Wednesday, October 31, 2012

When is it Time to Take a Vacation?

Answer: When Matt Yglesias calls you a "hero of rigor."

Kocherlakota Joins the COGCB

What's the COGCB? Club of Goofy Central Bankers. Case in point: his speech yesterday. Kocherlakota tells us, if we didn't already know, that
...the Fed’s policy stance is considerably more accommodative than it was five years ago.
Of course, in terms of what the Fed says it cares about (the twelve-month rate of increase in the pce deflator and the unemployment rate), the Fed has been considerably more accommodative than it would have been, pre-2008, if the same state of the world had occurred, as I pointed out here. Thus, the behavior of the FOMC has changed. Either it cares about some things it did not care about before (and in a particular way), or it cares about the same things in different ways - in particular it is less concerned about its price stability mandate.

The Fed may have good reasons for changing its behavior. If so, Fed officials should articulate those reasons in public, so that we can debate the issues. That seems to be what Kocherlakota is attempting here, and he goes even further than you might expect. His conclusion is:
...monetary policy is, if anything, too tight, not too easy.
If you have not fainted and fallen on the floor, take a couple of deep breaths, and we'll figure out what Narayana has on his mind. His argument is:

1. We had a big shock, and this calls for extreme measures.
2. The current inflation rate is too low.
3. Forecast inflation is too low:
Current monetary policy is typically thought to affect inflation with a one- to two-year lag. This means that we should always judge the appropriateness of current monetary policy using our best possible forecast of inflation, not current inflation. Along those lines, most FOMC participants expect that inflation will remain at or below 2 percent over the next one to two years.

Big shock: Of course we had the big shock - four years ago. Now, in 2012, what is it about that big shock that creates macroeconomic inefficiency that can be corrected by central bank action - and by central bank action that has not already been executed? Prices and wages are so sticky they have not adjusted? We are in the midst of some coordination failure that Ben Bernanke (or Narayana Kocherlakota for that matter) can fix? What? There has been massive intervention on an unprecedented scale, and the Fed was in the midst of a transformation of the maturity structure of its asset holdings when it embarked on its most recent asset purchase program. How much is enough? I'm sure you don't have any idea, and I'm afraid the Fed, in spite of its brave front, really has no idea either.

Is the inflation rate too low? It depends how you measure it. The Fed's inflation target is 2%, as measured by the pce deflator. The inflation rate, August 2011 to August 2012, is 1.5%, which is the number Kocherlakota uses. From September 2011 to September 2012, the number is 1.7%. If we look at a shorter horizon, the annualized percentage increase in August was 5.0%, and in September was 4.7%. If we look at a longer horizon, as I did here, from the beginning of 2007 or the beginning of 2009, you'll get something a little higher than 2.0% (I didn't have the most recent observation in the chart in the previous post). Too low? I don't think so.

Monetary policy should respond to forecast inflation, not actual inflation. You would think Kocheralakota would know better. Does a forecast give us any more information than what is in the currently available data? Of course not. It's the currently available data that we use to make the forecast. Further, the "best possible forecast of inflation" is not very good, in general, and the Fed's forecast of inflation is suspect. Any economic forecast is 90% judgement and 10% model, and the judgement of Fed forecasters is going to drive the forecast to something that justifies current policy actions. Arguing that the the Fed's forecasts could justify a more accommodative policy than what we already have is nonsense.

On April 8, 2010, Kocherlakota said this:
Deposit institutions are holding over a trillion dollars in excess reserves (that is, over 15 times what they are required to hold given their deposits). These excess reserves create the potential for high inflation. Suppose that households believe that prices will rise. They would then demand more deposits to use for transactions. Banks can readily accommodate this extra demand, because they are holding so many excess reserves. These extra deposits become extra money chasing the same amount of goods and so generate upward pressure on prices. The households’ inflationary expectations would, in fact, become self-fulfilling.
So the Kocherlakota of 2 1/2 years ago had some worries about the potential for inflation. Maybe he changed his mind for good reason? I don't think so. The new Kocheralakota seems to be a flimsy-excuse guy.

Addendum: Kocherlakota goes on in the same speech I quote from in the last paragraph to say this:
I hasten to say—and I want to stress—that I view this scenario as unlikely. For it to transpire, the country would need a combination of bad monetary policy and poor fiscal management. I do not foresee this combination as likely to occur.
That's important. We now have bad monetary policy and poor fiscal management (whoever is elected President next week - worse if it's Romney). The stunning thing here is that the old Kocherlakota didn't imagine that he would be the source of the bad monetary policy.

Thursday, October 25, 2012

FOMC Behavior and the Dual Mandate

In the most recent FOMC statement, we are told the following:
Consistent with its statutory mandate, the Committee seeks to foster maximum employment and price stability. The Committee remains concerned that, without sufficient policy accommodation, economic growth might not be strong enough to generate sustained improvement in labor market conditions. Furthermore, strains in global financial markets continue to pose significant downside risks to the economic outlook. The Committee also anticipates that inflation over the medium term likely would run at or below its 2 percent objective.
The argument is that we need more accommodative policy because: (i) the FOMC expects that, without such policy, there would be insufficient improvement in the state of the labor market and (ii) inflation is expected to continue below the FOMC's target of 2%. Thus, the FOMC anticipates that it will be missing on both sides of its dual mandate in the immediate future unless it does something about it.

Does the FOMC have it right? I have no idea whether the Fed is confident in its forecasts, whether we would think those forecasts are any good if we knew how they were done, etc., so all I can do is look at actual data. With respect to the "price stability" part of the dual mandate, the Fed has decided that the rate of increase in the PCE deflator is the appropriate measure of inflation. The first chart shows the twelve-month percentage increase in the PCE deflator.
As you can see, by this measure inflation is well below the 2% target - it's about 1.5% currently. But it was also well above the target for much of 2011 and early 2012. At that time, for example in the June 2011 policy statement, the FOMC was inclined to discount what it was seeing:
Inflation has moved up recently, but the Committee anticipates that inflation will subside to levels at or below those consistent with the Committee's dual mandate as the effects of past energy and other commodity price increases dissipate.
When the FOMC highlights the volatile nature of inflation on the upside and fails to mention it on the downside, one can't help but be suspicious. Further, suppose that we look at the price data in another way, as in the next chart.
In this chart, I show the pce deflator, and a 2% trend, beginning in January 2007. If we were judging the Fed's performance according to a price level target, as in the chart, we would think it was doing phenomenally well. Since early 2009, the actual PCE deflator has not strayed far from its target path, and is currently right on target, particularly with the large increase in the last month.

What about the other part of the Fed's mandate? We can argue about what an economically efficient level of aggregate economic activity is currently in the United States and what influence the Fed can have over it. But what if we take a standard central-banker-New-or-Old-Keynesian approach - something like what I describe in this post? The Congressional Budget Office (CBO) thinks that the current "short-term natural rate of unemployment" is 6%. I have no idea what they think that means, or if it makes any sense at all, but we'll use that number. The actual rate of unemployment is 7.8%, so by the CBO's criterion, the Fed is missing on this side of the mandate.

New Keynesians like to think about monetary policy in terms of a Taylor rule, which specifies a target for the federal funds rate as a function of the "output gap" and the deviation of the actual inflation rate from its target value. According to standard Taylor rules, the fed funds target should go down when the output gap rises and up when the inflation rate rises. You can even fit Taylor rules to the data. I fit one to quarterly data for 1987-2007, and obtained the following:

R = 2.02 - 1.48(U-U*) + 1.17P,

where R is the fed funds rate, U is the unemployment rate, U* is the CBO natural rate of unemployment (so U-U* is my measure of the output gap) and P is the year-over-year percentage increase in the PCE deflator. You can see how it fits the historical data in the next chart.
So that's how the Fed behaved in the past. If it were behaving in the same way today, what would it be doing? Given that U = 7.8, U* = 6.0, and P = 1.5, my Taylor rule predicts R = 1.1%.

Thus, the FOMC may see itself as missing on both sides of its mandate, but if it had been missing in the same way in 1997, the fed funds rate target would be at 1.1%, and the Fed would certainly not have been considering large-scale asset purchase programs, let alone making promises to keep the fed funds rate at 0-0.25% for almost three years into the future.

So, the Fed's behavior seems to have changed. Maybe change is good. Who is to say that behavior from 1987-2007 was optimal? One argument for an extended period of low nominal interest rates comes from Eggertsson and Woodford, and I raise some doubts about that here. Even if you buy Eggertsson Woodford, their extended-period arguments are a matter of fulfilling a commitment, not getting some extra accommodative mileage currently. People who use the Eggertsson-Woodford argument are imagining the extended period to be well off in the future, not where we are now.

But one view of the Fed's behavior could be that it has simply lost its inflation resolve. If that view catches on, look out.

More on Bubbles

This will perturb David Andolfatto, who has heard enough about bubbles. But Andolfatto perturbation is enjoyable, for some reason, so here goes. Ben Lester told me about this 1993 paper by Allen, Morris, and Postlewaite on bubbles. Here's the relevant quote:
I think that corresponds quite closely to what I had in mind here (and see this post as well). I'll leave you to judge whether Allen, Morris, and Postlewaite are better or worse economic theorists than Paul Krugman or Noah Smith.

Monday, October 22, 2012

Money and Bubbles

Money. Bubble. Liquidity. Fire sale. Those words are used a lot, particularly with reference to the recent financial crisis. Sometimes the words are used as if we all agree on what they mean, but if you engage anyone in a discussion about any of them, you'll find a distinct lack of agreement. I've seen several shouting matches in seminar rooms over what "bubble" means. Thus, it's not surprising that Noah Smith, Paul Krugman, and I don't think about bubbles in the same way.

From my previous post, here's my bubble definition, with examples:
What is a bubble? You certainly can't know it's a bubble by just looking at it. You need a model. (i) Write down a model that determines asset prices. (ii) Determine what the actual underlying payoffs are on each asset. (iii) Calculate each asset's "fundamental," which is the expected present value of these underlying payoffs, using the appropriate discount factors. (iv) The difference between the asset's actual price and the fundamental is the bubble. Money, for example, is a pure bubble, as its fundamental is zero. There is a bubble component to government debt, due to the fact that it is used in financial transactions (just as money is used in retail transactions) and as collateral. Thus bubbles can be a good thing. We would not compare an economy with money to one without money and argue that the people in the monetary economy are "spending too much," would we?

Noah Smith and I once had a conversation along these lines, and I thought we were making progress, but apparently not. Noah says the above paragraph is nonsense, since most payoffs on assets in monetary economies (like the one we live in) are denominated in terms of money. Thus, Noah reasons, if money is a bubble, then all assets are bubbles. How dumb could I be?

The payoffs on my stocks and bonds, and the sale of my house, may be denominated in dollars, but that does not mean that the value of those assets is somehow derived from the value of money. It's useful to ask what would happen if the monetary bubble "bursts." Think about an identical economy where money is not valued (that's always an equilibrium) and ask what happens. Everything changes of course, as now it's more difficult to carry out transactions - but not impossible. People will find other means to get the job done. Private financial intermediaries will issue substitutes for government money; people might engage in barter; people might use commodity monies. There is no reason why stocks and bonds and houses can't exist and be traded, with payoffs denominated in terms of something other than government-issued liabilities. Indeed, because private assets are substituting for government liabilities in exchange, some of those assets will have larger bubble components than in the monetary economy.

To give a practical example, think about monetary arrangements in the United States during the free banking era before the civil war. There was no fiat money or central bank. Transactions were executed primarily using the paper notes issued by private, state-chartered, banks, and using commodity money. Think of the role that gold played in that era. The price of gold had a bubble component as the stuff was used in exchange. It's not used in exchange today, so the bubble has gone away.

Here's Krugman's bubble definition:
I’d start by asking, what do we mean when we talk about bubbles? Basically, I’d argue, we mean that people are basing their decisions on beliefs about the future that are based on recent experience but can’t be fulfilled. E.g., people buy houses because they expect home prices to keep rising at a pace that would eventually leave nobody able to buy a first home...This sounds a lot like what happens in a Ponzi scheme...
It's different, right? My definition was based on rationality, and bubbles can be sustained forever. The crucial elements of a Krugman bubble are irrationality, and lack of sustainability. That's pretty much where the discussion ends. Krugman finds his notion of a bubble useful. I find mine useful. Krugman is in the Shiller bubble camp. I'm in the monetary theorist bubble camp.

Here's something interesting, though. Toward the end of his post, Krugman discusses fiat money, and Samuelson's overlapping generations (OG) model, which is one framework for thinking about money and what it does. No one took this model seriously as a model of money for a long time, perhaps because the tone of Samuelson's article is half-serious. However, Lucas used it in his 1972 paper, and this inspired Neil Wallace and his Minnesota students in the early 1980s to develop it further. The OG model captures Jevons's absence-of-double-coincidence problem in a nice way, it's easy to work with, and it admits complications like credit arrangements in a simple manner. Indeed, this book by Champ/Freeman/Haslag is essentially OG models for undergraduates.

One interesting feature of an equilibrium with valued money in the OG model, is that it looks like a Ponzi scheme - i.e. it has a feature Krugman associates with his bubble. And it's sustained forever. In each period, the young transfer goods to the old in the belief that they will receive goods when old from the next generation. Indeed, that arrangement looks just like social security, which is also a Ponzi scheme, though Krugman doesn't want to admit it. There's nothing wrong with it of course. Under the right conditions, social security can be an efficient and sustainable Ponzi scheme.

Saturday, October 20, 2012

The State of the World

There is some stuff in this Krugman blog post that is worth discussing. It's hard to get past his usual self-aggrandizement (I am a remarkably prescient forecaster; I am an island of clarity in an ocean of confusion; blah blah blah), but I'll try. We need to be tolerant, even when it's hard.

Here's what Krugman thinks is the key effect of the financial crisis:
Here’s how I interpret what we see in the historical data: financial crises leave an overhang of private-sector problems, principally excessive debt on the part of some subset of economic agents — households, in the case of the United States. Because these agents are either forced or strongly induced to slash spending, the “natural” rate of interest, the interest rate consistent with full employment, falls sharply — and in the case of a severe crisis, falls well below zero.
He's got the sign wrong. Suppose that for various reasons debt constraints bind more severely. That's in Eggertsson-Krugman for example. They just impose debt limits exogenously, but you could do something more sophisticated and tie the debt limits to the value of collateral, or what a would-be borrower stands to lose from default. In any case, what you get is more severe credit frictions, which make safe assets - government debt and safe private liabilities - more valuable. Why? These assets are now more useful at the margin in financial trade and as collateral. The safe market rate of interest is now too low, relative to where it should, or could, be. The "natural rate of interest" has not fallen. For more detail, see this post, point #1.

Krugman thinks economists and policymakers (past and present) are subject to "conceptual confusion." According to him, there are three facets to this:

1. Krugman is worried that people think he is being inconsistent (see my previous post for example). How can the Reinhart-Rogoff regularity (extended recovery after a financial crisis) be a regularity, and also represent an opportunity for Keynesian policy? Here's the heart of Krugman's argument:
...there are simple policy actions that could quickly end this depression now, there were simple policy actions that could have quickly ended depressions past. The problem is that now and then policy makers tend not to take these actions — which is why some of us write books.
A simple and quick solution is at hand. Easy. To buy this argument, you have to think that Krugman is really really smart, and the remainder of the human race is really really stupid. There are plenty of economists - with and without Nobel prizes - who don't think the solutions are easy.

2. Demand/supply and bubbles:
Over and over again one hears that we can’t expect to return to 2007 levels of employment, because there was a bubble back then. But what is a bubble? It’s a situation in which some people are spending too much — and we can’t expect those people to return to past spending habits.
What is a bubble? You certainly can't know it's a bubble by just looking at it. You need a model. (i) Write down a model that determines asset prices. (ii) Determine what the actual underlying payoffs are on each asset. (iii) Calculate each asset's "fundamental," which is the expected present value of these underlying payoffs, using the appropriate discount factors. (iv) The difference between the asset's actual price and the fundamental is the bubble. Money, for example, is a pure bubble, as its fundamental is zero. There is a bubble component to government debt, due to the fact that it is used in financial transactions (just as money is used in retail transactions) and as collateral. Thus bubbles can be a good thing. We would not compare an economy with money to one without money and argue that the people in the monetary economy are "spending too much," would we?

But the bubble component of housing prices after, say, 2000, does not appear to have been entirely a good thing, as it was built on false pretenses. Various kinds of deception resulted in housing prices - and prices of mortgage-related assets - that, by anyone's measure, exceeded what was socially optimal. As a result, I think we can make the case that pre-2008 real GDP in the US was higher than it would have been otherwise. Further, the housing-market and mortgage-market boom could have masked underlying changes taking place in US labor markets - for example David Autor's "hollowing out" phenomenon. One could argue that there was a cumulative effect in terms of the labor market adjustments needed, and that these adjustments took place during the recent recession, and are still taking place. See for example this paper by Jaimovich and Siu. That's why all the long-term unemployed. So that's not some confusion. People are talking about alternative ideas that have some legs, and may have quantitative significance. Why dismiss them?

3. Sectoral shifts:
One last point: we still keep hearing the “structural” argument, that we have to expect prolonged high unemployment because it takes time to turn construction workers into manufacturing workers or whatever. One answer is that this portrait of the economy is factually wrong: job losses have not been concentrated in a few sectors or professions, they have been broadly spread across the economy. But there’s also a conceptual answer: if shifting workers across sectors requires mass unemployment, how come the bubble years — when we were moving out of manufacturing into housing — weren’t high-unemployment years? Why does moving into the bubble sectors mean more jobs, but moving out into other sectors mean fewer jobs? I’ve never heard a coherent answer.
Answer: If you're not listening, you can't hear. There is plenty of unusual behavior in the recent labor market data: (i) the jobless recoveries that Krugman highlights here; (ii) the large drop in employment relative to output in the recent recession; (iii) the abnormally large fraction of long-term unemployed. One element of unusual behavior is the failure of residential construction to lead the recovery. David Autor and others (as mentioned above) have highlighted the recent shift out of middle-skill occupations. There is plenty here for any labor economist or macroeconomist to sink their teeth into. How do you tie together the financial crisis, the shifts in employment across sectors, and the changes in the skill mix? It's well-known that sectoral changes have macroeconomic consequences - economists have been discussing this at least since the early 1980s.

So, there is a lot going on. If we want to think of our current predicament as an aggregate demand management problem, we're missing all or most of what is important. It's certainly not simple, but it's a lot more interesting than an IS-LM model.

Friday, October 19, 2012

Financial Crises

I just received Gary Gorton's new book, Misunderstanding Financial Crises in the mail. This is as good an account of the financial crisis as any I have seen, and adds to Gary's previous book, Slapped by the Invisible Hand. Gary has an unusually broad grasp of banking history, modern banking theory, financial theory, and the practical aspects of modern finance and institutions. Indeed, some of his consulting work placed him at the center of the financial crisis. In the late 1980s, Gary taught me that securitization was important, long before most economists had any idea what that was about. I don't agree with everything he writes, but you can learn a lot from his new book.

Lucas once said that business cycles are all alike. Gorton wants to focus on what makes financial crises all alike. His key point seems to be that, in any financial crisis, we can find a run. In the United States, bank runs were a key feature of panic episodes during the National Banking era (1863-1913) and the Great Depression. Bank runs were certainly not a feature of the recent financial crisis, but Gorton thinks that "repo runs" were essentially the same phenomenon.

Gorton's idea is that any financial entity that intermediates across maturities can be subject to a run. The Diamond-Dybvig view is that bank runs are inherent to the liquidity transformation carried out by banks. A well-diversified bank transforms illiquid assets into liquid liabilities, subject to withdrawal. Everything is fine unless depositors anticipate that others will run on the bank, in which case we find ourselves in a bad equilibrium - a bank run. In the run equilibrium, it is optimal for each depositor to run to the bank to withdraw his or her deposit, since their best hope in this equilibrium is to get to the bank before the assets are exhausted.

A shadow bank is not quite like a Diamond-Dybvig bank. A typical shadow bank holds long-maturity liquid assets and finances its portfolio by rolling over short-term repos (repurchase agreements), using the underlying assets as collateral. One might think that, because the shadow bank's assets are liquid, a run could never occur. If financial market participants are reluctant to roll over the shadow bank's repos, it can sell assets to pay off its debts. The problem arises if there is a systemic revaluation of shadow bank assets. Then, an individual shadow bank could default because new repo holders are demanding large haircuts in their repo contracts. Worse, since all shadow banks are selling assets simultaneously, the prices of assets are further depressed (a fire sale), which amplifies the repo run.

A debt contract is an efficient arrangement that works extremely well in good times. The payments required under a debt contract are non-contingent, and there is no fuss about what it means to fulfill the terms of the contract. Problems occur in default states, however, particularly when there are multiple creditors. For a bank, the coordination problem that arises in the event of default is particularly severe, given the large number of small depositors. However, coordination can be very costly even if a financial institution's creditors consist of a few other financial institutions. In a Diamond-Dybvig model, coordination is formalized as "sequential service," which inhibits communication among the bank's depositors in a rather brutal fashion. Of course, a shadow bank run really has nothing to do with creditors "lining up" at the shadow bank, so we can't take sequential service literally if we want to think of repo runs as akin to Diamond-Dybvig runs.

Coordination costs that arise in a default involving multiple creditors are reflected in legal costs, and the time that assets are tied up in litigation. In the case of banking, deposit insurance minimizes those costs in a nice way. The FDIC stands in for all creditors, thus eliminating the replication of default costs among creditors and doing away with disputes among creditors. Further, resolution occurs quickly. Of course, we all know about the fallout from insuring the liabilities of financial intermediaries. Absent constraints on risk-taking, insurance creates a moral hazard problem, whereby intermediaries take on more risk than is socially optimal.

So if, as Gorton suggests, a financial crisis is defined by widespread runs on financial intermediaries, how is that helpful?

1. Does this mean that central banks should respond to every financial crisis in the same way? Probably not. Banking panics in the National Banking era and the Great Depression were essentially currency shortages. The recent financial crisis involved a shortage of safe assets, more broadly. A currency shortage can be solved with a central bank open market purchase of government debt. A shortage of safe assets may be a problem for fiscal policy - as asset swaps by the central bank will not change the net supply of safe assets. Further, central bank lending policies may depend on the particulars of the crisis.

2. If we think of a financial crisis as a run problem, and draw an analogy to banking and deposit insurance, this must mean we should insure everything that looks vaguely like banking. Moral hazard everywhere. Great.

3. One of the lessons of the financial crisis is that financial factors are important. Surprisingly, many people once thought otherwise, and some continue to think so. But the importance of financial factors is not confined to the events we want to call "financial crises." It seems wrongheaded to take episodes in history and put them in "crisis" and "non-crisis" bins.

You can see how fussing over what is a financial crisis and what is not can be unproductive. Case in point:

1. Reinhart and Rogoff define a financial crisis in a particular way, and argue that there is a regularity in the data. Recoveries after financial crises are protracted. People use that "fact" in different ways. Jim Bullard wants to argue that the Reinhart-Rogoff regularity tells us that the Fed should not held responsible for the slow recovery. Paul Krugman wants to use the Reinhart-Rogoff regularity to absolve the Obama administration. Of course, he is walking a fine line here as, in contrast to Bullard (apparently) he seems to think that appropriate monetary and fiscal policy would have left Reinhart and Rogoff with no regularity to talk about.

2. Mike Bordo and Joe Haubrich define a financial crisis differently (from Reinhart and Rogoff) and argue that, in the United States, it's hard to argue that the Reinhart/Rogoff regularity is in the data. Sometimes we see it. Sometimes we don't. John Taylor picks up on this. Like Krugman, he has an ax to grind - different ax though. According to Taylor, things are worse than they should be because of you-know-who.

Taylor does point out something useful, though, and quotes Bordo:
The mistaken view comes largely from the 2009 book “This Time Is Different,” by economists Carmen Reinhart and Kenneth Rogoff, and other studies based on the experience of several countries in recent decades. The problem with these studies is that they lump together countries with diverse institutions, financial structures and economic policies.
That's important. The U.S. financial system is unique in many ways. It still has many small banks; U.S. financial regulation is unusually complicated, with a confusing patchwork of overlapping regulatory authority; the U.S. supplies the world's reserve currency; the Fed intervenes in different ways because of peculiarities in our financial markets. The comparison with Canada is useful. Canada and the U.S. are similar in many ways, but it is difficult or impossible to find anything that Reinhart-Rogoff or Bordo-Haubrich would call a financial crisis, in all of Canadian history. How come? They have debt contracts, banking, and financial intermediation across maturities in Canada. Why no panics?

Why indeed. Definitions and data give us something, but they can't substitute for theories that can help us organize our thinking about the data. The immediate question is whether or not the monetary and fiscal authorities in the United States are doing the appropriate things. There are good reasons to think that recessions are not alike, and that the most recent recession has features that are different from previous ones in the United States - and different in important ways from episodes where we think that there was some element of "financial crisis." Even if we could figure out the Great Depression, and understood completely the policies that would have been appropriate at the time, that would be no guarantee of success under current conditions.

Sunday, October 14, 2012

Government Debt and Intergenerational Distribution

After reading Nick Rowe, Brad DeLong, and Paul Krugman, I now understand who bears the burden of the government debt. Any member of the younger generation who reads the stuff that these old guys have written on government debt will be hopelessly confused. And that will be a burden on us all.

The idea that a larger government debt is a burden for future generations is so strongly intuitive as to be part of the standard narrative for anyone who wants to tell you that more government debt is a bad idea. But is that correct? When I teach undergraduates about government finance, I find it instructive to start with the Ricardian equivalence theorem. In a frictionless world, government debt is irrelevant. A tax cut that increases the government debt today has no effect because everyone understands that government debt is just deferred taxation. Lifetime wealth does not change, and everyone saves their tax cut today so as to pay the higher future taxes that are required to pay off the government debt in the future. Ricardian equivalence is a useful starting point, as it makes clear what frictions might cause Ricardian equivalence to break down - and that's a route for thinking about how policy might work to improve matters. Distorting taxes, intragenerational distribution effects from tax policy, and credit market frictions all potentially make a difference. But that doesn't make Ricardian equivalence "wrong" or useless. Indeed, it is an important organizing principle, and needs to be taken seriously.

One key departure from Ricardian equivalence arises because of intergenerational redistribution effects from government tax policy. Ricardian equivalence works because changes in the timing of taxes do not matter for anyone's wealth - the present value of tax liabilities for each individual is unchanged. But what if the government is cutting one person's taxes today, and paying off the government debt in the future by taxing someone else? Surely that comes into play in reality, as the government can cut taxes today, increase the government debt, and potentially not pay off the debt for 100 years, at which time the people who were on the receiving end of the current tax cut are long dead?

Robert Barro had an answer for this. If generations are tied together through altruism (we care about our children) and bequests, then we behave as if we are chained together with our descendants, and might as well be infinite-lived households. If I receive a tax cut today, then I save more, and give my descendants a larger bequest that will allow them to pay their higher future taxes. But surely this is going to far. Following the logic of Barro's argument, chains of altruism tie everyone together in ways that make everything neutral. There is certainly altruism in the world, but I think we all recognize that some collective action is required to make outcomes more efficient.

A standard vehicle for thinking about intergenerational distribution is the overlapping generations (OG) model (DeLong knows that the OG model exists, which is a start, at least). Peter Diamond's version of the OG model, with capital accumulation and production, is useful, but I'll simplify here to get the ideas across. Suppose two-period-lived people, with two generations alive at each date - young and old. The population grows at rate n, and each person is endowed with y units of consumption good when young, and zero when old. Storage is possible, with r the rate of return on storage from one period to the next. This is useful, as it ties down the real interest rate at r (so long as there is some storage in equilibrium). Suppose also that the government has access to lump sum taxes on the young and the old. The government can also issue one-period real bonds, with the real interest rate on government bonds = r in equilibrium.

Suppose that the government increases the quantity of government debt in the current period, T, by an amount b per young person currently alive. Since the real interest rate is constant at r forever, the effect of this change in government debt on economic welfare for each generation depends only on how taxes change for the rest of time. Suppose that the increase in debt is reflected in a lump sum transfer of x(T) to each person currently alive, so

x(T) = [b(1+n)]/(2+n)

Now, clearly the current old are better off as a result. They receive a transfer and are better off. What about everyone else - the current young, and future generations? For those people, it depends.

Scenario 1: Pay off the debt at T+1: In this scenario, to pay off the principal and interest on the government debt in period T+1 requires total taxes per young person alive equal to [b(1+r)]/(1+n). Supposing the government spreads the tax burden equally across people alive in period T+1, the tax per person at T+1 is

-x(T+1) = [b(1+r)]/(2+n)

Therefore, as long as n > 0, so the population is growing, the young people in period T (who are old at T+1) are better off, as their lifetime wealth increases. The old in period T+1 are worse off, as they pay a higher tax.

In this case, it is clear that the government debt is a burden on future generations - specifically the next one, which pays the taxes to retire the government debt.

Scenario 2: Hold the debt constant forever at its higher level: Under this scenario, the government debt per young person in period T+i is


(where x^i is x to the power i). If taxes are the same for young and old during any period, this implies that the tax per person in period T+i is

-x(T+i) = rb/[(2+n)(1+n)^(i-2)]

In this case, the young in period T are better off in present value terms, but anyone born in periods T+1 and beyond is worse off.

Compared to scenario 1, we are now spreading the burden of the government debt across all future generations. Note however, that the government debt per capita vanishes in the limit. The experiment increases the government debt by a fixed real amount and, as the population grows, the burden of the debt in per capita terms falls.

Scenario 3: Hold debt per capita constant at the higher level forever: In this case, if everyone at a point in time bears the same tax, the tax in period T+i is

-x(T+i) = [b(r-n)]/(2+n)

As in scenarios 1 and 2, the young in period T are better off, but now things are more interesting. If r > n, so that the real interest rate is higher than the population growth rate, we get a similar scenario to what we had previously, except we are now spreading the burden of the debt equally across future generations. However, if r < n, we actually make everyone better off. When r < n, it is socially inefficient to use the storage technology, and an optimal arrangement would have government debt driving out use of the storage technology. Diamond's classic paper shows how this works in a more standard neoclassical growth framework. In general, an economy with r < n is dynamically inefficient, and government debt is one means (as is social security) for effecting the appropriate intergenerational transfers. The interesting thing about the scheme in scenario 3 when r < n is that it works as a Ponzi scheme - effectively, each generation borrows from the next, and everyone is better off. Magic!

What's the bottom line? Increasing the quantity of government debt does indeed represent a transfer in wealth from future generations to those currently alive (or, in reality, those currently working). The only important qualification arises if society is not efficiently distributing wealth across generations. In that case, the inefficiency can be corrected with an increase in government debt. But unless we want to argue that the intergenerational redistribution being done by the U.S. social security system is insufficient, that seems a difficult argument to make.

Friday, October 5, 2012

How Central Bankers Think

When I read Charles Evans's most recent speech, I was struck by these two sentences:
Whenever the economy operates below its potential, the key mechanism that returns the economy back to potential is a fall in real interest rates. This decline reduces the supply of saving and boosts the demand for investment, resulting in increased spending.
Ben Bernanke tells us about how the Fed can speed the adjustment to "potential:"
Generally, if economic weakness is the primary concern, the Fed acts to reduce interest rates, which supports the economy by inducing businesses to invest more in new capital goods and by leading households to spend more on houses, autos, and other goods and services. Likewise, if the economy is overheating, the Fed can raise interest rates to help cool total demand and constrain inflationary pressures.
Those two quotes encapsulate the dominant school of central banking thought on the FOMC. We could even write this down formally - it's simple. Here are three equations that describe the economy and what monetary policy does, according to Evans, Bernanke, Kocherlakota, and I think Rosengren and Yellen, at least:

(1) r = R - i
(2) u = f(r-r*)
(3) I = i + g(r-r*)

In equations (1)-(3),

r = real interest rate
R = nominal interest rate
i = anticipated inflation rate
u = unemployment rate
r* = efficient real interest rate
I = actual inflation rate
f(.) is an increasing function
g(.) is a decreasing function

In this model, R is set by the Fed, i is exogenous, and r* is exogenous. Equation (1) is a Fisher relation. The real rate is the nominal rate minus the anticipated inflation rate. When the Fed says that "inflation expectations are well-anchored," what they mean is that i is independent of how the Fed sets R (as the Fed can always say reassuring things about how it is committed to price stability, apparently). Thus, when the Fed moves the nominal rate R, the Fed thinks it moves the real rate in lockstep.

The economically efficient real interest rate r* is what Woodford would call the "Wicksellian natural rate." In Woodford's world, this would be the equilibrium real interest rate if all wages and prices were flexible. This may or not be what Bernanke has in mind - he tends to leave unspecified the sources of the inefficiencies he is trying to correct. Equation (2) states that the unemployment rate rises if there is an increase in the difference between the actual real interest rate and the efficient real interest rate. But how do we know what r* is? From Charles Evans's point of view, that's easy. If Fed policy stayed unchanged (so that r is unchanged), and the unemployment rate went up, r* must have gone down.

Equation (3) states that the the current inflation rate is the anticipated rate of inflation, plus a term that depends negatively on the difference between the actual real rate and the efficient real rate. Therefore, if the real rate is above the efficient rate, the inflation rate will be low. Implicit in equations (2) and (3) is the Phillips curve - a central part of FOMC religion. FOMC statements and public discussion by Fed officials are replete with Phillips curve language. Note that changes in R which translate into changes in the same direction in r will move u and I in opposite directions - a movement along the Phillips curve.

Some readers will recognize the similarity between equations (1)-(3) and the three equations that some New Keynesian researchers work with. Typically the three New Keynesian equations are: (i) IS curve; (ii) Taylor rule; and (iii) Phillips curve. What I wrote down above is roughly the same idea, and that's no accident. What Woodford and his disciples did was to essentially reverse-engineer actual Fed policymaking. That's why central bankers are infatuated with New Keynesian economics - it rationalizes what they do.

Here is how the FOMC sees its current predicament. r* is really low, making u really high. The optimal thing to do would be to lower R to reduce r and reduce u. But R = 0 and can't go lower. Solution: There are long-term Rs that we can go after, which are currently above zero. Let's buy long-maturity assets and lower long-term Rs instead. You can read about that in Bernanke's speech.

But what's wrong with this approach? These ideas may be easy for Fed officials to explain to the Rotary club, but if you recite the ideas enough you start to believe them. How can anyone think that our current problem is that efficient real interest rates fall far below actual real interest rates? A short rate of -2% is too high? Why? TIPS yields of -1.65% (5 year) or 0.37% (30 year) are too high? Again, why? No one thinks that monetary policy has any long run consequences. Over what horizon do FOMC members think that they can essentially peg a real interest rate? Didn't exploitation of a perceived Phillips curve tradeoff get us in trouble in the past? Aren't we worried about that? Why did Phillips-curve "theory" become the dominant theory of inflation among policymakers? Why did monetary theory disappear? We know there are problems with the targeting of monetary aggregates, but surely most of us think that exchange in some class of assets is what drives long-run inflation. Where's the money?

Bullard and Price Level Targeting

Jim Bullard (St. Louis Fed President) gave a talk yesterday which has some interesting ideas in it. In the second chart in this post, you can see that the pce deflator, which is the Fed's favorite measure of the price level, is currently very close to a 2% growth path starting in 2007. Bullard's slides show that this is the case even if our base year is as far in the past as 1995. Thus, while the Fed speaks a language that might make you think it believes past inflation is irrelevant, in practice the Fed appears to be targeting a 2% price level path.

But why do that? Bullard makes the case that this is the prescription that comes out of a Woodford sticky-price model. While I think price and wage stickiness is at best of second-order importance, I think there are other good reasons to think that price level targeting is a good idea. For example, such a policy rule appears to have good properties in terms of minimizing the inflation uncertainty associated with long-term nominal debt contracts.

But there is more on Bullard's mind. He seems to want to make the case that, in spite of the weak recovery we have been experiencing in the U.S., monetary policy has been close to optimal. In support of that argument, he enlists Reinhart and Rogoff, who document a slow recovery as a regularity in the aftermath of previous financial crises. One could say that Reinhart and Rogoff don't have a theory to explain the regularity, or that there may be something policy could do in response to the slow recovery, in spite of the regularity. For example, some people think that high inflation, which would redistribute wealth from creditors to debtors, would be a good idea in current circumstances. In response to that, Bullard argues that inflation redistribution would do more harm to the Fed's credibility than it's worth. One could argue as well that, if redistribution is the answer to our problems, then fiscal policy is a more efficient vehicle for doing it than is monetary policy.

A point that Bullard emphasizes is that nominal GDP targeting in the current circumstances would be the wrong policy. Basically, there is nothing much that monetary policy can do about the slow recovery so, for example, targeting 5% nominal GDP growth would give us too much inflation. By implication, it seems Woodford is contradicting himself. Bullard argues that the policy we have (optimal, he says) is the one Woodford would recommend. But Woodford's Jackson Hole paper seems to view nominal GDP targeting favorably. Woodford-optimal policy and NGDP targeting appear to be two different things.