Category Archives: Scott Sumner is Coherent

Ideology and Macroeconomics

Posted on by

Scott Sumner writes,

I am amazed by how many proponents of fiscal policy don’t understand that it’s symmetrical. Fiscal policy doesn’t mean more government; it means more government during recessions and less government during booms, with no overall change in the average level of government. Anyone who doesn’t even get to that level of understanding, who doesn’t think in terms of policy regimes, is simply not part of the serious conversation.

I agree with the first two sentences, but not with the last.

Yes, in theory, there should be economists who, as they argued for more stimulus in 2009, should at the same time have been arguing for entitlement reform or other reductions in future spending. Other things equal, the bigger debt that we have accumulated over the past five years would make a non-ideological macroeconomist want to propose tighter fiscal policy somewhere down the road.

But “nonideological” and macroeconomics are nearly oxymorons. Name a prominent economist who believes that fiscal expansion is important during recessions and who also is to the right of the median economist on issues like school choice or taxing the rich or the usefulness of regulation. Or try to name a prominent economist who is to the left of the median economist on those issues and who does not believe that fiscal expansion is important.

I know that I was more to the left generally 30 years ago, and I was a confirmed Keynesian. I am more to the right today, and I am a skeptic of Keynesianism.

I do not think that being on the left (right) on other issues necessarily causes you to be a supporter (skeptic) of Keynesianism. However, I do think that people try to avoid affiliative dissonance and cognitive dissonance.

Cognitive dissonance is an issue because if your general view is that market failures are small and difficult for government to correct, then it is hard to fit Keynesianism in with that belief. If your general view is that market failures are significant and require government intervention, then it is hard to fit the skepticism toward Keynesianism in with that belief.

Affiliative dissonance is my own expression. It just means that if the people with whom you feel an affinity on issues W, X, and Y take a position on issue Z with which you disagree, that makes you uncomfortable. Other things equal, this will make it easier to get you to change your mind on issue Z.

We know from Daniel Kahneman (and others) that we are good at rationalizing opinions that may be arrived at on the basis of intuition. I am not saying that therefore we will never find truth in macroeconomics. What I am saying is that if you close your ears every time you detect someone’s ideology embedded in what they say about macroeconomics, then you will not hear anything.

What do Low Unemployment Insurance Claims Tell Us?

Posted on by

Scott Sumner raises the issue.

with today’s numbers (308,000 on the 4 week average) we have fallen below the 0.1% level, meaning that fewer than 1/1000ths of Americans now file for unemployment comp each week. That’s not just boom conditions, it’s peak of boom conditions. The only other times this occurred since 1969 (when unemployment was 3.5%) were just a few weeks at the peak of the 2000 tech boom and a few weeks at the peak of the 2006 housing boom. In other words, the puzzle is now even greater than 6 weeks ago, as the unemployment rate is still at recession levels (7.3%.) Something very weird is going on in the labor markets.

Some possibilities:

1. We are in a “slow-churn” economy, with fewer jobs being created or destroyed.

2. The financial crisis time-shifted layoffs from 2010-2013 back to 2008-2009. Without the financial crisis, firms would have gradually noticed that they were carrying unproductive workers. The financial crisis made them realize this suddenly. By now, the companies are lean and do not need to shed anyone. As a result, the unemployment rate, instead of climbing steadily from 5 percent to 7 percent, first overshot and is in the process of dropping back.

3. Michael Mandel is right, and the unemployment rate is poised to drop swiftly.

1930s Bank Failures and Output: DYTVSC?

Posted on by

Jeffrey Miron and Natalia Rigol write,

assume it takes at least a month for bank failures to disrupt credit intermediation and thereby lower output. Under this assumption, any contemporaneous relation between output and failures is assumed to represent the impact of output on failures. We can then determine the effects of failures on output by excluding contemporaneous failures from the regressions. At a minimum, it seems reasonable to consider this specification.

Table 3 presents the results. In these regressions, bank failures have no predictive power for
output; indeed, the coefficients on bank failures imply that failures predict increases in output. Thus, if the
identifying assumption implicit in Table 3 is correct, these data and this specification provide no evidence
(indeed, contradict) the view that bank failures cause output declines.

DYTVSC means “Did you two visit the same country?” The person with the opposite point of view is Ben Bernanke.

Scott Sumner has more.

Monetary Offset

Posted on by

See Scott Sumner’s short paper.

estimates of fiscal multipliers become little more than forecasts of central bank incompetence. If the Fed is doing its job, then it will offset fiscal policy shocks and keep nominal spending growing at the desired level. Ben Bernanke would deny engaging in explicit monetary offset, as the term seems to imply something close to sabotage. But what if he were asked, “Mr. Bernanke, will the Fed do what it can to prevent fiscal austerity from leading to mass unemployment?” Would he answer “no”?

The Keynesian point of view is that the Fed “ran out of ammunition” when the Fed Funds rate went to zero. At this point, I do not know what to say to people who take that view. If it were true, then the fiscal multiplier should be bigger than one would otherwise expect. Yet it seems to have turned out smaller–the stimulus did less than predicted, and the austerity did less damage than predicted. And to me, it seems obvious that as long as there is stuff that the Fed can buy, including long-term bonds and foreign currency, it has “ammunition.” But the Keynesians routinely dismiss as incompetent anyone who who claims that the Fed cannot run out of ammunition. Perhaps Scott’s paper will force them to actually defend their position, although chances are that they will just continue to ignore or insult those with whom they disagree.

My own views do not align with either Sumner or the Keynesians. PSST is an alternative to the AS-AD story.

Booms, Busts, and Money

Posted on by

George Selgin writes,

it seems to me that there is a good reasons for not buying into Friedman’s view that there is no such thing as a business cycle, or Sumner’s equivalent claim that there is no such thing as a monetary-policy-induced boom. The reason is that there is too much anecdotal evidence suggesting that doing so would be imprudent. The terms “business cycle” and “boom,” together with “bubble” and “mania,” came into widespread use because they were, and still are, convenient if inaccurate names for actual economic phenomena. The expression “business cycle,” in particular, owes its popularity to the impression many persons have formed that booms and busts are frequently connected to one another, with the former proceeding the latter; and it was that impression that inspired Mises and Hayek do develop their “cycle” or boom-bust theory rather than a mere theory of busts, and that has inspired Minsky, Kindleberger, and many others to describe and to theorize about recurring episodes of “Mania, Panic, and Crash.” Nor is the connection intuitively hard to grasp: the most severe downturns do indeed, as monetarists rightly emphasis, involve severe monetary shortages. But such severe shortages are themselves connected to financial crashes, which connect, or at least appear to connect, to prior booms, if not to “manias.” That the nature of the connections in question, and the role monetary policy plays in them, remains poorly understood is undoubtedly true. But our ignorance of these details hardly justifies proceeding as if booms never happened, or as if monetary policymakers should never take steps to avoid fueling them.

Read the whole thing. The conventional wisdom, as of 2007, was that no matter what happens in financial markets, the Fed can keep employment high if it avoids disappointing people’s inflation expectations. That conventional wisdom disappeared during the crisis of 2008. At that time, Chairman Bernanke decided that ordinary monetary policy was not going to work. Instead, bailouts were needed in order to prevent a catastrophic recession. In the event, we had a bad recession. Now, the conventional wisdom is that he was right and that he made the recession less catastrophic. My alternative hypothesis is that we got more or less the same recession we would have had without bailouts (and without the stimulus, for that matter). It is impossible to go back and run the relevant experiment to determine who is right. I am willing to admit I may be wrong, but I think that those who espouse the conventional wisdom ought to be equally modest.

Scott Sumner became a notorious radical by sticking with the pre-crisis conventional wisdom rather than adopting the post-crisis conventional wisdom. Meanwhile, as Selgin points out, the Austrian alternative that there is such a thing as an unsustainable boom has been picked up by everyone from erstwhile descendants of Milton Friedman to President Obama (in the latter case, it fits in with the narrative that everything bad that takes place during his Administration is the fault of George Bush and/or Congressional Republicans).

My instincts are:

1. Downplay the role of the financial crisis, as opposed to ongoing structural adjustment.

2. Having said that, rapid expansion and contraction of the banking sector is bound to require a lot of short-term structural adjustment elsewhere, particularly in the contraction phase.

3. I am finding myself more and more reverting to what I call the MIT view (before Dornbusch and Fischer) that asset shuffling by the Fed (including all the conventional tools of monetary policy as well as the unconventional ones) does not have much impact.

David Andolfatto’s Challenge

Posted on by

He writes,

The PCE inflation rate since 1990 averaged 2.09% per annum.

What’s interesting about this diagram is that even though the Fed does not officially target the PCE price level, the data above suggests that the Fed is behaving as if it does.

As a price-level (PL) target is equivalent to a nominal GDP (NGDP) target in a wide class of macroeconomic models (especially under the assumption of constant productivity growth), then what more does the NGDP crowd expect from an official NGDP target? Seems to me that they are just asking for more price inflation and wishfully hoping that some of the subsequent rise in NGDP will take the form of real income.

Tell me I’m wrong (and why).

Pointer from Mark Thoma.

1. I made a similar point when I wrote,

From April of 2003 through April of 2008, the rate of growth of the CPI averaged 3.2 percent. From April of 2008 through April of 2013, it averaged 1.6 percent. If in 2007 you had asked macroeconomists to predict the consequences of a decline in the inflation rate of that magnitude, how many would have told you to expect unemployment to rise above 7 percent? None of them would have foreseen it. My guess is that many of the macroeconomists would have regarded a drop in inflation of 1.6 percentage points as close to a non-event for unemployment.

2. I downloaded the quarterly PCE data from 1990 Q1 through 2013 Q3. The average inflation rate (simple average, not compounded) was 2.16 percent for the whole period. For the sub-period prior to 2008 Q3 it was 2.34 percent. For the sub-period since it has been 1.43 percent. So, suppose the Fed’s inflation target was actually 2.34 percent. In that case, it recently undershot its target by 0.91 percentage points. As I pointed out above, that hardly seems like enough to cause Armageddon.

In fact, it was not even the worst miss. From 1997 Q3 through 1999 Q3, the PCE inflation rate averaged 1.10 percent, which would be a miss of 1.20 percentage points. If 0.91 caused the Great Recession, why was 1.20 consistent with strong growth?

3. One of Scott Sumner’s arguments against targeting prices is that prices are mis-measured. I think if you go with the mis-measurement argument, you have to explain away the 1997-1999 anomaly by saying that there was more inflation and less real growth than what the statisticians reported; turning to 2008-Q3 to present, you explain the anomaly by saying that there was less inflation and more real growth than what the statisticians reported. I think those are pretty difficult cases to make. The sorry-looking employment figures for the last five years are consistent with weak real growth. (NOTE: I wrote this before Scott Sumner replied, but I think I anticipated most of Scott’s points. These days, I almost never publish a post immediately. Scheduling them in advance instead makes me more careful.)

4. I hope that Andolfatto sees that there is a larger point to be made here than to pick on Scott Sumner. If a “wide class” of models suggests that you would not see a Great Recession arising from a small miss to an inflation or price-level target, then I think it is time to open one’s mind to other ideas.

(The term “wide class” sticks in my craw. In my table of contents, you may recall that I said “the macroeconomics profession became narrow, inbred, and retarded.” The book is progressing well, but I have left that phrase out for now.)

Rules vs. Discretion

Posted on by

Scott Sumner writes,

I’m all for a rules-based approach to policy. But unfortunately Taylor fails to make his case. You’d think a fan of rules-based policy would provide a razor sharp critique of Fed policy, but Taylor’s critique is anything but clear

Scott Sumner’s rallying cry is “Target the Forecast!” (for nominal GDP) and John Taylor’s rallying cry is “Follow the Taylor Rule!”

Some remarks:

1. I think I saw the clash between Sumner and Taylor coming even before Sumner did.

2. “Target the Forecast!” is, in a generic sense, what the Fed has been doing since the 1960s. The FOMC discussions revolve around forecasts. Fed staff scrutinize data closely in order to divine what it means for the forecast. Alan Greenspan was an intense and promiscuous data-scrutinizer. Taylor would argue that, until late in his term, Greenspan’s target-the-forecast approach happened to line up with the Taylor rule.

3. What is the result? As Ed Leamer puts it in chapter 15 of Macroeconomic Patterns and Stories,

On the basis of circumstantial evidence, the Federal Reserve Board, by raising rates late in expansions, can take some blame for almost all our recessions.

Below is circumstantial evidence of the sort he describes. See how large moves in the Fed Funds rate tend to lead large moves in the unemployment rate.

4. Leamer also says,

With all these patterns, it is a mystery* whether monetary policy can be said to cause anything or merely reacts to things that would have occurred anyway. But if I felt the need, I could suppress the doubt and tell with confidence the following story.

[Expansions start out with mild inflation. Then price pressures build as the expansion matures] But the Fed fiddles as inflation smolders. The ever-so-gradual increase in inflation is not enough to get the Fed to respond, but like a small brush fire, inflation soon enough gets out of control…By the time the data are in, and the Fed rate-setting committee has deliberated enough to make absolutely sure that it is time to make a change in monetary policy, inflation is burning fiercely and it takes a heavy spray of higher interest rates to put the fire out. That creates an inverted yield curve, a credit crunch** for housing, and an unpleasant recession. Oh, Oh, we’re sorry, say the Fed Governors, who knock down interest rates to try to get housing and the rest of the economy back on their feet.

5. From this historical perspective “target the forecast” is what got us where we are today. Sure, it looks like a great idea now, when we think that the expansion is still not mature and price pressures are still nowhere to be seen. But eventually “target the forecast” will once again result in a failure to change policy in time. We will continue to experience needless, Fed-induced cyclical behavior unless the Fed is lashed to a rule.

6. You might characterize my beliefs as:

Probability that “target the forecast” (using some market prediction for nominal GDP) would stabilize the economy = .15

Probability that “stick to a rule” would stabilize the economy = .10

Probability that monetary policy “merely reacts to what would have occurred anyway” = .75

*The “mystery” arises in part because the Fed only controls the short-term nominal interest rate, and it is the long-term real interest rate that most plausibly drives spending. In chapter 5, Leamer writes,

the interest-rate on the 10-year has a life of its own, sometimes moving with the 3-month rate, but not always. That should make one wonder how much impact the Fed has on the longer-term rates and also wonder how much it matters.

In chapter 15 he says that

long-term interest rates were generally elevating at the ends [of economic expansions since 1960]. Maybe that is what killed off housing. Maybe that would have occurred even if the Fed had not taken action.

**Leamer was writing prior to 2008, when a different sort of credit crunch arose. I will discuss the Leamer credit-crunch model in a subsequent post.

What Does Japan Exemplify?

Posted on by

Noah Smith writes,

It seems to me that the standard New Keynesian sticky-price story just cannot explain Japan. The “short run” for Japan is over and done. We are not looking at a “short-run” fluctuation caused by sticky prices.

Pointer from Tyler Cowen.

In textbook terms, Smith is saying that the long-run aggregate supply curve is vertical, so aggregate demand does not matter.

Paul Krugman objects. In textbook terms, Krugman is saying (I am pretty sure) that in a liquidity trap the aggregate demand curve is vertical, so that the long-run aggregate supply curve does not matter.

So it gets back to whether one believes in a liquidity trap. Krugman writes,

the only reason deflation “works” in the standard model is that it increases the real money supply, which leads to lower interest rates; in effect, it acts like an expansionary monetary policy.

It depends what you mean by standard model. Some economists would put the Pigou Effect into the standard model. Krugman can argue–and has argued–that the Pigou Effect does not apply in Japan. Anyway, I personally don’t stake my case against the liquidity trap on the Pigou effect.

But the standard model actually has another channel by which lower prices raise demand, imparting a downward slope to the aggregate demand curve. It comes from the trade balance. If wages and prices go down in Japan, and this is not offset by an appreciation of the yen, then Japanese goods become cheaper in America and American goods become more expensive in Japan. As a result, the demand for Japanese output goes up. If Krugman has an argument that refutes this, I would like to see it.

Not surprisingly, Scott Sumner has a take.

The BOJ has produced 20 years worth of adverse AD shocks. In both 2000 and 2006 they raised interest rates despite the fact that Japan was experiencing deflation. In 2006 they cut the monetary base by 20%.

So, it’s a sequence of aggregate demand curve shifts. If the Bank of Japan would just hold still for a sec, the economy would find its way back to the long-run aggregate supply curve. As usual, Sumner is coherent, but I am not persuaded.

Smith continues,

But I don’t think Japan is living in an RBC world either. Because in an RBC world, keeping interest rates at zero for decades, and printing a bunch of money (as the Bank of Japan did in the mid-2000s), should cause inflation (without helping growth). Instead, we see persistent deflation. So an RBC model of the common type can’t be describing Japan’s world either.

RBC being “real business cycle,” in which slow productivity growth is the driver of a recession. For me personally, this is even less plausible than the Keynesian story. PSST is not RBC.

I think we need to get away from static thinking, including AS-AD and RBC. In static thinking, there is a full-employment equilibrium out there, if everyone would just adjust to it (match the right person with the right job, or cut wages by enough, or whatever). In the dynamic world of PSST, new opportunities to reconfigure production constantly arise. Some of these create ZMP situations, in which (some) workers’ value to the firm drops essentially to zero. These workers are released into the economy as free resources. Entrepreneurs can try to pick up these free resources and do something profitable with them. But it may take some time for entrepreneurs to figure out exactly what this “something profitable” might consist of.

I know nothing about Japan. But if I were looking for the source of its problems, I would examine the cultural, legal, and institutional factors that surround the formation of new businesses. If what you had during Japan’s post-war resurgence was an economy based on top-down industrial policy and cronyism, and what you need to fix the problem today is bottom-up entrepreneurial energy and creativity, then, as Noah suspects, the solution is not going to come from wiggling interest rates and deficits.

Nominal GDP and Employment

Posted on by

The chart comes from Dan Diamond. Pointer from Tyler Cowen.

Let’s pretend that health care is the whole economy. The top line is the growth rate of nominal GDP. The lower line is the growth rate of employment. Growth in nominal GDP is growth in real GDP plus inflation. Growth in real GDP is growth in number of workers plus growth in output per worker. Inflation is growth in compensation per worker plus growth in the price markup over compensation. Putting this all together, we have

growth of nominal GDP = (growth of number of workers + growth of output per worker) + (growth of compensation per worker + growth of the price markup over compensation)

Scott Sumner would say that the two most reliable numbers here are the ones shown in the chart–growth in nominal GDP and growth in the number of workers. The division between nominal GDP and real GDP depends on making the correct quality adjustment for prices, which Sumner would argue is much less reliable than the other two measures. (I think everyone would agree that quality-adjustment is less reliable, but some of us prefer to believe that it is not much less reliable.)

The difference between the two lines on the chart consists of productivity growth, wage growth, and growth in the price markup. Diamond says that wage growth does not account for the slowdown in nominal GDP (although wage growth did decline–I think by about a percentage point, based on the data in Diamond’s link). He alludes to a mix shift. If people shift away from prescription drugs and toward other services, those other services could have lower productivity and/or a lower price markup.

In any case, it looks as if either productivity growth has declined or the growth in the price markup has declined. This should show up as a decline in corporate profits in the health care industry. Can anyone find data? I have trouble navigating the Commerce Department’s web site.

I did stumble across this paper, which tries to decompose the rise in health care spending from 2003 to 2007.

Our decomposition also sheds light on productivity in the treatment of cancer. Over the four-year sample period, expenditure per capita rose twice as fast for malignant neoplasms (48 percent growth in expenditure per capita) than non-malignant neoplasms (24 percent growth in expenditure per capita). A large reason for the discrepancy is the difference between growth in the cost of treatment (that is, expenditure per episode of care). Service prices for malignant neoplasms grew over twice as fast as service prices for non-malignant neoplasms. This may indicate that more expensive and innovative services are playing a role in cancer spending growth.

This is interesting, but for present purposes it is of little use. The chart above shows a slowdown in the rate of growth in overall health spending between 2003 and 2007.

But my main point is that we expect nominal GDP growth and employment growth to line up. If they do not, something must be going on with either productivity, compensation, or price markups. This is a matter of accounting.

Sumner vs. Williams

Posted on by

John Williams writes,

The intuition for policy attenuation is that uncertainty about the effects of policy creates ex post policy errors that cause economic outcomes to differ from the policymaker’s intentions. The magnitude of the policy error is multiplicative in the policy action; that is, the larger the action, the greater the expected squared error. Therefore, the expected size of the policy error is affected by the size of the policy action, creating a bias toward muted policy actions.

Think of the shower-tuning analogy. Suppose the water is too cold. If you know exactly how the water temperature will change as you move the knob, you turn it quickly. If you are not sure how much the water temperature will change for a given amount of turning, you turn the knob more slowly, to avoid getting scalded.

Scott Sumner comments,

Williams misses the bigger picture, those “demand shocks” were contractionary Fed policy. More specifically they were caused by the failure of the Fed to do NGDP level targeting. The Fed set the wrong target in each year, and this caused the vast majority of the “demand shocks”. With a policy of NGDPLT along a five percent trend line, the recession would have been far milder, with unemployment probably peaking at 6 to 7 percent.

A digression for people of the concrete steppes. No, it wasn’t just “errors of omission”. After growing at 5 percent per year during the Great Moderation, the Fed brought growth in the base to a sudden halt from July 2007 to April 2008. Yet the Fed saw itself as a valiant knight fighting off recession by cutting the Fed funds rate, as if interest rates were a reliable measure of the stance of monetary policy. They aren’t. But even if they were, the Fed drove real interest rates sharply higher in the second half of 2008, a time when they weren’t at the zero bound. So there were plenty of affirmative actions taken by the Fed to drive us into a deep slump. BTW, the base isn’t a reliable indicator either, only NGDP expectations count.

Sumner thinks that the Fed ignored the temperature of the water and just stared at the knob.