"the Fed will have to raise rates somewhat to avoid major overheating."
If there is inflation, which is quite possible, it will not be because of an "overheated" economy, it will be because people have what they consider excess savings from the checks and unemployment bonuses, combined with delayed consumption - mostly of things like recreation and entertainment outside the home. The situation would probably be somewhat like that when WW II rationing ended in 1946. There was inflation then, but all sorts of consumer goods had been unavailable for years. Raising interest rates would have no effect on this type of consumption. Even if businesses went crazy because of increased prices and consumption (which would be in limited sectors) there is no reason to believe that "somewhat" increased interest rates would stop them from overexpanding. But most likely any inflation would die down once the excess savings are gone
.
The idea that raising interest rates can easily defeat inflation is a strange myth anyway. The Fed raised federal funds rates for years on end to record levels during the 70's and did not stop inflation from going unacceptably high. Inflation came down only when oil price quit rising, at the end of 1974 and the beginning of 1980.
Shouldn’t you also account for the largely “exogenous” boost to PCE that will come from consumption, largely of services, returning to normal as the virus recedes? This can be funded through the saving rate returning to around 7.5% from the current level. And while some of the increase in services consumption will be accompanied by less spending on goods—like food away from home(PCE services) for food at home(PCE nondurables) — there is still a large net stimulus that will take place after substitution effects. The income to fund this is already present and more will be produced as reopening proceeds. It seems to me that around 2.5% of GDP, or more, is set to return if “herd Immunity” is achieved over the summer, without requiring any additional stimulus.
So how does this post stand up after a year (and a few months)?
Inflation is falling but only after rate rises, unemployment is falling too and there has been no meaningful recession. The IRA amount was considerably less than expected.
How about a review of the projections in the then crystal ball?
FRED is really addictive. Here is a FRED chart that has not appeared in your posts but probably should. M2 grew 25% last year and looks set to grow at least 15% this year. Q.E.D. inflation is a serious (hopefully not an existential) threat.) It don't take a weatherman to know which way the wind blows. https://fred.stlouisfed.org/series/WM2NS
I tried to re-create the “prime-age Phillips curve” because the Phillips curve bothers me a lot. I got the core inflation and prime age employment ratio data from the Federal Reserve Bank of St. Louis. The scatter plot for the years 2000 to 2020 looks very similar to the one in the newsletter, although not exactly the same. There is a visible positive slope to the data. But when the full data set from 1948 to 2020 is plotted, it shows a cloud of points with no obvious trend. There doesn’t seem to be any correlation between the employment ratio and core inflation. The trend line for every 20 year period ending from 1988 to 2014 has a negative slope, meaning a higher employment ratio is associated with lower inflation.
I tried plotting inflation vs unemployment rate from 1948 to 2016, which is also a cloud of points with no apparent correlation.
Milton Friedman said of the Phillips Curve, “it contains a basic defect—the failure to distinguish between nominal wages and real wages”. I agree so I plotted inflation against real (inflation adjusted) median income change. There is a lot of scatter in the data (1948 – 2016) but the trend line has a negative slope so that higher income growth is associated with lower inflation. This is the opposite of the thinking that went into the Phillips curve.
I also looked a United Kingdom data. Plots of inflation vs unemployment show a negative slope between 1971 and 1985 and a positive slope from 1985 to 2019.
The data do not consistently show that higher wages cause inflation, or that low unemployment causes inflation. The Phillips curve theory only works sometimes, depending on what data you look at.
In the United States real median incomes grew rapidly between 1948 and 1973, faster than real GDP per capita. Real median income grew by 2.5% per year. The average inflation rate during that time was 2.4%.
“Our World in Data” shows very strong real median income growth in the United Kingdom between 1996 and 2007. It averaged 3.2% per year. U. K. inflation during that time was 1.87% per year.
There actually has been strong real wage growth without high inflation. This has been tested and proven.
The Phillips curve idea is partly based on the notion that corporations would have to raise prices to pay for higher wages. This is not true. The large corporate tax cut in 2017 was not followed by price cuts (the CPI went up 1.9% in 2018), so why should wage hikes be followed by price increases? Corporations can just reduce their spending on stock buybacks. I don’t think it is a coincidence that the strong wage growth I mentioned earlier happened when stock buybacks were illegal.
The economic theories that wage growth creates inflation, and higher minimum wage costs jobs, and tax cuts create growth, lead to the conclusion that the only viable economy is one in which the peasants do the work and the rich class gets the money. Someone might think that rich people pay economists to come up with these theories. Especially since none of them pan out when you look at the data.
In 1973 the United States had the world’s highest median income. Since then the United States has had one of the worst if not the worst wage growth among first world countries. Several other countries now have higher median incomes including some socialist hellholes like Norway, Sweden, and Denmark. It didn’t have to be like this. The money was there for wages to double, but government policy changes in the 70’s and 80’s diverted the national income from the workers to the rich. One of those policies was to use the Phillip’s curve as a rationale to suppress wages.
"if the relief package turns out to be significantly bigger than needed, the Fed will have to raise rates somewhat to avoid major overheating. And this will be a terrible thing because … why, exactly? True, bond markets might experience something of a shock, something like what happened in 1994 — a shock that burned some investors, but did nothing to derail a very solid real economic recovery"
In response to your question, raising interest rates would be a terrible, terrible thing because, as everyone knows, low interest rates are bad because they cause bubbles...no...wait.....
The ironic thing is that the people worrying about inflation are some of the same people (not Blanchard or Summers) that were complaining about low interest rates punishing coupon clippers.
A conclusion of the Boskin Commission was that the CPI overstates inflation. Therefore, could one reasonably conclude that inflationary pressures are real?
Besides automatic changes in unemployment insurance as employment picks up, are there elements of the American Rescue Plan that can be re-programmed in real time?
Thank you for doing this substack with some depth and that it is independent of NYT.
Paul Krugman,
Your term 'asymmetry of fiscal risks', meaning that :
it's easier> to manage interest rates
than> to approve fiscal rescue,
maybe reworded to 'economic policy asymmetry',
is realist, important and deserves to enter Macroeconomics textbooks.
Congrats!
"the Fed will have to raise rates somewhat to avoid major overheating."
If there is inflation, which is quite possible, it will not be because of an "overheated" economy, it will be because people have what they consider excess savings from the checks and unemployment bonuses, combined with delayed consumption - mostly of things like recreation and entertainment outside the home. The situation would probably be somewhat like that when WW II rationing ended in 1946. There was inflation then, but all sorts of consumer goods had been unavailable for years. Raising interest rates would have no effect on this type of consumption. Even if businesses went crazy because of increased prices and consumption (which would be in limited sectors) there is no reason to believe that "somewhat" increased interest rates would stop them from overexpanding. But most likely any inflation would die down once the excess savings are gone
.
The idea that raising interest rates can easily defeat inflation is a strange myth anyway. The Fed raised federal funds rates for years on end to record levels during the 70's and did not stop inflation from going unacceptably high. Inflation came down only when oil price quit rising, at the end of 1974 and the beginning of 1980.
Thanks Paul K. Your voice is needed again
Shouldn’t you also account for the largely “exogenous” boost to PCE that will come from consumption, largely of services, returning to normal as the virus recedes? This can be funded through the saving rate returning to around 7.5% from the current level. And while some of the increase in services consumption will be accompanied by less spending on goods—like food away from home(PCE services) for food at home(PCE nondurables) — there is still a large net stimulus that will take place after substitution effects. The income to fund this is already present and more will be produced as reopening proceeds. It seems to me that around 2.5% of GDP, or more, is set to return if “herd Immunity” is achieved over the summer, without requiring any additional stimulus.
elucidating, refreshing, aware - thank you, paul k
So how does this post stand up after a year (and a few months)?
Inflation is falling but only after rate rises, unemployment is falling too and there has been no meaningful recession. The IRA amount was considerably less than expected.
How about a review of the projections in the then crystal ball?
FRED is really addictive. Here is a FRED chart that has not appeared in your posts but probably should. M2 grew 25% last year and looks set to grow at least 15% this year. Q.E.D. inflation is a serious (hopefully not an existential) threat.) It don't take a weatherman to know which way the wind blows. https://fred.stlouisfed.org/series/WM2NS
I tried to re-create the “prime-age Phillips curve” because the Phillips curve bothers me a lot. I got the core inflation and prime age employment ratio data from the Federal Reserve Bank of St. Louis. The scatter plot for the years 2000 to 2020 looks very similar to the one in the newsletter, although not exactly the same. There is a visible positive slope to the data. But when the full data set from 1948 to 2020 is plotted, it shows a cloud of points with no obvious trend. There doesn’t seem to be any correlation between the employment ratio and core inflation. The trend line for every 20 year period ending from 1988 to 2014 has a negative slope, meaning a higher employment ratio is associated with lower inflation.
I tried plotting inflation vs unemployment rate from 1948 to 2016, which is also a cloud of points with no apparent correlation.
Milton Friedman said of the Phillips Curve, “it contains a basic defect—the failure to distinguish between nominal wages and real wages”. I agree so I plotted inflation against real (inflation adjusted) median income change. There is a lot of scatter in the data (1948 – 2016) but the trend line has a negative slope so that higher income growth is associated with lower inflation. This is the opposite of the thinking that went into the Phillips curve.
I also looked a United Kingdom data. Plots of inflation vs unemployment show a negative slope between 1971 and 1985 and a positive slope from 1985 to 2019.
The data do not consistently show that higher wages cause inflation, or that low unemployment causes inflation. The Phillips curve theory only works sometimes, depending on what data you look at.
In the United States real median incomes grew rapidly between 1948 and 1973, faster than real GDP per capita. Real median income grew by 2.5% per year. The average inflation rate during that time was 2.4%.
“Our World in Data” shows very strong real median income growth in the United Kingdom between 1996 and 2007. It averaged 3.2% per year. U. K. inflation during that time was 1.87% per year.
There actually has been strong real wage growth without high inflation. This has been tested and proven.
The Phillips curve idea is partly based on the notion that corporations would have to raise prices to pay for higher wages. This is not true. The large corporate tax cut in 2017 was not followed by price cuts (the CPI went up 1.9% in 2018), so why should wage hikes be followed by price increases? Corporations can just reduce their spending on stock buybacks. I don’t think it is a coincidence that the strong wage growth I mentioned earlier happened when stock buybacks were illegal.
The economic theories that wage growth creates inflation, and higher minimum wage costs jobs, and tax cuts create growth, lead to the conclusion that the only viable economy is one in which the peasants do the work and the rich class gets the money. Someone might think that rich people pay economists to come up with these theories. Especially since none of them pan out when you look at the data.
In 1973 the United States had the world’s highest median income. Since then the United States has had one of the worst if not the worst wage growth among first world countries. Several other countries now have higher median incomes including some socialist hellholes like Norway, Sweden, and Denmark. It didn’t have to be like this. The money was there for wages to double, but government policy changes in the 70’s and 80’s diverted the national income from the workers to the rich. One of those policies was to use the Phillip’s curve as a rationale to suppress wages.
"if the relief package turns out to be significantly bigger than needed, the Fed will have to raise rates somewhat to avoid major overheating. And this will be a terrible thing because … why, exactly? True, bond markets might experience something of a shock, something like what happened in 1994 — a shock that burned some investors, but did nothing to derail a very solid real economic recovery"
↑Oh, this is something I would loooove to see!
Do you think $1.9 trillion is big enough, or would you recommend an even larger package (assuming it was politically feasible)?
In response to your question, raising interest rates would be a terrible, terrible thing because, as everyone knows, low interest rates are bad because they cause bubbles...no...wait.....
Gap theories of inflation are not only not useful, but very misleading...
https://marcusnunes.substack.com/p/the-inflation-kerfuffle
The ironic thing is that the people worrying about inflation are some of the same people (not Blanchard or Summers) that were complaining about low interest rates punishing coupon clippers.
A conclusion of the Boskin Commission was that the CPI overstates inflation. Therefore, could one reasonably conclude that inflationary pressures are real?
Besides automatic changes in unemployment insurance as employment picks up, are there elements of the American Rescue Plan that can be re-programmed in real time?