As a non-economist statistician (in training), I've really been appreciating your real-time and extremely uncertainty-conscious coverage of these trends. Makes me wish a bit that I was doing macro right now!
However, I don't quite follow the logic about price controls here. If one set off a feedback loop of rising price expectations (e.g. with a nasty but convincing rumor) without changing the amount of goods which are demanded or which can be produced in the economy, it seems to me that short-circuiting the feedback loop by inserting a temporary but strict price control would do exactly the thing you want (stop the expectation feedback loop).
As long as you don't set the price level below their levels prior to the inflationary spiral, I don't see why this would lead to emptying store shelves (again assuming that the real levels of aggregate supply and demand are unchanged).
I can see why this kind of collapse would happen if aggregate demand were actually increasing (relative to aggregate supply). But isn't the "demand rising while supply struggles to catch up" scenario exactly the transitory scenario in which we're not so concerned anyway? So I don't understand why one couldn't say that either we have a transitory supply bottleneck, in which case we can just wait, or we have an expectation contagion which we can control with price controls?
I understand that in reality we probably have some mix of the two and some care should be taken in selecting a level for price controls that isn't too far from whatever equilibrium we would reach at the end of the transitory supply constraints, but that seems like an empirical problem, not a conceptual one. Anyway, as I said, I'm learning all this stuff as I go, so I'm really just curious to see what I might be missing here.
>after you adjust for inflation it turns out that real wages are going down. That’s not good, and economists ought to be thinking much more about why nominal wages can’t seem to keep up when inflation spikes.
I mean, I don't know if this generalizes, but in this case specifically, how can you not blame the covid 19 crisis which reduced global output as things were less efficiently produced under covid restrictions?
Isn't it normal, when there is less to buy, that prices should rise a little?
Isn't it the best case scenario, for central banks to let prices adjust upwards a bit (and real wages adjust downwards a bit, fixing the problem of nominal rigidities) instead of wage stickiness resulting in chaotic rifts in the labor market where, yeah, some might have kept a few percentage more of real wages but many others' wages would have dropped to zero as they succumbed to unemployment.
The global pie was shrunken by covid, a slight adjustment in real prices, is probably the smoothest way for the smaller pie to get distributed. Keeping employers starved for labor is the best way to speed up the recovery.
Real unexpected supply constraints should cause a bit of inflation, even when central banks are not over-printing. In such an broad crisis, the best we might be able to do is to work on solving those production inefficiencies. We should not have central banks throw a monkey wrench in the labor markets to prevent temporary price adjustments that should be happening given the very real circumstances. Prices can rise, and real wages can drop, even under insufficient monetary aggregate demand, if the problems are physically real, not just nominal.
I definitely agree that improvements in supply-side productivity and efficiency trend to manifest in increased economic activity rather than disinflation, but I want to push back a little bit here on the idea that the Fed is "dogmatically insisting that inflation will go back down because they want to support Congress’ spending plans."
Firstly I want to say that we should be focusing on the PCE indexes when trying to estimate inflation accurately - after all, that is what the Fed is targeting and it is a more representative basket with better methodology. While PCE inflation clocked in at 6.3% annualized in June, trimmed mean PCE only hit 2.3%. That is the essence of transitory inflation - a few items with extreme price fluctuations are pushing the headline indexes up while mean inflation remains low. I think it is likely we will experience several more months of changing headline inflation, but we are not currently experience abnormally high "core" inflation.
Second, you should not use the "real compensation per hour" index from FRED that you linked. That index does not use PCE as a deflator and has suffered from serious composition problems as people have left and rejoined the workforce. It would be much better to use the Employment Cost Index, which does not suffer from the same composition problems, in conjunction with PCE.
Finally, I really see no signals of long-term inflation risk. 10 year TIPS breakeven inflation expectations are at 2.4% and 5 year expectations are 2.5%. Given that TIPS track CPI (which tends to show higher inflation than PCE) and breakevens contain risk premiums that increase their values above *true* inflation expectations, it is perfectly likely that real inflation expectations are actually below 2%. The DKW and Cleveland Fed models even show inflation expectations at 1.58% and 1.84% respectively. Given the level of unemployment, the size of the output gap, and well-anchored expectations, the risk of too low long term inflation is still larger than the risk of too high long-term inflation.
I think he is making a pretty strong case here that managing inflation by improving the functioning of the supply side is, if not some silver bullet or the single most important thing we could be doing, still _an_ important thing that we _should_ be doing, in order to keep the economy on track in the brave new world of full employment.
Dylan Matthews had a piece in Vox (and was on the Weeds) discussing the Regulation Q thesis of 1970s stagflation.
What do you think of that explanation, ie Volcker rate hikes didn’t reset expectations but rather the interest rate feedback loops were broken and once those were fixed, ie Reg Q repealed, regular monetary policy was able to smooth out the business cycle?
I know your piece isn’t focused on stagflation but if the Reg Q thesis is true, then the Fed just has much more power via interest rates to break high inflation than we previously thought. So we could get a softer landing than what we got with Volcker.
Inflation can be fiscal and/or monetary and still be sectoral. It's well understood that monetary stimulus especially affects homes and cars. The spending of elevated savings post(ish)-pandemic and stimulus checks also seems to be concentrated on cars and homes. The Fed has I think been deliberately goosing housing and cars to compensate for sectors that by nature will recover more slowly. And also this does indeed come amid supply issues, pandemic inspired second home buying, crypto-exacerbated chip shortages.
There is a big difference between persistent high inflation, like the US had in the 70s, Greece and Italy had before the Euro, or Turkey again has now, and hyperinflation. The drivers you described cause persistent high inflation, not hyperinflation, which is driven by extreme continuous increases in nominal fiscal spending, typically amid a collapse of fiscal revenue. The US was not at risk of hyperinflation in the 70s. Unless Biden comes out with a 35 trillion proposal we don't need to consider hyperinflation risk
When I took macro the Prof hammered home the notion of a fixed basket of goods. And in fact we probably spent close to two weeks going over the various indexes people use and each time we use the same basket. But why is there such variance in the the goods people choose to measure inflation? Is it more indicative of broader systemic economic trends? Also don’t the current measures take into account substitution hedonics and such? How from quarter to quarter can we get a feel for these types of preference trade offs?
I agree that one can't just eliminate the biggest sources of inflation and then say that inflation is fine. But wouldn't a bit more disaggregation of the components of current inflation be at least somewhat educational? We know about huge surges in used car prices, just as previously there was a spectacular increase in lumber prices (followed by a spectacular collapse: https://www.calculatedriskblog.com/2021/08/update-framing-lumber-prices-down-year.html). Has anyone looked at the dominance of certain commodities/products in current inflation and compared that to previous bouts of inflation, especially early on? My guess is that there are some highly unusual spikes in a limited set of areas, and that this buttresses the transitory argument.
I agree you can't fight inflation with supply-side policy. But that's because inflation is a macro phenomena. If the CPI is going up because of car manufacturing bottlenecks and a computer chip foundry fire, that calls for a supply-side solution and it's not really inflation.
My understanding of the argument for fighting inflation through supply-side policy is that there is a model where supply shock -> price increase -> inflationary expectations -> inflation. If we have this model in mind, wouldn't it make sense to attack the underlying supply problems early, before inflationary expectations set in? This wouldn't necessarily substitute for monetary policy in the long-run, but it could hold off the need to tighten for a while.
I guess I also think though that we really, really need a proper boom right now, and it's worth risking a lot to keep it going. It seems to me that any statement from the fed that would hold down inflationary expectations would also act as a signal to markets that the party is over.
As a non-economist statistician (in training), I've really been appreciating your real-time and extremely uncertainty-conscious coverage of these trends. Makes me wish a bit that I was doing macro right now!
However, I don't quite follow the logic about price controls here. If one set off a feedback loop of rising price expectations (e.g. with a nasty but convincing rumor) without changing the amount of goods which are demanded or which can be produced in the economy, it seems to me that short-circuiting the feedback loop by inserting a temporary but strict price control would do exactly the thing you want (stop the expectation feedback loop).
As long as you don't set the price level below their levels prior to the inflationary spiral, I don't see why this would lead to emptying store shelves (again assuming that the real levels of aggregate supply and demand are unchanged).
I can see why this kind of collapse would happen if aggregate demand were actually increasing (relative to aggregate supply). But isn't the "demand rising while supply struggles to catch up" scenario exactly the transitory scenario in which we're not so concerned anyway? So I don't understand why one couldn't say that either we have a transitory supply bottleneck, in which case we can just wait, or we have an expectation contagion which we can control with price controls?
I understand that in reality we probably have some mix of the two and some care should be taken in selecting a level for price controls that isn't too far from whatever equilibrium we would reach at the end of the transitory supply constraints, but that seems like an empirical problem, not a conceptual one. Anyway, as I said, I'm learning all this stuff as I go, so I'm really just curious to see what I might be missing here.
>after you adjust for inflation it turns out that real wages are going down. That’s not good, and economists ought to be thinking much more about why nominal wages can’t seem to keep up when inflation spikes.
I mean, I don't know if this generalizes, but in this case specifically, how can you not blame the covid 19 crisis which reduced global output as things were less efficiently produced under covid restrictions?
Isn't it normal, when there is less to buy, that prices should rise a little?
Isn't it the best case scenario, for central banks to let prices adjust upwards a bit (and real wages adjust downwards a bit, fixing the problem of nominal rigidities) instead of wage stickiness resulting in chaotic rifts in the labor market where, yeah, some might have kept a few percentage more of real wages but many others' wages would have dropped to zero as they succumbed to unemployment.
The global pie was shrunken by covid, a slight adjustment in real prices, is probably the smoothest way for the smaller pie to get distributed. Keeping employers starved for labor is the best way to speed up the recovery.
Real unexpected supply constraints should cause a bit of inflation, even when central banks are not over-printing. In such an broad crisis, the best we might be able to do is to work on solving those production inefficiencies. We should not have central banks throw a monkey wrench in the labor markets to prevent temporary price adjustments that should be happening given the very real circumstances. Prices can rise, and real wages can drop, even under insufficient monetary aggregate demand, if the problems are physically real, not just nominal.
I definitely agree that improvements in supply-side productivity and efficiency trend to manifest in increased economic activity rather than disinflation, but I want to push back a little bit here on the idea that the Fed is "dogmatically insisting that inflation will go back down because they want to support Congress’ spending plans."
Firstly I want to say that we should be focusing on the PCE indexes when trying to estimate inflation accurately - after all, that is what the Fed is targeting and it is a more representative basket with better methodology. While PCE inflation clocked in at 6.3% annualized in June, trimmed mean PCE only hit 2.3%. That is the essence of transitory inflation - a few items with extreme price fluctuations are pushing the headline indexes up while mean inflation remains low. I think it is likely we will experience several more months of changing headline inflation, but we are not currently experience abnormally high "core" inflation.
Second, you should not use the "real compensation per hour" index from FRED that you linked. That index does not use PCE as a deflator and has suffered from serious composition problems as people have left and rejoined the workforce. It would be much better to use the Employment Cost Index, which does not suffer from the same composition problems, in conjunction with PCE.
Finally, I really see no signals of long-term inflation risk. 10 year TIPS breakeven inflation expectations are at 2.4% and 5 year expectations are 2.5%. Given that TIPS track CPI (which tends to show higher inflation than PCE) and breakevens contain risk premiums that increase their values above *true* inflation expectations, it is perfectly likely that real inflation expectations are actually below 2%. The DKW and Cleveland Fed models even show inflation expectations at 1.58% and 1.84% respectively. Given the level of unemployment, the size of the output gap, and well-anchored expectations, the risk of too low long term inflation is still larger than the risk of too high long-term inflation.
Have you read MattY's most recent?
https://www.slowboring.com/p/tariffs-inflation
I think he is making a pretty strong case here that managing inflation by improving the functioning of the supply side is, if not some silver bullet or the single most important thing we could be doing, still _an_ important thing that we _should_ be doing, in order to keep the economy on track in the brave new world of full employment.
Dylan Matthews had a piece in Vox (and was on the Weeds) discussing the Regulation Q thesis of 1970s stagflation.
What do you think of that explanation, ie Volcker rate hikes didn’t reset expectations but rather the interest rate feedback loops were broken and once those were fixed, ie Reg Q repealed, regular monetary policy was able to smooth out the business cycle?
I know your piece isn’t focused on stagflation but if the Reg Q thesis is true, then the Fed just has much more power via interest rates to break high inflation than we previously thought. So we could get a softer landing than what we got with Volcker.
I don't agree with the conclusion of this article.
We are already well into the wage-price spiral portion - tough talk from the Fed won't do anything.
Much agreed. But two points to make on details.
Inflation can be fiscal and/or monetary and still be sectoral. It's well understood that monetary stimulus especially affects homes and cars. The spending of elevated savings post(ish)-pandemic and stimulus checks also seems to be concentrated on cars and homes. The Fed has I think been deliberately goosing housing and cars to compensate for sectors that by nature will recover more slowly. And also this does indeed come amid supply issues, pandemic inspired second home buying, crypto-exacerbated chip shortages.
There is a big difference between persistent high inflation, like the US had in the 70s, Greece and Italy had before the Euro, or Turkey again has now, and hyperinflation. The drivers you described cause persistent high inflation, not hyperinflation, which is driven by extreme continuous increases in nominal fiscal spending, typically amid a collapse of fiscal revenue. The US was not at risk of hyperinflation in the 70s. Unless Biden comes out with a 35 trillion proposal we don't need to consider hyperinflation risk
For us dummies. What does the Fed do to address inflation?
When I took macro the Prof hammered home the notion of a fixed basket of goods. And in fact we probably spent close to two weeks going over the various indexes people use and each time we use the same basket. But why is there such variance in the the goods people choose to measure inflation? Is it more indicative of broader systemic economic trends? Also don’t the current measures take into account substitution hedonics and such? How from quarter to quarter can we get a feel for these types of preference trade offs?
I agree that one can't just eliminate the biggest sources of inflation and then say that inflation is fine. But wouldn't a bit more disaggregation of the components of current inflation be at least somewhat educational? We know about huge surges in used car prices, just as previously there was a spectacular increase in lumber prices (followed by a spectacular collapse: https://www.calculatedriskblog.com/2021/08/update-framing-lumber-prices-down-year.html). Has anyone looked at the dominance of certain commodities/products in current inflation and compared that to previous bouts of inflation, especially early on? My guess is that there are some highly unusual spikes in a limited set of areas, and that this buttresses the transitory argument.
I agree you can't fight inflation with supply-side policy. But that's because inflation is a macro phenomena. If the CPI is going up because of car manufacturing bottlenecks and a computer chip foundry fire, that calls for a supply-side solution and it's not really inflation.
Assets up, real wages down. Inflation is upward redistribution at its best.
My understanding of the argument for fighting inflation through supply-side policy is that there is a model where supply shock -> price increase -> inflationary expectations -> inflation. If we have this model in mind, wouldn't it make sense to attack the underlying supply problems early, before inflationary expectations set in? This wouldn't necessarily substitute for monetary policy in the long-run, but it could hold off the need to tighten for a while.
I guess I also think though that we really, really need a proper boom right now, and it's worth risking a lot to keep it going. It seems to me that any statement from the fed that would hold down inflationary expectations would also act as a signal to markets that the party is over.
What do Smets Wouters’07 or Slobodyon Wouters’12 say?