed a large increase in household borrowing. But this is not the case for the earlier rise in household leverage in 1983–1990, when the debt– income ratios rose by 20 points despite a sharp fall in new borrowing by households.

As we explained:

For both the 1980s episode of rising leverage and for the post-1980 period as a whole, the entire rise in debt–income ratios is explained by the rise in nominal interest rates relative to nominal income growth. Unlike the debt deflation of the 1930s, this ‘debt disinflation’ has received little attention from economists or in policy discussions.

Over the full 1984–2011 period, the household sector debt–income ratio almost exactly doubled… Over the preceding 20 years, debt–income ratios were essentially constant. Yet households ran cumulative primary deficits equal to just 3 percent of income over 1984–2012 (compared to 20 percent in the preceding period). The entire growth of household debt after 1983 is explained by the combination of higher interest payments, which contributed an additional 3.3 points per year to leverage after 1983 compared with the prior period, and lower inflation, which reduced leverage by 1.3 points per year less.

We concluded:

From a policy standpoint, the most important implication of this analysis is that in an environment where leverage is already high and interest rates significantly exceed growth rates, a sustained reduction in household debt–income ratios probably cannot be brought about solely or mainly via reduced expenditure relative to income. …There is an additional challenge, not discussed in this paper, but central to both Fisher’s original account and more recent discussions of ‘balance sheet recessions’: reduced expenditure by one sector must be balanced by increased expenditure by another, or it will simply result in lower incomes and/or prices, potentially increasing leverage rather than decreasing it. To the extent that households have been able to run primary surpluses since 2008, it has been due mainly to large federal deficits and improvement in US net exports.

We conclude that if reducing private leverage is a policy objective, it will require some combination of higher growth, higher inflation, lower interest rates, and higher rates of debt chargeoffs. In the absence of income growth well above historical averages, lower nominal interest rates and/or higher inflation will be essential. … Deleveraging via low interest rates …  implies a fundamental shift in monetary policy. If interest-rate policy is guided by the desired trajectory of debt ratios, it no longer can be the primary instrument assigned to managing aggregate demand. This probably also implies a broader array of interventions to hold down market rates beyond traditional open market operations, policies sometimes referred to as ‘financial repression.’ Historically, policies of financial repression have been central to almost all episodes where private (or public) leverage was reduced without either high inflation or large-scale repudiation.

These papers only went through 2011. I’ve thought for a while it would be interesting to revisit this analysis for the more recent period of falling household debt ratios. 

With the help of Arjun’s student Advait Moharir, we’ve now brought the same analysis forward to the end of 2019. Stopping there was partly a matter of data availability — the BEA series on interest payments we use is published with a considerable lag. But it’s also a logical period to look at, since it brings us up to the start of the pandemic, which one would want to split off anyway.

The table below is a reworked version of tables in the two papers, updated through 2019. (I’ve also adjusted the periodization slightly.) 

Due to …
Period Annual PP Change in Debt Ratio Primary Deficit Interest Growth Inflation Defaults
1929 – 1931 3.7 -5.5 2.9 2.8 2.9 *
1932 – 1939 -1.2 -1.5 2.4 -1.6 -0.7 *
1940 – 1944 -3.8 -1.6 1.3 -2.5 -1.9 *
1945 – 1963 2.6 2.5 2.6 -1.5 -0.8 *
1964 – 1983 0.0 0.8 5.1 -2.4 -3.5 *
1984 – 1999 1.7 -0.3 7.5 -2.9 -2.1 -0.4
2000 – 2008 4.5 2.4 7.2 -1.7 -2.5 -0.8
2009 – 2013 -5.4 -3.7 5.8 -3.1 -2.3 -2.4
2014 – 2019 -2.0 -1.4 4.6 -3.4 -1.3 -0.6

Again, our central finding in the earlier papers was that if we compare the 1984-2008 period of rising debt ratios to the previous two decades of stable debt ratios, there was no rise in the primary deficit. For 1984-2008 as a whole, annual new borrowing exceeded debt service payments by 0.7 percent of income on average, almost exactly the same as during the 1964-1983 period. (That’s the weighted average of the two sub-periods shown in the table.) Even during the housing boom period, when new borrowing did significantly exceed debt service, this explained barely a third of the difference in annual debt-ratio growth (1.6 out of 4.5 points).

The question now is, what has happened since 2008? What has driven the fall in debt ratios from 130 percent of household income in 2008 to 92 percent on the eve of the pandemic?

In the immediate aftermath of the crisis, sharply reduced borrowing was indeed the main story. Of the 10-point swing in annual debt-ratio growth (from positive 4.5 points per year to negative 5.4), 6 points is accounted for by the fall in net borrowing (plus another 1.5 points from higher defaults). But for the 2014-2019 period, the picture is more mixed. Comparing those six years to the whole 1984-2008 period of rising debt, we have a 4.7 point shift in debt ratio growth, from positive 2.7 to negative 2. Of that, 2.1 points is explained by lower net borrowing, while almost 3 points is explained by lower interest. (The contribution of nominal income growth was similar in the two periods.) So if we ask why household debt ratios continued to fall over the past decade, rather than resuming their rise after the immediate crisis period, sustained low interest rates are at least as important as household spending decisions. 

Another way to see this is in the following graph, which compares three trajectories: The actual one in black, and two counterfactuals in red and blue. The red counterfactual is constructed by combining the average 1984-2008 level of net borrowing as a fraction of income to the actual historical rates of interest, nominal income growth and defaults. The blue counterfactual is similarly constructed by combining the average 1984-2008 effective interest rate with historical levels of net borrowing, nominal income growth and defaults. In other words, the red line shows what would have happened in a world where households had continued to borrow as much after 2008 as in the earlier period, while the blue line shows what would have happened if households had faced the same interest rates after 2008 as before. 

As the figure shows, over the 2008-2019 period as a whole, the influence of the two factors is similar — both lines end up in the same place. But the timing of their impact is different. In the immediate wake of the crisis, the fall in new borrowing was decisive — that’s why the red and black lines diverge so sharply. But in the later part of the decade, as household borrowing moved back toward positive territory and interest rates continued to fall, the more favorable interest environment became more important. That’s why the blue line starts rising after 2012 — if interest rates had been at their earlier level, the borrowing we actually saw in the late 2010s would have implied rising debt ratios. 

As with the similar figures in the papers, this figure was constructed by using the law of motion for debt ratios:

where b is the debt-income ratio, d is the primary deficit, is the effective interest rate (i.e. total interest payments divided by the stock of debt), g is income growth adjusted for inflation, π is the inflation rate, and sfa is a stock-flow adjustment term, in this case the reduction of debt due to defaults. The exact sources and definitions for the various variables can be found in the papers. (One note: We do not have a direct measurement of the fraction of household debt written off by default for the more recent period, only the fraction of such debt written down by commercial banks. So we assumed that the ratio of commercial bank writeoffs of household debt to total writeoffs was the same for the most recent period as for the period in which we have data for both.)

Starting from the actual debt-ratio in the baseline year (in this case, 2007), each year’s ending debt-income ratio is calculated using the primary deficit (i.e. borrowing net of debt service payments), the share of debt written off in default, nominal income growth and the interest rate. All but one of these variables are the actual historical values; for one, I instead use the average value for 1984-2007. This shows what the path of the debt ratio would have been if that variable had been fixed at its earlier level while the others evolved as they did historically.  In effect, the difference between these counterfactual lines and the historical one shows the contribution of that variable to the difference between the two periods.

Note that the interest rate here is not the current market rate, but the effective or average rate, that is, total interest payments divided by the stock of debt. For US households, this fell from around 6 percent in 2007 to 4.4 percent by 2019 — less than the policy rate did, but still enough to create a very different trajectory, especially given the compounding effect of interest on debt over time. So while expansionary monetary policy is not the whole story of falling debt ratios since 2008, it was an important part of it. As I recently argued in Barrons, the deleveraging of US households is unimportant and under appreciated benefit of the decade of low interest rates after the crisis.

 

At the International Economy: What’s Wrong with Abundant Liquidity?

(I am an occasional contributor to roundtables of economists in the magazine The International Economy. This month’s topic was the possible dangers of “today’s giant swirling ocean of liquidity”.)

Imagine a city that experiences a miraculous improvement in its transit system. Thanks to some mix of new technologies and organizational improvements, the subways and buses are now able to carry far more passengers at lower cost and the same level of service. Would we see that as good news, or as bad? It’s true that Uber drivers and gas station owners would be unhappy as abundant public transportation reduced demand for their services. And retailers and restaurants might face challenges in managing a sudden flood of new customers. But no one, presumably, would think the city should deliberately give up the improvements and return transit service back to its old level. 

The point of this little fable should be obvious: liquidity, like transportations services, is useful. Having more of it is better than having less. 

What liquidity is useful for, fundamentally, is making promises. It functions as a kind of collective trust. The world is full of socially useful projects that can’t be carried out because even a well-grounded expectation of future benefits can’t be turned into a claim on resources today. Liquidity is the fuel for these transactions. In a world of abundant credit and low interest rates, it’s easier for me to turn my future income into ownership of a home, or a business to turn future profits into new plant and equipment, or a government to turn future revenue into improved public services.

Someone with a great business plan but no capital of their own might try to get the labor and inputs they need to bring it about by promising workers and vendors a share in the profits. Unless the business can be launched with just the resources of immediate family and friends, though, it’s not likely to get off the ground this way. The role of the bank is to allow strangers, and not just those who already know and trust each other, to contribute to the plan, by accepting — after appropriate scrutiny — the entrepreneur’s promise, and offering its own generally-negotiable promise to the suppliers of labor and other resources. 

Yes, when you make it easier to make promises, some of them won’t pan out. But we would like people to make more provision for future needs, not less, even if our knowledge of those needs is less than perfect. The most dynamic parts of the economy are the ones where there are the most risky projects, some of which inevitably fail.

Of course asset owners are unhappy about lower yields. But that’s no different from the complaints we always hear from incumbents when production improvements make something cheaper. Asset owners’ complaints are no more reason to deny us the socially useful services of liquidity than those of the proverbial buggy-whip makers were to deny us the services of cars. (Less reason, actually, given the concentration of financial wealth among the wealthiest families and institutions.)

Interest rates today are lower than at almost any time in history. So are the prices of food or clothing. We should see abundant liquidity the same way we see these other forms of abundance  — as the fruit of the technological and institutional that has made us so much materially richer than our ancestors.

Inflation, Interest Rates and the Fed: A Dissent

Last week, my Roosevelt colleague Mike Konczal said on twitter that he endorsed the Fed’s decision to raise the federal funds rate, and the larger goal of using higher interest rates to weaken demand and slow growth. Mike is a very sharp guy, and I generally agree with him on almost everything. But in this case I disagree. 

The disagreement may partly be about the current state of the economy. I personally don’t think the inflation we’re seeing reflects any general “overheating.” I don’t think there’s any meaningful sense in which current employment and wage growth are too fast, and should be slower. But at the end of the day, I don’t think Mike’s and my views are very different on this. The real issue is not the current state of the economy, but how much confidence we have in the Fed to manage it. 

So: Should the Fed be raising rates to control inflation? The fact that inflation is currently high is not, in itself, evidence that conventional monetary policy is the right tool for bringing it down. The question we should be asking, in my opinion, is not, “how many basis points should the Fed raise rates this year?” It is, how conventional monetary policy affects inflation at all, at what cost, and whether it is the right tool for the job. And if not, what should we be doing instead?

What Do Rate Hikes Do?

At Powell’s press conference, Chris Rugaber of the AP asked an excellent question: What is the mechanism by which a higher federal funds rate is supposed to bring down inflation, if not by raising unemployment?1 Powell’s answer was admirably frank: “There is a very, very tight labor market, tight to an unhealthy level. Our tools work as you describe … if you were moving down the number of job openings, you would have less upward pressure on wages, less of a labor shortage.”

Powell is clear about what he is trying to do. If you make it hard for businesses to borrow, some will invest less, leading to less demand for labor, weakening workers’ bargaining power and forcing them to accept lower wages (which presumably get passed on to prices, tho he didn’t spell that step out.) If you endorse today’s rate hikes, and the further tightening it implies, you are endorsing the reasoning behind it: labor markets are too tight, wages are rising too quickly, workers have too many options, and we need to shift bargaining power back toward the bosses.

Rather than asking exactly how fast the Fed should be trying to raise unemployment and slow wage growth, we should be asking whether this is the only way to control inflation; whether it will in fact control inflation; and whether the Fed can even bring about these outcomes in the first place.

Both hiring and pricing decisions are made by private businesses (or, in a small number of cases, in decentralized auction markets.) The Fed can’t tell them what to do. What it can do – what it is doing – is raise the overnight lending rate between banks, and sell off some part of the mortgage-backed securities and long-dated Treasury bonds that it currently holds. 

A higher federal funds rate will eventually get passed on to other interest rates, and also (and perhaps more importantly) to credit conditions in general — loan standards and so on. Some parts of the financial system are more responsive to the federal funds rate than others. Some businesses and activities are more dependent on credit than others.

Higher rates and higher lending standards will, eventually, discourage borrowing. More quickly and reliably, they will raise debt service costs for households, businesses and governments, reducing disposable income. This is probably the most direct effect of rate hikes. It still depends on the degree to which market rates are linked to the policy rate set by the Fed, which in practice they may not be. But if we are looking for predictable results of a rate hike, higher debt service costs are one of the best candidates. Monetary tightening may or may not have a big effect on unemployment, inflation or home prices, but it’s certainly going to raise mortgage payments — indeed, the rise in mortgage rates we’ve seen in recent months presumably is to some degree in anticipation of rate hikes.

Higher debt service costs disposable income for households and retained earnings for business, reducing consumption and investment spending respectively. If they rise far enough, they will also lead to an increase in defaults on debt.

(As an aside, it’s worth noting that a significant and rising part of recent inflation is owners’ equivalent rent, which is a BLS estimate of how much homeowners could hypothetically get if they rented out their homes. It is not a price paid by anyone. Meanwhile, mortgage payments, which are the main actual housing cost for homeowners, are not included in the CPI. It’s a bit ironic that in response to a rise in a component of “housing costs” that is not actually a cost to anyone, the Fed is taking steps to raise what actually is the biggest component of housing costs.)

Finally, a rate hike may cause financial assets to fall in value — not slowly, not predictably, but eventually. This is the intended effect of the asset sales.

Asset prices are very far from a simple matter of supply and demand — there’s no reason to think that a small sale of, say 10-year bonds will have any discernible effect on the corresponding yield (unless the Fed announces a target for the yield, in which case the sale itself would be unnecessary.) But again, eventually, sufficient rate hikes and asset sales will presumably lead asset prices to fall. When they do fall, it will probably by a lot at once rather than a little at a time – when assets are held primarily for capital gains, their price can continue rising or fall sharply, but it cannot remain constant. If you own something because you think it will rise in value, then if it stays at the current price, the current price is too high.

Lower asset values in turn will discourage new borrowing (by weakening bank balance sheets, and raising bond yields) and reduce the net worth of households (and also of nonprofits and pension funds and the like), reducing their spending. High stock prices are often a major factor in periods of rising consumption, like the 1990s; a stock market crash could be expected to have the opposite impact.

What can we say about all these channels? First, they will over time lead to less spending in the economy, lower incomes, and less employment. This is how hikes have an effect on inflation, if they do. There is no causal pathway from rate hikes to lower inflation that doesn’t pass through reduced incomes and spending along the way. And whether or not you accept the textbook view that the path from demand to prices runs via unemployment wage growth, it is still the case that reduced output implies less demand for labor, meaning slower growth in employment and wages.

That is the first big point. There is no immaculate disinflation. 

Second, rate hikes will have a disproportionate effect on certain parts of the economy. The decline in output, incomes and employment will initially come in the most interest-sensitive parts of the economy — construction especially. Rising rates will reduce wealth and income for indebted households. 2. Over time, this will cause further falls in income and employment in the sectors where these households reduce spending, as well as in whatever categories of spending that are most sensitive to changes in wealth. In some cases, like autos, these may be the same areas where supply constraints have been a problem. But there’s no reason to think this will be the case in general.

It’s important to stress that this is not a new problem. One of the things hindering a rational discussion of inflation policy, it seems to me, is the false dichotomy that either we were facing transitory, pandemic-related inflation, or else the textbook model of monetary policy is correct. But as the BIS’s Claudio Borio and coauthors note in a recent article, even before the pandemic, “measured inflation [was] largely the result of idiosyncratic (relative) price changes… not what the theoretical definition of inflation is intended to capture, i.e. a generalised increase in prices.” The effects of monetary policy, meanwhile, “operate through a remarkably narrow set of prices, concentrated mainly in the more cyclically sensitive service sectors.”

These are broadly similar results to a 2019 paper by Stock and Watson, which finds that only a minority of prices show a consistent correlation with measures of cyclical activity.3 It’s true that in recent months, inflation has not been driven by auto prices specifically. But it doesn’t follow that we’re now seeing all prices rising together. In particular, non-housing services (which make up about 30 percent of the CPI basket) are still contributing almost nothing to the excess inflation. Yet, if you believe the BIS results (which seem plausible), it’s these services where the effects of tightening will be felt most.

This shows the contribution to annualized inflation above the 2% target, over rolling three-month periods. My analysis of CPI data.

The third point is that all of this takes time. It is true that some asset prices and market interest rates may move as soon as the Fed funds rate changes — or even in advance of the actual change, as with mortgage rates this year. But the translation from this to real activity is much slower. The Fed’s own FRB/US model says that the peak effect of a rate change comes about two years later; there are significant effects out to the fourth year. What the Fed is doing now is, in an important sense, setting policy for the year 2024 or 2025. How  confident should we be about what demand conditions will look like then? Given how few people predicted current inflation, I would say: not very confident.

This connects to the fourth point, which is that there is no reason to think that the Fed can deliver a smooth, incremental deceleration of demand. (Assuming we agreed that that’s what’s called for.) In part this is because of the lags just mentioned. The effects of tightening are felt years in the future, but the Fed only gets data in real time. The Fed may feel they’ve done enough once they see unemployment start to rise. But by that point, they’ll have baked several more years of rising unemployment into the economy. It’s quite possible that by the time the full effects of the current round of tightening are felt, the US economy will be entering a recession. 

This is reinforced when we think about the channels policy actually works through. Empirical studies of investment spending tend to find that it is actually quite insensitive to interest rates. The effect of hikes, when it comes, is likelier to be through Minskyan channels — at some point, rising debt service costs and falling asset values lead to a cascading chain of defaults.

In and Out of the Corridor

A broader reason we should doubt that the Fed can deliver a glide path to slower growth is that the economy is a complex system, with both positive and negative feedbacks; which feedbacks dominate depends on the scale of the disturbance. In practice, small disturbances are often self-correcting; to have any effect, a shock has to be big enough to overcome this homeostasis.

Axel Leijonhufvud long ago described this as a “corridor of stability”: economic units have buffers in the form of liquid assets and unused borrowing capacity, which allow them to avoid adjusting expenditure in response to small changes in income or costs. This means the Keynesian multiplier is small or zero for small changes in autonomous demand. But once buffers start to get exhausted, responses become much larger, as the income-expenditure positive feedback loop kicks in.

The most obvious sign of this is the saw-tooth pattern in long-run series of employment and output. We don’t see smooth variation in growth rates around a trend. Rather, we see two distinct regimes: extended periods of steady output and employment growth, interrupted by shorter periods of negative growth. Real economies experience well-defined expansions and recessions, not generic “fluctuations”.

This pattern is discussed in a very interesting recent paper by Antonio Fatas, “The Elusive State of Full Employment.” The central observation of the paper is that whether you measure labor market slack by the conventional unemployment rate or in some other way (the detrended prime-age employment-population ratio is his preferred measure), the postwar US does not show any sign of convergence back to a state of full employment. Rather, unemployment falls and employment rises at a more or less constant rate over an expansion, until it abruptly gives way to a recession. There are no extended periods in which (un)employment rates remain stable.

One implication of this is that the economy spends very little time at potential or full employment; indeed, as he says, the historical pattern should raise questions whether a level of full employment is meaningful at all.

the results of this paper also cast doubt on the empirical relevance of the concepts of full employment or the natural rate of unemployment. … If this interpretation is correct, our estimates of the natural rate of unemployment are influenced by the length of expansions. As an example, if the global pandemic had happened in 2017 when unemployment was around 4.5%, it is very likely that we would be thinking of unemployment rates as low as 3.5% as unachievable.

There are many ways of arriving at this same point. For example, he finds that the (un)employment rate at the end of an expansion is strongly predicted by the rate at the beginning, suggesting that what we are seeing is not convergence back to an equilibrium but simply a process of rising employment that continues until something ends it.

Another way of looking at this pattern is that any negative shock large enough to significantly slow growth will send it into reverse — that, in effect, growth has a “stall speed” below which it turns into recession. If this weren’t the case, we would sometimes see plateaus or gentle hills in the employment rate. But all we see are sharp peaks. 

In short: Monetary policy is an anti-inflation tool that works, when it does, by lowering employment and wages; by reducing spending in a few interest-sensitive sectors of the economy, which may have little overlap with those where prices are rising; whose main effects take longer to be felt than we can reasonably predict demand conditions; and that is more likely to provoke a sharp downturn than a gradual deceleration.

Is Macroeconomic Policy the Responsibility of the Fed?

One reason I don’t think we should be endorsing this move is that we shouldn’t be endorsing the premise that the US is facing dangerously overheated labor markets. But the bigger reason is that conventional monetary policy is a bad way of managing the economy, and entails a bad way of thinking about the economy. We should not buy into a framework in which problems of rising prices or slow growth or high unemployment get reduced to “what should the federal funds rate do?”

Here for example is former CEA Chair Jason Furman’s list of ways to reduce inflation:

What’s missing here is any policy action by anyone other than the Fed. It’s this narrowing of the discussion I object to, more than the rate increase as such.

Rents are rising rapidly right now — at an annual rate of about 6 percent as measured by the CPI. And there is reason to think that this number understates the increase in market rents and will go up rather than down over the coming year. This is one factor in the acceleration of inflation compared with 2020, when rents in most of the country were flat or falling. (Rents fell almost 10 percent in NYC during 2020, per Zillow.) The shift from falling to rising rents is an important fact about the current situation. But rents were also rising well above 2 percent annually prior to the pandemic. The reason that rents (and housing prices generally) rise faster than most other prices generally, is that we don’t build enough housing. We don’t build enough housing for poor people because it’s not profitable to do so; we don’t build enough housing for anyone in major cities because land-use rules prevent it. 

Rising rents are not an inflation problem, they are a housing problem. The only way to deal with them is some mix of public money for lower-income housing, land-use reform, and rent regulations to protect tenants in the meantime. Higher interest rates will not help at all — except insofar as, eventually, they make people too poor to afford homes.

Or energy costs. Energy today still mostly means fossil fuels, especially at the margin. Both supply and demand are inelastic, so prices are subject to large swings. It’s a global market, so there’s not much chance of insulating the US even if it is “energy independent” in net terms. The geopolitics of fossil fuels means that production is both vulnerable to interruption from unpredictable political developments, and subject to control by cartels. 

The long run solution is, of course, to transition as quickly as possible away from fossil fuels. In the short run, we can’t do much to reduce the cost of gasoline (or home heating oil and so on), but we can shelter people from the impact, by reducing the costs of alternatives, like transit, or simply by sending them checks. (The California state legislature’s plan seems like a good model.) Free bus service will help both with the short-term effect on household budgets and to reduce energy demand in the long run. Raising interest rates won’t help at all — except insofar as, eventually, they make people too poor to buy gas.

These are hard problems. Land use decisions are made across tens of thousands of local governments, and changes are ferociously opposed by politically potent local homeowners (and some progressives). Dependence on oil is deeply baked into our economy. And of course any substantial increase in federal spending must overcome both entrenched opposition and the convoluted, anti-democratic structures of our government, as we have all been learning (again) this past year. 

These daunting problems disappear when we fold everything into a price index and hand it over to the Fed to manage. Reducing everything to the core CPI and a policy rule are a way of evading all sorts of difficult political and intellectual challenges. We can also then ignore the question how, exactly, inflation will be brought down without costs to the real economy,  and how to decide if these costs are worth it. Over here is inflation; over there are the maestros with their magic anti-inflation device. All they have to do is put the right number into the machine.

It’s an appealing fantasy – it’s easy to see why people are drawn to it. But it is a fantasy.

A modern central bank, sitting at the apex of the financial system, has a great deal of influence over markets for financial assets and credit. This in turn allows it to exert some influence — powerful if often slow and indirect — on production and consumption decisions of businesses and households. Changes in the level and direction of spending will in turn affect the pricing decisions of business. These effects are real. But they are no different than the effects of anything else — public policy or economic developments — that influence spending decisions. And the level of spending is in turn only one factor in the evolution of prices. There is no special link from monetary policy to aggregate demand or inflation. It’s just one factor among others — sometimes important, often not.

Yes, a higher interest rate will, eventually reduce spending, wages and prices. But many other forces are pushing in other directions, and dampening or amplifying the effect of interest rate changes. The idea that there is out there some “r*”, some “neutral rate” that somehow corresponds to the true inter temporal interest rate — that is a fairy tale

Nor does the Fed have any special responsibility for inflation. Once we recognize monetary policy for what it is — one among many regulatory and tax actions that influence economic rewards and incomes, perhaps influencing behavior — arguments for central bank independence evaporate. (Then again, they did not make much sense to begin with.) And contrary to widely held belief, the Fed’s governing statutes do not give it legal responsibility for inflation or unemployment. 

That last statement might sound strange, given that we are used to talking about the Fed’s dual mandate. But as Lev Menand points out in an important recent intervention, the legal mandate of the Fed has been widely misunderstood. What the Federal Reserve Act charges the Fed with is

maintain[ing the] long run growth of the monetary and credit aggregates commensurate with the economy’s long-run potential to increase production, so as to promote effectively the goals of maximum employment, stable prices, and moderate long-term interest rates.

There are two things to notice here. First, the bolded phrase: The Fed’s mandate is not to maintain price stability or full employment as such. It is to prevent developments in the financial system that interfere with them. This is not the same thing. And as Menand argues (in the blog post and at more length elsewhere), limiting the Fed’s macroeconomic role to this narrower mission was the explicit intent of the lawmakers who wrote the Fed’s governing statutes from the 1930s onward. 

Second, price stability, maximum employment and moderate interest rates (an often forgotten part of the Fed’s mandate) are not presented as independent objectives, but as the expected consequences of keeping credit growth on a steady path. As Menand writes:

The Fed’s job, as policymakers then recognized, was not to combat inflation—it was to ensure that banks create enough money and credit to keep the nation’s productive resources fully utilized…

This distinction is important because there are many reasons that, in the short-to-medium term, the economy might not achieve full potential—as manifested by maximum employment, price stability, and moderate long-term interest rates. And often these reasons have nothing to do with monetary expansion, the only variable Congress expected the Fed to control. For example, supply shortages of key goods and services can cause prices to rise for months or even years while producers adapt to satisfy changing market demand. The Fed’s job is not to stop these price rises—even if policymakers might think stopping them is desirable—just as the Fed’s job is not to … lend lots of money to companies so that they can hire more workers. The Fed’s job is to ensure that a lack of money and credit created by the banking system—an inelastic money supply—does not prevent the economy from achieving these goals. That is its sole mandate.

As Menand notes, the idea that the Fed was directly responsible for macroeconomic outcomes was a new development in the 1980s, an aspect of the broader neoliberal turn that had no basis in law. Nor does it have any good basis in economics. If a financial crisis leads to a credit crunch, or credit-fueled speculation develops into an asset bubble, the central bank can and should take steps to stabilize credit growth and asset prices. In doing so, it will contribute to the stability of the real economy. But when inflation or unemployment come from other sources, conventional monetary policy is a clumsy, ineffectual and often destructive way of responding to them. 

There’s a reason that the rightward turn in the 1980s saw the elevation of central banks as the sole custodians of macroeconomic stability. The economies we live in are not in fact self-regulating; they are subject to catastrophic breakdowns of various forms, and even when they function well, are in constant friction with their social surroundings. They require active management. But routine management of the economy — even if limited to the adjustment of the demand “thermostat,” in Samuelson’s old metaphor — both undermine the claim that markets are natural, spontaneous and decentralized, and opens the door to a broader politicization of the economy. The independent central bank in effect quarantines the necessary economic management from the infection of democratic politics. 

The period between the 1980s and the global financial crisis saw both a dramatic elevation of the central bank’s role in macroeconomic policy, and a systematic forgetting of the wide range of tools central banks used historically. There is a basic conflict between the expansive conception of the central bank’s responsibilities and the narrow definition of what it actually does. The textbooks tell us that monetary policy is the sole, or at least primary, tool for managing output, employment and inflation (and in much of the world, the exchange rate); and that it is limited to setting a single overnight interest rate according to a predetermined rule. These two ideas can coexist comfortably only in periods of tranquility when the central bank doesn’t actually have to do anything. 

What has the Fed Delivered in the Past?

Coming back to the present: The reason I think it is wrong to endorse the Fed’s move toward tightening is not that there’s any great social benefit to having an overnight rate on interbank loans of near 0. I don’t especially care whether the federal funds rate is at 0.38 percent or 1.17 percent next September. I don’t think it makes much difference either way. What I care about is endorsing a framework that commits us to managing inflation by forcing down wages, one that closes off discussion of more progressive and humane — and effective! — ways of controlling inflation. Once the discussion of macroeconomic policy is reduced to what path the federal funds rate should follow, our side has already lost, whatever the answer turns out to be.

It is true that there are important differences between the current situation the end of 2015, the last time the Fed started hiking, that make today’s tightening more defensible. Headline unemployment is now at 3.8 percent, compared with 5 percent when the Fed began hiking in 2015. The prime-age employment rate was also about a point lower then than now. But note also that in 2015 the Fed thought the long-run unemployment rate was 4.9 percent. So from their point of view, we were at full employment. (The CBO, which had the long-run rate at 5.3 percent, thought we’d already passed it.) It may be obvious in retrospect (and to some of us in the moment) that in late 2015 there was still plenty of space for continued employment growth. But policymakers did not think so at the time.

More to the point, inflation then was much lower. If inflation control is the Fed’s job, then the case for raising rates is indeed much stronger now than it was in December 2015. And while I am challenging the idea that this should be the Fed’s job, most people believe that it is. I’m not upset or disappointed that Powell is moving to hike rates now, or is justifying it in the way that he is. Anyone who could plausibly be in that position would be doing the same. 

So let’s say a turn toward higher rates was less justified in 2015 than it is today. Did it matter? If you look at employment growth over the 2010s, it’s a perfectly straight line — an annual rate of 1.2 percent, month after month after month. If you just looked at the employment numbers, you’d have no idea that the the Fed was tightening over 2016-2018, and then loosening in the second half of 2019. This doesn’t, strictly speaking, prove that the tightening had no effect. But that’s certainly the view favored by Occam’s razor. The Fed, fortunately, did not tighten enough to tip the economy into recession. So it might as well not have tightened at all. 

The problem in 2015, or 2013, or 2011, the reason we had such a long and costly jobless recovery, was not that someone at the Fed put the wrong parameter into their model. It was not that the Fed made the wrong choices. It was that the Fed did not have the tools for the job.

Honestly, it’s hard for me to see how anyone who’s been in these debates over the past decade could believe that the Fed has the ability to steer demand in any reliable way. The policy rate was at zero for six full years. The Fed was trying their best! Certainly the Fed’s response to the 2008 crisis was much better than the fiscal authorities’. So for that matter was the ECB’s, once Draghi took over from Trichet. 4 The problem was not that the central bankers weren’t trying. The problem was that having the foot all the way down on the monetary gas pedal turned out not to do much.

As far as I can tell, modern US history offers exactly one unambiguous case of successful inflation control via monetary policy: the Volcker shock. And there, it was part of a comprehensive attack on labor

It is true that recessions since then have consistently seen a fall in inflation, and have consistently been preceded by monetary tightenings. So you could argue that the Fed has had some inflation-control successes since the 1980s, albeit at the cost of recessions. Let’s be clear about what this entails. To say that the Fed was responsible for the fall in inflation over 2000-2002, is to say that the dot-com boom could have continued indefinitely if the Fed had not raised rates. 

Maybe it could have, maybe not. But whether or not you want to credit (or blame) the Fed for some or all of the three pre-pandemic recessions, what is clear is that there are few if any cases of the Fed delivering slower growth and lower inflation without a recession. 

According to Alan Blinder, since World War II the Fed has achieved a soft landing in exactly two out of 11 tightening cycles, most recently in 1994. In that case, it’s true, higher rates were not followed by a recession. But nor were they followed by any discernible slowdown in growth. Output and employment grew even faster after the Fed started tightening than before. As for inflation, it did come down about two years later, at the end of 1996 – at exactly the same moment as oil prices peaked. And came back up in 1999, at exactly the moment when oil prices started rising again. Did the Fed do that? It looks to me more like 2015 – a tightening that stopped in time to avoid triggering a recession, and instead had no effect. But even if we accept the 1994 case, that’s one success story in the past 50 years. (Blinder’s other soft landing is 1966.)

I think the heart of my disagreement with progressives who are support tightening is whether it’s reasonable to think the Fed can adjust the “angle of approach” to a higher level of employment. I don’t think history gives us much reason to believe that they can. There are people who think that a recession, or at least a much weaker labor market, is the necessary cost of restoring price stability. That’s not a view I share, obviously, but it is intellectually coherent. The view that the Fed can engineer a gentle cooling that will bring down inflation while employment keeps rising, on the other hand, seems like wishful thinking.

That said, of the two realistic outcomes of tightening – no effect, or else a crisis – I think the first is more likely, unless they move quite a bit faster than they are right now. 

So what’s at stake then? If the Fed is doing what anyone in their position would do, and if it’s not likely to have much impact one way or another, why not make some approving noises, bank the respectability points, and move on? 

Four Good Reasons to Be Against Rate Hikes (and One that Isn’t)

I think that it’s a mistake to endorse or support monetary tightening. I’ll end this long post by summarizing my reasons. But first, let me stress that a commitment to keeping the federal funds rate at 0 is not one of those reasons. If the Fed were to set the overnight rate at some moderate positive level and then leave it there, I’d have no objection. In the mid-19th century, the Bank of France kept its discount rate at exactly 4 percent for something like 25 years. Admittedly 4 percent sounds a little high for the US today. But a fixed 2 percent for the next 25 years would probably be fine.

There are four reasons I think endorsing the Fed’s decision to hike is a mistake.

  1. First, most obviously, there is the risk of recession. If rates were at 2 percent today, I would not be calling for them to be cut. But raising them is a different story. Last week’s hike is no big deal in itself, but there will be another, and another, and another. I don’t know where the tipping point is, where hikes inflict enough financial distress to tip the economy into recession. But neither does the Fed. The faster they go, the sooner they’ll hit it. And given the long lags in monetary transmission, they probably won’t know until it’s too late. People are talking a lot lately about wage-price spirals, but that is far from the only positive feedback in a capitalist economy. Once a downturn gets started, with widespread business failures, defaults and disappointed investment plans, it’s much harder to reverse it than it would have been to maintain growth. 

I think many people see trusting the Fed to deal with inflation as the safe, cautious position. But the fact that a view is widely held doesn’t mean it is reasonable. It seems to me that counting on the Fed to pull off something that they’ve seldom if ever succeeded at before is not safe or cautious at all.5 Those of us who’ve been critical of rate hikes in the past should not be too quick to jump on the bandwagon now. There are plenty of voices calling on the Fed to move faster. It’s important that there also be some saying, slow down. 

2. Second, related to this, is a question I think anyone inclined to applaud hikes should be asking themselves: If high inflation means we need slower growth, higher unemployment and lower wages, where does that stop? Inflation may come down on its own over the next year — I still think this is more likely than not. But if it doesn’t come down on its own, the current round of rate hikes certainly isn’t going to do it. Looking again at the Fed’s FRB/US model, we see that a one point increase in the federal funds rate is  predicted to reduce inflation by about one-tenth of a point after one year, and about 0.15 points after two years. The OECD’s benchmark macro model make similar predictions: a sustained one-point increase in the interest rate in a given year leads to an 0.1 point fall in inflation the following year, an 0.3 fall in the third year and and an 0.5 point fall in the fourth year.

Depending which index you prefer, inflation is now between 3 and 6 points above target.6 If you think conventional monetary policy is what’s going to fix that, then either you must have have some reason to think its effects are much bigger than the Fed’s own models predict, or you must be imagining much bigger hikes than what we’re currently seeing. If you’re a progressive signing on to today’s hikes, you need to ask yourself if you will be on board with much bigger hikes if inflation stays high. “I hope it doesn’t come to that” is not an answer.

3. Third, embracing rate hikes validates the narrative that inflation is now a matter of generalized overheating, and that the solution has to be some form of across-the-board reduction in spending, income and wages. It reinforces the idea that pandemic-era macro policy has been a story of errors, rather than, on balance, a resounding success.

The orthodox view is that low unemployment, rising wages, and stronger bargaining power for workers are in themselves serious problems that need to be fixed. Look at how the news earlier this week of record-low unemployment claims got covered: It’s a dangerous sign of “wage inflation” that will “raise red flags at the Fed.”  Or the constant complaints by employers of “labor shortages” (echoed by Powell last week.) Saying that we want more employment and wage growth, just not right now, feels like trying to split the baby. There is not a path to a higher labor share that won’t upset business owners.

The orthodox view is that a big reason inflation was so intractable in the 1970s was that workers were also getting large raises. From this point of view, if wages are keeping pace with inflation, that makes the problem worse, and implies we need even more tightening. Conversely, if wages are falling behind, that’s good. Alternatively, you might think that the Powell was right before when he said the Phillips curve was flat, and that inflation today has little connection with unemployment and wages. In that case faster wage growth, so that living standards don’t fall, is part of the solution not the problem. Would higher wages right now be good, or bad? This is not a question on which you can be agnostic, or split the difference. I think anyone with broadly pro-worker politics needs to think very carefully before they accept the narrative of a wage-price spiral as the one thing to be avoided at all costs.

Similarly, if rate hikes are justified, then so must be other measures to reduce aggregate spending. The good folks over at the Committee for a Responsible Federal Budget just put out a piece arguing that student loan forbearance and expanded state Medicare and Medicaid funding ought to be ended, since they are inflationary. And you have to admit there’s some logic to that. If we agree that the economy is suffering from excessive demand, shouldn’t we support fiscal as well as monetary measures to reduce it? A big thing that rate hikes will do is raise interest payments by debtors, including student loan debtors. If that’s something we think ought to happen, we should think so when it’s brought about in other ways too. Conversely, if you don’t want to sign on to the CFRB program, you probably want to keep some distance from Powell.

4. Fourth and finally, reinforcing the idea that inflation control is the job of the Fed undermines the case for measures that actually would help with inflation. Paradoxical as it may sound, one reason it’s a mistake to endorse rate hikes is precisely because rising prices really are a problem. High costs of housing and childcare are a major burden for working families. They’re also a major obstacle to broader social goals (more people living in dense cities; a more equal division of labor within the family). Rate hikes move us away from the solution to these problems, not towards it. Most urgently and obviously, they are entirely unhelpful in the energy transition. Tell me if you think this is sensible: “Oil prices are rising, so we should discourage people from developing alternative energy sources”. But that is how conventional monetary policy works. 

The Biden administration has been strikingly consistent in articulating an alternative vision of inflation control – what some people call a progressive supply-side vision. In the State of the Union, for example, we heard:

We have a choice. One way to fight inflation is to drive down wages and make Americans poorer. I think I have a better idea … Make more cars and semiconductors in America. More infrastructure and innovation in America. …

First, cut the cost of prescription drugs. We pay more for the same drug produced by the same company in America than any other country in the world. Just look at insulin. … Insulin costs about $10 a vial to make. … But drug companies charge … up to 30 times that amount. …. Let’s cap the cost of insulin at $35 a month so everyone can afford it.7

Second, cut energy costs for families an average of $500 a year by combating climate change. Let’s provide investment tax credits to weatherize your home and your business to be energy efficient …; double America’s clean energy production in solar, wind and so much more; lower the price of electric vehicles,…

Of course weatherizing homes is not, by itself, going to have a big effect on inflation. But that’s the direction we should be looking in. If we’re serious about managing destructive price increases, we can’t leave the job to the Fed. We need to be looking for a mix of policies that directly limit price increases using  administrative tools, that cushion the impact of high prices on family budgets in the short run, and that deal with the supply constraints driving price increases in the long run. 

The interest rate hike approach is an obstacle to all this, both practically and ideologically. A big reason I’m disappointed to see progressives accepting  the idea that inflation equals rate hikes, is that there has been so much creative thinking about macroeconomic policy in recent years. What’s made this possible is increasing recognition that the neoliberal, central bank-centered model has failed. We have to decide now if we really believed that. Forward or backward? You can’t have it both ways.

Today’s Inflation Won’t be Solved by the Fed

(This post originally ran as an opinion piece in Barron’s.)

The U.S. today is experiencing inflation. This is not controversial. But what exactly does it mean?

In the textbook, inflation is a rise in all prices together, caused by an excessive increase in the money supply. But when we measure it, inflation is just a rise in the average price of goods and services. That average might reflect a uniform rise in prices due to excessive money creation. Or, as today, it might instead be the result of big rises in the prices of a few items, for their own reasons.

Over the past year, prices have risen by 7.5%, far above the usual 2% target set by the Federal Reserve. But 70% of that 5.5 points of excess inflation has come from two categories that make up just 15% of the consumption basket: energy (2 points) and new and used cars (1.9 points). Used cars alone make up barely 4% of the consumption basket, but accounted for a third of the excess inflation.

Some commentators have argued that inflation is just a matter of too much money. If that were true, it’s hard to see why so much of it would be flowing to cars. (And before you say cheap financing: Rates on auto loans were lower through most of the 2010s.)

In recent months, vehicle and energy prices have begun to stabilize, while food and housing prices have picked up. These price increases hit family budgets harder. A car purchase can usually be put off, but not rent or groceries. But this is still a story about specific sectors following their own dynamics.

Energy prices are global, and their periodic rise and fall depends mostly on the politics of oil-producing regions (as we are being reminded today). As recently as the summer of 2014, gas prices were higher than they are now, before falling precipitously. No doubt they will fall again, but in the short run there is not much to do about them—though it may be possible to shield people from their impact. In the longer run, decarbonization will leave us less vulnerable to the gyrations of the oil market.

As for vehicles, it’s no mystery why prices soared. Early in the pandemic, automakers expected a long period of depressed demand, and cut back production plans. When the economy bounced back rapidly, automakers found themselves short of key inputs, especially semiconductors. Combine this with a pandemic-induced shift in demand from services to goods, and you have a formula for rapid price increases. The effect was strongest for used cars, whose supply is essentially fixed in the short run.

Housing has made a smaller contribution so far—0.6 of the 5.5 points of excess inflation—but given the way the Bureau of Labor Statistics measures them, housing prices are likely to rise sharply over the coming year. This is a problem. But, it was also a big problem before the pandemic, when rents were rising by nearly 4% annually. Housing affordability is a serious issue in the U.S. But if the question is why inflation is higher today than in 2018 or 2019, housing is not the answer.

Finally, there are food prices, which have contributed about 0.7 points to excess inflation over the past year, and more in recent months. Food prices, like energy prices, are famously volatile; there’s a reason they are both excluded from the Fed’s measure of “core” inflation. They’re also an area where market power may be playing a major role, given the high concentration in food processing. Monopolies may be reluctant to fully exploit their power in normal times; price increases elsewhere in the economy give them a chance to widen their margins.

The great majority of the excess inflation over the past year has come from these four areas. Other sectors—including labor-intensive services where prices have historically risen more quickly—have contributed little or nothing.

The point is not that these price increases don’t matter. Food, housing and energy are necessities of life. People are naturally unhappy when they have to pay more for them. The point is that current price rises are not symptoms of economy-wide overheating.

Some of these prices, like autos, will come back down on their own as supply-chain kinks work themselves out. Others, like housing, will not, and call for a policy response. But that response is not raising interest rates, which would only make the problem worse. The main reason why housing costs are rising is that the U.S. does not build enough of it, especially in the expensive metro areas where employment opportunities are concentrated. Construction is one of the most interest-sensitive sectors of the economy. Rate hikes will cause supply to fall further short of demand.

Some might say that the Fed still controls the overall level of spending in the economy. If people spent less on used cars, wouldn’t they spend more on something else? This ignores the existence of balance sheets. Households hold cash, and finance many purchases—including cars—with debt. Lower used-car prices wouldn’t mean higher prices elsewhere, but higher household savings and less debt.

An inability to build housing where people want to live, dependence on fossil fuels, fragile supply chains and the monopolization of key industries: These are all serious economic problems. But they are not monetary-policy problems. Looking at them through the lens of a textbook story of inflation will not get us any closer to solving them.

 

No Maestros: Further Thoughts

One of the things we see in the questions of monetary policy transmission discussed in my Barron’s piece is the real cost of an orthodox economics education. If your vision of the economy is shaped by mainstream theory, it is impossible to think about what central banks actually do.

The models taught in graduate economics classes feature an “interest rate” that is the price of goods today in terms of identical goods in the future. Agents in these models are assumed to be able to freely trade off consumption today against consumption at any point in the future, and to distribute income from any time in the future over their lifetime as they see fit, subject only to the “no Ponzi” condition that over infinite time their spending must equal their income. This is a world, in other words, of infinite liquidity. There are no credit markets as such, only real goods at different dates.1

Monetary policy in this framework is then thought of in terms of changing the terms at which goods today trade for goods tomorrow, with the goal of keeping it at some “natural” level. It’s not at all clear how the central bank is supposed to set the terms of all these different transactions, or what frictions cause the time premium to deviate from the natural level, or whether the existence of those frictions might have broader consequences. 2 But there’s no reason to get distracted by this imaginary world, because it has nothing at all to do with what real central banks do.

In the real world, there are not, in general, markets where goods today trade for identical goods at some future date. But there are credit markets, which is where the price we call “the interest rate” is found. The typical transaction in a credit market is a loan — for example, a mortgage. A mortgage does not involve any trading-off of future against present income. Rather, it is income-positive for both parties in every period.

The borrower is getting a flow of housing services and making a flow of mortgage payments, both of which are the same in every period. Presumably they are getting more/better housing services for their mortgage payment than they would for an equivalent rental payment in every period (otherwise, they wouldn’t be buying the house.) Far from getting present consumption at the expense of future consumption, the borrower probably expects to benefit more from owning the house in the future, when rents will be higher but the mortgage payment is the same.

The bank, meanwhile, is getting more income in every period from the mortgage loan than it is paying to the holder of the newly-created deposit. No one associated with the bank is giving up any present consumption — the loan just involves creating two offsetting entries on the bank’s books. Both parties to the transaction are getting higher income over the whole life of the mortgage.

So no one, in the mortgage transaction, is trading off the present against the future. The transaction will raise the income of both sides in every period. So why not make more mortgages to infinity? Because what both parties are giving up in exchange for the higher income is liquidity. For the homeowner, the mortgage payments yield more housing services than equivalent rent payments, but they are also harder to adjust if circumstances change. Renting gives you less housing for your buck, but it’s easier to move if it turns out you’d rather live somewhere else. For the bank, the mortgage loan (its asset) carries a higher interest rate than the deposit (its liability), but involves the risk that the borrower will not repay, and also the risk that, in a crisis, ownership of the mortgage cannot be turned into immediate cashflows while the deposit is payable on demand.

In short, the fundamental tradeoff in credit markets – what the interest rate is the price of – is not less now versus more later, but income versus liquidity and safety.3

Money and credit are hierarchical. Bank deposits are an asset for us – they are money – but are a liability for banks. They must settle their own transactions with a different asset, which is a liability for the higher level of the system. The Fed sits at the top of this hierarchy. That is what makes its actions effective. It’s not that it can magically change the terms of every transaction that involves things happening at different dates. It’s that, because its liabilities are what banks use to settle their obligations to each other, it can influence how easy or difficult they find it to settle those liabilities and hence, how willing they are to take on the risk of expanding their balance sheets.

So when we think about the transmission of monetary policy, we have to think about two fundamental questions. First, how much do central bank actions change liquidity conditions within the financial system? And second, how much does real activity depends on the terms on which credit is available?

We might gloss this as supply and demand for credit. The mortgage, however, is typical of credit transactions in another way: It involves a change in ownership of an existing asset rather than the current production of goods and services. This is by far the most common case. So some large part of monetary policy transmission is presumably via changes in prices of assets rather than directly via credit-financed current production. 4 There are only small parts of the economy where production is directly sensitive to credit conditions.

One area where current production does seem to be sensitive to interest rates is housing construction. This is, I suppose, because on the one hand developers are not large corporations that can finance investment spending internally, and on the other hand land and buildings are better collateral than other capital goods. My impression – tho I’m getting well outside my area of expertise here – is that some significant part of construction finance is shorter maturity loans, where rates will be more closely linked to the policy rate. And then of course the sale price of the buildings will be influenced by prevailing interest rates as well. As a first approximation you could argue that this is the channel by which Fed actions influence the real economy. Or as this older but still compelling article puts it, “Housing IS the business cycle.

Of course there are other possible channels. For instance, it’s sometimes argued that during the middle third of the 20th century, when reserve requirements really bound, changes in the quantity of reserves had a direct quantitative effect on the overall volume of lending, without the interest rate playing a central role one way or the other. I’m not sure how true this is — it’s something I’d like to understand better — but in any case it’s not relevant to monetary policy today. Robert Triffin argued that inventories of raw materials and imported commodities were likely to be financed with short term debt, so higher interest rates would put downward pressure on their prices specifically. This also is probably only of historical interest.

The point is, deciding how much, how quickly and how reliably changes in the central bank’s policy rate will affect real activity (and then, perhaps, inflation) would seem to require a fairly fine-grained institutional knowledge about the financial system and the financing needs of real activity. The models taught in graduate macroeconomics are entirely useless for this purpose. Even for people not immersed in academic macro, the fixation on “the” interest rate as opposed to credit conditions broadly is a real problem.

These are not new debates, of course. I’ve linked before to Juan Acosta’s fascinating article about the 1950s debates between Paul Samuelson and various economists associated with the Fed.5 The lines of debate then were a bit different from now, with the academic economists more skeptical of monetary policy’s ability to influence real economic outcomes. What Fed economist Robert Roosa seems to have eventually convinced Samuelson of, is that monetary policy works not so much through the interest rate — which then as now didn’t seem to have big effect on investment decision. It works rather by changing the willingness of banks to lend — what was then known as “the availability doctrine.” This is reflected in later editions of his textbook, which added an explanation of monetary policy in terms of credit rationing.

Even if a lender should make little or no change in the rate of interest that he advertises to his customers, there may probably still be the following important effect of “easy money.” …  the lender will now be rationing out credit much more liberally than would be the case if the money market were very tight and interest rates were tending to rise. … Whenever in what follows I speak of a lowering of interest rates, I shall also have in mind the equally important relaxation of the rationing of credit and general increase in the availability of equity and loan capital to business.

The idea that “the interest rate” is a metaphor or synecdoche for a broader easing of credit conditions is important step toward realism. But as so often happens, the nuance has gotten lost and the metaphor gets taken literally.

At Barron’s: There Are No Maestros

(A week ago, I had an opinion piece in Barron’s, which I am belatedly posting here. I talk a bit more about this topic in the following post.)

In today’s often acrimonious economic debates, one of the few common grounds is reverence for the Fed. Consider Jay Powell: First nominated to the Fed’s board of governors by President Obama, he was elevated to FOMC chair by Trump and renominated by Biden His predecessors Bernanke, Greenspan and Volcker were similarly first appointed by a president from one party, then reappointed by a president from the other. Politics stops at Maiden Lane.

There are disagreements about what the Fed should be doing — tightening policy to rein in inflation, or holding back to allow for a faster recovery. But few doubt that it’s the Fed’s job to make the choice, and that once they do, they can carry it out.

Perhaps, though, we should take a step back and ask if the Fed is really all-powerful. You might like to see inflation come down; I’d like to see stronger labor markets. But can the Fed give either of us what we want?

During the so-called Great Moderation, it was easy to have faith in the Fed. In the US, as in most rich countries, governments had largely turned over the job of macroeconomic management to independent central banks, and were enjoying an era of stable growth with low inflation. Magazine covers could, without irony, feature the Fed chair as  “Pope Greenspan and His College of Cardinals,” or (when the waters got choppier) the central figure in the “committee to save the world.”

Respectable opinion of the 1990s and 2000s was captured in a speech by Christina Romer (soon to be Obama’s chief economist), declaring that “the Federal Reserve, is directly responsible for the low inflation and the virtual disappearance of the business cycle in the last 25 years. …The story of stabilization policy of the last quarter century is one of amazing success.”

Romer delivered those words in late 2007. At almost exactly that moment, the US was entering its then-deepest recession since World War II.

The housing bubble and financial crisis raised some doubts about whether that success had been so amazing after all. The subsequent decade of slow growth and high unemployment, in the face of a Fed Funds rate of zero and multiple rounds of QE, should have raised more. Evidently the old medicine was no longer working – or perhaps had never worked as well as we thought.

In truth, there were always reasons for doubt.

One is that, as Milton Friedman famously observed, monetary policy acts with long and variable lags. A common  estimate is that the peak impact of monetary policy changes comes 18 to 24 months later, which is cripplingly slow for managing business cycles. Many people – including at the Fed – believe that today’s inflation is the transitory result of the pandemic. When the main effects of today’s tightening are felt two years from now, how confident are we that inflation will still be too high?

More fundamentally, there’s the question of what links monetary policy to inflation in the first place. Prices are, after all, set by private businesses; if they think it makes sense to raise prices, the Fed has no mind-control ray to convince them otherwise.

In the textbook story, changes in the Federal funds rate are passed through to other interest rates. A higher cost of borrowing discourages investment spending, reducing demand, employment and wages, which in turn puts downward pressure on prices. This was always a bit roundabout; today, it’s not clear that critical links in the chain function at all.

Business investment is financed with long-term debt; the average maturity of a corporate bond is about 13 years. But long rates don’t seem particularly responsive to the Federal funds rate. Between Fall 2015 and Spring 2019, for example, the Fed raised its policy rate by 2.5 points. Over this same period, the 10-year Treasury rate was essentially unchanged, and corporate bond yields actually fell. Earlier episodes show a similar non-response of long rates to Fed actions.

Nor is it obvious that business investment is particularly sensitive to interest rates, even long ones. One recent survey of the literature by Fed economists finds that hurdle rates for new investment “exhibit no apparent relation to market interest rates.”

Former Fed chair Ben Bernanke puzzled over “the black box” of monetary policy transmission. If it doesn’t move interest rates on the long-term debt that businesses mostly issue, and if even longer rates have no detectable effect on investment, how exactly is monetary policy affecting demand and inflation? It was a good question, to which no one has offered a very good answer.

To be sure, no one would claim that the Fed is powerless. Raise rates enough, and borrowers unable to roll over their loans will face default; as asset values fall and balance sheets weaken, households will have no choice but to drastically curtail consumption.

But being able to sink a ship is not the same as being able to steer it. The fact that the Fed can, if it tries hard enough, trigger a recession, does not mean that it can maintain steady growth. Perhaps it’s time to admit that there are no central banking “maestros” who know the secret of maintaining full employment and price stability. Balancing these critical social objectives requires a variety of tools, not just a single interest rate. And it is, for better or worse, the responsibility of our elected governments.