I say misguided, although many smart people think the Fed is intentionally undershooting its 2% inflation target, or treating it like a ceiling.  That would of course be evil, because it would mean the Fed is lying. Call me naive, but I still don’t quite accept that the Fed is a Trump-like institution. I believe they are misguided.

So what are the implications of my theory?  How can we test it?

If’ I’m right, then I believe that the Fed will eventually see that its reliance on the Phillips curve model has been a mistake. Low unemployment does not cause high inflation.  I expect this realization to occur at some point during 2018, at which time the Fed will switch to an easier money policy—to boost inflation.  I believe this because the market believes it, and (like Larry Summers) I’m a market monetarist.

This is one reason why I expect this expansion to be the longest in American history.  It won’t be the best (the 60s, the 80s, and the 90s were all better), but it will be the longest.  Switching to an easier money policy in the 9th year of an expansion is unusual.  It will prolong the expansion for at least a few more years.

I do not expect the Fed’s undershoot of inflation to cause a recession (although I wouldn’t entirely rule it out–it just seems unlikely.)  The economy has basically adjusted to 1.7% inflation.  The real problems with this are:

1.  A loss of Fed credibility, which will hurt them when the next crisis occurs.

2.  More zero bound episodes.

So the Fed needs to fix this problem.

BTW, there is nothing intrinsically wrong with 1.7% inflation during a period of low unemployment, if the Fed is a flexible inflation targeter.  Indeed in a sense that’s desirable.  But only if the Fed runs above 2% inflation during recessions.  And that’s why the Fed’s Phillips curve thinking is so pernicious.  The Fed fully expects inflation to fall during the next recession—the opposite of what they should be doing.  In that case the Fed needs to generate above 2% inflation during booms, in order to average 2% over the entire business cycle.  They are not doing so.

PS.  Stephen Kirchner directed me to an excellent Martin Sandbu column in the FT.  It does a great job analyzing the recent letter calling for a higher inflation target.  Indeed a far better job than I did in my recent analysis.  Here is the conclusion:

None of this means the target should not be reconsidered. But if there is going to be a change to what the Fed aims to achieve, one can do much better than a higher inflation rate target. One attractive possibility is to target a steadily growing price level rather than an inflation rate, which would require policymakers to pursue higher-than-target inflation for a while to make up for lower-than-target inflation in the past. Another is to consider targeting a path for the nominal size of the economy — nominal gross domestic product level targeting — which would allow for greater monetary stimulus when it is likely to do the most good.

PS.  The Larry Summers link above may be of interest to some people.

For several days I’ve been reading about a letter by 22 “liberal” (i.e. left-of center) economists, calling for a higher inflation target.  That’s not really my first choice, although I do see their point.  When I finally read the letter I was pleasantly surprised by its tone, which was quite moderate:

Even if a 2 percent inflation target set an appropriate balance a decade ago, it is increasingly clear that the underlying changes in the economy would mean that, whatever the correct rate was then, it would be higher today. To ensure the future effectiveness of monetary policy in stabilizing the economy after negative shocks – specifically, to avoid the zero lower bound on the funds rate – this fall in the neutral rate may well need to be met with an increase in the long-run inflation target set by the Fed.

As you know, I favor NGDP targeting.  But let’s suppose that we decided back in the early 2000s on a 4.5% NGDP target, expecting it to deliver 1.5% trend inflation.  Today we’d expect that same NGDP target to deliver roughly 3% inflation.  So in that sense they are correct—if we must target inflation, then due to changing circumstances a higher rate would now be appropriate.

There is another sense in which they are also correct.  The original call for a 2% target, instead of say 1% or zero, was partly based on the assumption that 2% was high enough to keep us away from the zero interest rate bound.  Japan had a zero percent inflation target, and was persistently at or close to the zero bound.  So if that was the criterion used for the 2% target in the early 2000s, then the exact same criterion would call for a higher target rate today.

And their concluding recommendations are quite modest and reasonable:

Policymakers must be willing to rigorously assess the costs and benefits of previously-accepted policy parameters in response to economic changes. One of these key parameters that should be rigorously reassessed is the very low inflation targets that have guided monetary policy in recent decades. We believe that the Fed should appoint a diverse and representative blue ribbon commission with expertise, integrity, and transparency to evaluate and expeditiously recommend a path forward on these questions. We believe such a process will strengthen the Fed as an institution and its conduct of monetary policy, and help ensure wise policymaking for the years and decades to come.

Isn’t the Congressional GOP proposing a similar commission?

The title of this blog post might seem a bit provocative, but that’s not my intention. These 22 left-of-center economists have signed a letter calling for the Fed to make its policy regime so robust that fiscal stimulus is no longer needed, indeed is no longer even effective. That’s effectively a “conservative” outcome.  I’m not sure that’s what they all intended—presumably people like Joe Stiglitz did not have that intention—but fiscal policy becomes superfluous when the Fed is not stuck at the zero bound (as we learned during the 1984-2007 period).  That’s not just my view, Paul Krugman used to make much the same argument.

If I wrote the letter it would have been much shorter:

Dear Fed, Please choose one of the half dozen policy regimes that insure monetary policy is always capable of providing the desired level of aggregate demand, so that fiscal stimulus is never needed.

PS.  I also have an Econlog post on Fed policy.

PPS.  At the half time of game 4, Paul Pierce had a look of wonder in his eyes.  Referring to Cleveland’s 86 points at the half, he said something to the effect, “My reaction is that’s what it takes to beat the Warriors?”  If the public understood just how small an adjustment it would take to fix monetary policy they have the opposite reaction: “That’s all it would have taken to prevent the Great Recession?”  The policy elite better hope the public never finds out.

Vaidas Urba directed me to a very interesting talk by Vítor Constâncio, Vice-President of the European Central Bank. Here’s one item that caught my eye:

Increasing the inflation target real interest rates could then be effective to eliminate negative output and unemployment gaps. These benefits of a higher inflation target can be outweighed by the broader costs of higher inflation, depending on the chosen level. Historically, relatively higher inflation has usually been associated with more volatile inflation.^[30] Moreover, the ECB and many other inflation-targeting central banks have shown that other monetary policy tools are available when interest rates cannot be lowered further.

It’s true that the availability of alternative tools makes a higher inflation target unnecessary, but the first argument is a bit misleading.  While its true that “Historically, relatively higher inflation has usually been associated with more volatile inflation”, this isn’t actually relevant to the debate over whether the inflation target should be set at 2%, 3% or 4%.  Yes, inflation was both higher and more volatile during the 1960s-80s, but inflation was lower and more volatile during the gold standard era.

The greater stability of inflation since 1990 is due to the fact that central banks have started targeting inflation, whereas there was no consistent inflation target under either the Great Inflation or the gold standard.  Inflation will be more stable when it’s being targeted, regardless of whether the target is set at 2%, 3% or 4%. Now if you are talking about an extremely high target, say 17%, then I’d expect more volatility, as it’s unlikely the next government or central bank head would agree with that sort of specific number.  The real issue is not how high the target is, but rather the degree of consensus.  Maybe there is currently more consensus around 2% than 3%, but if the zero bound continues to be an issue then that consensus may reverse in the future.

The discussion also mentions NGDPLT, but doesn’t really offer any useful analysis of that proposal.

David Levey directed me to an interesting Vox essay by Athanasios Orphanides, which compares the debt situation in Japan and Italy:

For Japan, the dramatic rise of the debt ratio before the crisis reflects the lack of nominal growth.  While the long-term government bond yield appeared to be low (consistently below 2%), nominal GDP growth was even lower (about zero, on average). The adverse debt dynamics worsened after 2007, with the recession following the Global Crisis. Part of the problem was overly tight monetary policy: policy rates were constrained by the zero lower bound (ZLB), but the Bank of Japan was reluctant to employ the required QE policies. However, since 2013 the Bank of Japan has embarked on a decisive QE programme which has simultaneously boosted nominal GDP growth and depressed long-term government bond yields. Since September 2016, as part of its ‘Quantitative and Qualitative Monetary Easing with Yield Curve Control’ policy, the Bank of Japan has communicated explicitly its intention to keep the 10-year yield on government bonds close to zero and short-term interest rates negative until inflation rises to 2%, in line with its definition of price stability. This monetary policy has stabilised Japan’s debt dynamics and has provided the Japanese government more time to implement structural reform measures and complete the fiscal adjustment needed to bring its primary deficit under control.

Abenomics has modestly boosted inflation in Japan, but the rate remains below the BOJ’s 2% target.  In another sense, however, the policy has been a big success. Unemployment has fallen to 2.8%, NGDP growth has accelerated, and the debt ratio has stopped rising.  Japan is no longer on the road to bankruptcy.

That’s why I favored the monetary “arrow” of Abenomics, and I’m pleased to see even a partial success.  In contrast, Italy lacks its own currency and will have to combine fiscal austerity with supply-side reforms to solve its problems.  That’s much tougher to achieve.

PS.  Hester Peirce is my colleague at the Mercatus Center, where she focuses on financial regulatory issues.  In this link, she’s on a panel with Ben Bernanke, discussing Dodd-Frank provisions such as the “orderly liquidation authority”.

I just listened to a very interesting Macro Musings podcast—with David Beckworth interviewing Tyler Cowen.  Tyler sees the Fed passivity during the last decade as part of a broader increase in risk aversion across American society, part of what he calls the rise in complacency.  I like that hypothesis, and look forward to reading his new book on my vacation next week.

I’m less convinced by his argument that this also explains the public’s obsession with low inflation, and refusal to contemplate NGDP targeting.  I would argue that the public is not opposed to NGDP targeting, indeed they’ve never even heard of the idea.  So let’s try to disentangle what Tyler had in mind in his comments on this subject:

1.  The public might prefer the specific policy of inflation targeting to the specific policy of NGDP targeting.

2.  The public might prefer the outcome (in terms of economic performance) of inflation targeting over the outcome produced by NGDP targeting.

The first interpretation is a complete nonstarter.  If you asked the average American to discuss strict inflation targeting, flexible inflation targeting, symmetric inflation targeting, asymmetric targeting, growth rate targeting, level targeting, etc., you’d produce a glazed look on their faces.  And even that understates the problem.  Most people only understand the concept of supply side inflation; they have absolutely no understanding of demand side inflation.  Hence they think inflation is bad because it lowers their real income, but that’s only true of supply side inflation.  Even worse, the Fed only controls demand side inflation, so Fed policy has no impact on the only kind of inflation the public understands.  I used to ask my class whether the cost of living had actually increased if both wages and prices rose by exactly 10%.  And over 95% always got it wrong, claiming that the cost of living had not actually increased.  (Actually, it rose by 10%.)

And even that understates the problem, because very few Americans even understand that the Fed can choose between 2% and 4% trend inflation like someone choosing between massaman curry and pad thai for their entree.  They don’t know that that is what the Fed does.  How many Americans know that the Fed’s tight money caused the Great Recession?

So I’m going to assume that what Tyler meant was that Americans would be happier with the outcome of inflation targeting as compared to the outcome of NGDP targeting.  I doubt that.  People might say they prefer 0% inflation to 2% inflation, but I doubt that they’d say they prefer America’s 2009 economy to its 2007 economy.  But the 2009 economy is what you get when you reduce inflation to 0%.  Actually, we were much closer to NGDP targeting in the period before 2007 than in the years immediately after 2007.  And yet Americans seemed much more unhappy with the post-2007 economy.  Now of course it’s entirely possible that if the Fed had kept NGDP growing at 5% after 2007, the public would have been even angrier than they were with the actual policy (a fall of 3% in NGDP between mid 2008 and mid-2009.)  But I doubt it.  I think they would have been a bit disgruntled by the stagflation, but nothing more.

Both David and Tyler believe that the Fed now views 2% inflation as a ceiling, not a symmetrical target.  I’m still not convinced; let’s look at this again in 10 years.  If it’s still a ceiling, I’ll throw in the towel.  It was clearly a symmetrical target before 2007—why would the Fed have suddenly changed?

PS.  David and Tyler also discussed the declining rate of innovation in the arts.  Here’s a quote from a different Tyler Cowen interview:

If you think about Renaissance Florence, at its peak, its population, arguably, was between 60,000 and 80,000 people. And there were surrounding areas; you could debate the number. But they had some really quite remarkable achievements that have stood the test of time and lasted, and today have very high market value. Now, in very naive theories of economics, that shouldn’t be possible. People in Renaissance Florence, they didn’t produce a refrigerator that we’re still using or a tech company that we still consult.

But there’s something different about, say, the visual arts, where that was possible, and it was done with small numbers. So there’s something about the inputs to some kinds of production we don’t understand. I would suggest if we’re trying to figure out, like what makes Silicon Valley work, actually, by studying how they did what they did in the Florentine Renaissance is highly important. You learn what are the missing inputs that make for other kinds of miracles.

That’s an interesting point–future generations will be more interested in our art than our technology.

Here’s another question to consider. In the arts, there are periods of rapid innovation (the Renaissance), followed by periods of stasis. In David’s interview, Tyler cites the rapid innovation in pop music during the 1960s, relative to the slower innovation today. Is that different from the observation that explorers no longer discover as many new lands as they did in the heyday of Portuguese and Spanish exploration?  Are we less talented at exploration, or are there simply fewer as yet undiscovered lands? Perhaps the analogy is silly, but let’s take it a step further.  The discovery of new planets, and the technology that allows us to get there, would presumably lead to another Golden Age of discovery.  Doesn’t the invention of new art forms (the novel, film, electronic music, etc.) open up vast new fields for artistic discovery? Isn’t Robert Gordon’s argument that innovation in existing fields has diminishing returns, and that we need to develop entirely new fields to supercharge innovation?  Thus we understand the physics of the non-microscopic world so well that it’s getting really hard to radically improve our houses, cars, airplanes, ships and washing machines.  The only area of rapid innovation is at the microscopic level (biotech, computer chips, etc.), which opened up only in the past 50 years or so.  Gordon doesn’t expect more such fields to open up, and thus predicts diminishing rates of innovation in the fields that we already have.  (I’m agnostic on that claim.)

The following Venn diagram helps to explain how I visualize macro:

There are three basic fields within macro:

1.  Equilibrium nominal

2.  Equilibrium real

3.  Disequilibrium sticky wage/price (interaction)

I’ll take these one at a time.

1. Within equilibrium nominal there are important concepts:

A.  The quantity theory of money

B.  The Fisher effect

C.  Purchasing power parity

The first suggests that a change in M will cause a proportionate change in P.  The second suggests that a change in inflation will cause an equal change in nominal interest rates.  The third suggests that a change in the inflation differential between two countries will cause an equal change in the rate of appreciation of the nominal exchange rate.

All three concepts implicitly hold something constant; either the real demand for money, the real interest rate, or the real exchange rate.  In all three cases the concept is most useful when the money supply and price level are changing very rapidly, especially if those changes persist for long periods of time.

2.  Equilibrium real macro can be thought of as looking at economic shocks that do not rely on wage/price stickiness.  These include changes in population, technology, capital, preferences, government policies, weather conditions, taxes, etc.  These can cause changes in the real demand for money, the real interest rate, the real exchange rate, the unemployment rate, real GDP, and many other real variables.

3.  Disequilibrium macro looks at nominal shocks that cause changes in real variables, but only because of wage/price stickiness.  Thus because prices are sticky, an increase in the money supply will temporarily cause higher real money demand, a lower real interest rate, and a lower real exchange rate.  These changes occur even if there is no fundamental real shock hitting the economy.  The effects are temporary, and go away once wages and prices have adjusted.

If wages and prices are sticky then an increase in the money supply will also cause a temporary rise in employment and real output.

And that’s basically all of macro.  (This is how I’d try to explain macro to a really bright person, if I were given only 15 minutes.)

I also believe that understanding the implications of this three part schema makes one a better macroeconomist. Talented macroeconomists like Paul Krugman tend to have good instincts as to which real world issues belong in each category. Here’s my own view on a few examples. For simplicity, I’ll denote these three areas: nominal, real, and interaction:

1.  Most business cycles in ancient times were real, with some interaction.  The 1500 to 1650 inflation was nominal.

2.  Recessions such as 1893, 1908, 1921 and 1982 were mostly interaction.

3.  The Great Depression was all three.

4.  The Great Inflation was mostly nominal, especially in high inflation countries.

5. The 1974 recession was more “real” than usual.  Ditto for the WWII output boom.

6.  The real approach works best for short run shocks to specific industries such as housing and oil, plus long run growth.  The nominal approach works best for high inflation rates and long run inflation.  The interaction approach works best for real GDP fluctuations in large diversified economies.

7.  If one set of economists say the Japanese yen is too strong, and another set say it’s too weak, they are probably using different frameworks.  Those who say its too strong are using a nominal framework, and are likely worried about deflation.  A weaker yen would boost inflation.  Those who think the yen is too weak are using a real framework.  Rather than worry about deflation, they worry that Japan has a current account surplus.  These views seem to contradict, but it’s theoretically possible for both to be right.  Perhaps the nominal exchange rate for the yen is too strong, and the real exchange rate is too weak.  You would then weaken the nominal exchange rate by printing money, and strengthen the real exchange rate by reducing Japanese saving rates.  (I don’t favor the latter, just saying that’s the proper implication of the misguided worry about Japanese CA surpluses.)

8.  Don’t let your policy preferences drive your analysis.  Throughout all of my life, it’s been assumed that monetary shocks drive real output by causing changes in the unemployment rate.  Not changes in trend productivity growth or population growth or labor force participation or any number of other variables.  Money matters because it affects unemployment.  If the unemployment rate is telling you that monetary policy is no longer holding back growth, the proper response is not to double down on your belief that we need easier money and then look for new theories to justify it, but rather to conclude that whatever problems we still have are now “real”, not “interaction.”

A good macroeconomist knows that all three fields of macro are very important, and which models apply to each of the three fields, and which field is most applicable to each real world macro issue.