The disturbing implication of a Phillips curve that turned out to be flat is that the Fed has, for the last 48 years, pursued a policy that has kept millions of people unemployed, kept millions in poverty, and suppressed trillions of dollars in wage growth under the false belief that it would control inflation. This Phillips curve “mistake” has caused an incredible amount of damage to the American working class.
Wage growth essentially stopped after 1973 because of Nixon’s wage and price controls. Wage suppression continued with the “Council on Wage and Price Stability”, and then the Fed policy to keep unemployment high based on the Phillips curve. If wages had continued to grow with economic growth, the median income would be twice as high as it is now. The Rand paper, “Trends in Income From 1975 to 2018”, said “From 1975 to 2018, the difference between the aggregate taxable income for those below the 90th percentile and the equitable growth counterfactual totals $47 trillion.” That was a pretty expensive “mistake”!
America had terrific wage growth between 1933 and 1973. Median income went up by about a factor of 5. Most of that time there was only moderate inflation. For instance, between 1948 and 1973 real median income grew by 2.5% per year. The average inflation rate during that time was 2.4%. But between 1974 and 1981 the average inflation rate was 9.2% while median incomes shrank 1.2% per year. This is the opposite of the “higher wages cause inflation” theory.
The poverty rate went from about 70% in 1933 to 11% in 1973, and then stopped going down. If not for wage suppression poverty could have been virtually eliminated by now.
Shouldn’t the economy exist to benefit the people, instead of sacrificing people for the supposed benefit of the economy? What’s the point? That’s pretty messed up.
The 70’s inflation was not caused by an oil shock. In his 1968 paper Milton Friedman said, “Every major inflation has been produced by monetary expansion.” And Nixon did pressure his Fed chairman into an expansionary monetary policy to goose the economy in order to help him get re-elected (from “How Richard Nixon Pressured Arthur Burns: Evidence from the Nixon Tapes”). Nixon knew this would create inflation but was willing to accept the tradeoff. He said, “I’ve never seen anybody beaten on inflation in the United States. I’ve seen many people beaten on unemployment.”
The 2002 – 2008 oil shock was even bigger than the 70’s oil shock, but it did not produce inflation. That should have driven a stake through the heart of the “oil shock creates inflation” idea.
“Stagflation” wasn’t a real thing. From 1975 to 1979 the average GDP growth rate was 3.68%. That’s a funny kind of stagnation. That growth was better growth than Trump’s “greatest economy in history” and better than the 3.1% average GDP growth during the 80’s. There were 10.345 million jobs created during the Carter administration. The job growth rate of 12.82% was better than any presidential term since then. The labor participation rate was also growing rapidly. The number of manufacturing jobs hit an all-time high in 1979. I have wondered for years why people keep saying there was stagflation in the 70’s. Marcus Nunes made my day with his comment! Stagflation is like the Emperor’s New Clothes, not really there.
Robert Mundell was the economist who identified the economic condition of the late 70’s as “stagflation”. By some coincidence he had a way to fix it, Supply Side Economics. In other words the snake oil salesman diagnosed the disease that he said he could cure. His student Art Laffer cooked up a curve to show that tax cuts create economic growth so that the tax cuts would pay for themselves.
Reagan’s tax cuts did not produce any extra growth and the tax cuts did not pay for themselves (see “The Economic Consequences of Major Tax Cuts for the Rich” by Hope and Limberg). The tax cuts for the rich were paid for with public debt. From 1933 to 1981 high top tax rates effectively imposed a cap on top incomes. Reagan’s tax cuts removed that cap and allowed unlimited incomes. This enabled the top 0.1% to take the $47 trillion that the bottom 90% didn’t get.
In summary, a fake “stagflation” boogeyman was used as a justification for using the bogus Phillips Curve as rationale to suppress wages and the bogus Laffer Curve as rationale to cut top tax rates. That’s how trillions of dollars in income were redistributed from the wage earners to the rich.
In the 70s I was a young exec. leading a small team on public policy issues in the midwest and Rocky Mtn. region. The land was awash with high anxiety over commodity shortages, hyper inflation, the pragmatic value of taking on huger personal debt, and screwball investment schemes. It seemed like emotions drove most decisions, even as speakers rolled through town promoting rational choice economics. It just didn’t pass the smell test, but we were young, junior, and doubted our own suspicions in the face of senior expertise.
It suggests that wage setting conventions play a large role in determining inflation. In the 60s and 70s wage increases to account for inflation would often be baked into employment contracts. Econ admits inflation is hard to estimate; it’s likely also hard for employers and workers to estimate, so they have to rely on conventions to combat wage erosion from a force everyone knows is there but no one knows the size of for sure. In the dark ages of manufacturing, yearly cost increases used to be estimated by taking last year’s cost and adding a set percentage. A similar approach to inflation embedded in wage negotiations could lead inflation dynamics to be unmoored from the usual factors Econ considers.
Excelent post Professor Krugmam. I am Jose Luis Oreiro, professor of economics at University of Brasília. My field of research is on macroeconomics and develepment economics. I had a research group named Structuralist Development Macroeconomics with researchers from US (Peter Skott from UMass), Italy, Spain and Brazil. Peter wrote an interesting paper on the nature of inflation in developing or dual economies in Lewis' s sense. If you want to read it please visit our website www.sdmrg.com.br. With my best regards.
There were at least three very large shocks in the early 1970’s that affected the price level and may gave jointly destabilized inflation expectations: the release of wage and price controls, the Russian Grain Deal and associated food price shock, and the oil price shock. Plus Nixon put great pressure on Arthur Burns and the Fed to remain accommodative in order to not hurt his reelection chances in 1972.
Is "stagflation" the combination of higher than normal inflation with higher than normal unemployment? Did that actually occur in the 1970's, and how do you square that with the fact that total employment rose dramatically for demographic reasons? If there was high unemployment, how much was that a matter of so many more people looking for work? Perhaps the 70's should be heralded as the last great decade of real growth, combined with more jobs created than ever--a success--and not the failure Friedman suggested. The failure began afterwards, largely due to his influence.
I think of myself as a political economist, and not a classically trained mathematical economist. (As a graduate student in Sociology, I did take several economics courses and audited a few others, but it is not the same as the rigor of a sequence of courses. Okay. I've given you my caveat and background).
I remember the 1970s - I was able to put my babysitting money in certificates of deposit that earned 6 to 7% interest. At the time, in high school, I vaguely understood that the reason for higher interest rates were to reduce inflation.
But part of the inflation in the 1970s was due to a unique event in a certain time period and economy. Essentially the Arab oil cartel dramatically increased oil prices and produced the inflationary response. (The rational response would have been for oil companies to actually pay countries a reasonable price for their raw materials and help with education and technological investment, like decent human beings. But I digress).
But in the early 80s, several trends happened: 1) increased innovation, especially in computing and telecommunications,
2) outsourcing of work in the United States, partly because of the first trend. (Of course, there were many other factors, I was just thinking about these two in particular).
Innovation in computing and other technology reduced prices for consumer goods for positive good reasons: production methods and much better computer processing decreased the price of consumer goods, thus decreasing inflationary pressure on prices.
In the early 1980s, another part of the inflationary pressure was wages. In response, United States companies, with the full support of the Reagan administration - remember the Carribean Basin Initiative - outsourced work to lower wage countries.
It is most likely that wages were less inflationary after the early 1980s because companies saved so much money outsourcing work to other countries and paying much lower wages overseas. This also undercut unions, reduced their bargaining power, and resulted in decades of underpayment of United States workers.
As a political economist, I am interested in events in certain times and regions. Mathematical modeling is very useful for many problems. But sometimes it is used to dazzle rather than illuminate.
Paul, if you wanted to talk about the Phillips curve problems you should have gone to the time when it was first “discovered”. Bill Phillips always regretted having written his famous paper and Samuelson and Solow did not estimate it their famous version for the United States but they drew by hand. Several papers show that neither Phillips nor Samuelson and Solow data support the existence of a trade off between wage or price inflation and the unemployment rate. In other words, the Phillips curve never existed in the first place so it is not a surprise that Nakamura and Steinsson found it has been flat which is observationally equivalent to say it does not exist. You seem to have had doubts since the missing inflation episode after the financial crisis and you wrote that explaining inflation was not the core objective of Keynesian macro. As you point out here, even MS require a leap of faith for their results to make sense and that is the case of any other paper that claims to have rescued the Phillips curve. Prominent economists such as Robert Hall and Roger Farmer have already dump it. Friedman had the chance to kill for good in his famous 1968 speech by reexamining Samuelson and Solow data but instead he decided to believe them and give the famous trade off a new life as a short run event. The spirals of inflation and unemployment you show in your chart are also explainable by saying the Phillips curve never existed and this Okham razor type argument also explains all the other episodes of this famous modelas failures.
Sorry, but the price of oil can NOT be blamed for the stagflation of the late 70's/early 80's though it contributed a little bit. The core CPI number (which subtracts out the price of food and oil) was very high. Adding the price of oil didn't add that much to prices. I take Krugmen's point though that we didn't see the kind of economic growth in the 1970's that should have led to the runaway inflation that occurred. It also can't be blamed on government spending. Government spending as a percentage of GDP was not as high in the 70's as it was after the great recession or even now in 2021.
I'm not an economist but it seems to me that after nearly 50 years we OUGHT to be able to come to a consensus regarding what caused the stagflation of this period. It changed the course of American history. Yet I've heard so many conflicting theories and seen no evidence of a "smoking gun" as of yet. The most convincing argument I've heard was that it was the result of the "Nixon shock" and the end of Bretton Woods. The value of the dollar is said to have fallen by a THIRD in the 1970's. I don't know enough about economics though to know if this is enough to be the major causal factor.
The pattern looks like an example of hysteresis due to "memory" or lag in a non-linear system. I'm not that well-versed in the concept, but it seems reasonable that if expectations were at odds with the actual future values (which could be triggered by events such as unexpected "shocks") it could lead to hysteresis.
Inflation is expected to accelerate when aggregate monetary demand exceeds potential supply. The problem with the NAIRU is in assuming that a single labor statistic, the unemployment rate, is a reliable proxy for the true output gap. That's just too simplistic. Potential output is not only limited by the supply of labor but the composition of labor (skills needed vs. skills available) as well as limits on the supply of basic commodities (for example, oil) and the propensity of business to make investments that boost productivity. Indeed, it's possible and even likely that a low unemployment rate leads to increased potential because businesses will increase their investments in productivity boosting technology in order to save on labor costs. The NAIRU idea then is an oversimplification of the complex relationship between aggregate demand and potential supply. It may appear to be workable over in some periods only to fail in others, just as history shows.
Is it possible that the natural rate of unemployment is just significantly lower than 40 years ago. The communications revolution over the past 40 years may have allowed the job market to operate more efficiently and reduce the natural rate of employment. It is certainly easier to get in contact in contact with applicants. No more phone tag. No more running around town looking for job applications.
The disturbing implication of a Phillips curve that turned out to be flat is that the Fed has, for the last 48 years, pursued a policy that has kept millions of people unemployed, kept millions in poverty, and suppressed trillions of dollars in wage growth under the false belief that it would control inflation. This Phillips curve “mistake” has caused an incredible amount of damage to the American working class.
Wage growth essentially stopped after 1973 because of Nixon’s wage and price controls. Wage suppression continued with the “Council on Wage and Price Stability”, and then the Fed policy to keep unemployment high based on the Phillips curve. If wages had continued to grow with economic growth, the median income would be twice as high as it is now. The Rand paper, “Trends in Income From 1975 to 2018”, said “From 1975 to 2018, the difference between the aggregate taxable income for those below the 90th percentile and the equitable growth counterfactual totals $47 trillion.” That was a pretty expensive “mistake”!
America had terrific wage growth between 1933 and 1973. Median income went up by about a factor of 5. Most of that time there was only moderate inflation. For instance, between 1948 and 1973 real median income grew by 2.5% per year. The average inflation rate during that time was 2.4%. But between 1974 and 1981 the average inflation rate was 9.2% while median incomes shrank 1.2% per year. This is the opposite of the “higher wages cause inflation” theory.
The poverty rate went from about 70% in 1933 to 11% in 1973, and then stopped going down. If not for wage suppression poverty could have been virtually eliminated by now.
Shouldn’t the economy exist to benefit the people, instead of sacrificing people for the supposed benefit of the economy? What’s the point? That’s pretty messed up.
The 70’s inflation was not caused by an oil shock. In his 1968 paper Milton Friedman said, “Every major inflation has been produced by monetary expansion.” And Nixon did pressure his Fed chairman into an expansionary monetary policy to goose the economy in order to help him get re-elected (from “How Richard Nixon Pressured Arthur Burns: Evidence from the Nixon Tapes”). Nixon knew this would create inflation but was willing to accept the tradeoff. He said, “I’ve never seen anybody beaten on inflation in the United States. I’ve seen many people beaten on unemployment.”
The 2002 – 2008 oil shock was even bigger than the 70’s oil shock, but it did not produce inflation. That should have driven a stake through the heart of the “oil shock creates inflation” idea.
“Stagflation” wasn’t a real thing. From 1975 to 1979 the average GDP growth rate was 3.68%. That’s a funny kind of stagnation. That growth was better growth than Trump’s “greatest economy in history” and better than the 3.1% average GDP growth during the 80’s. There were 10.345 million jobs created during the Carter administration. The job growth rate of 12.82% was better than any presidential term since then. The labor participation rate was also growing rapidly. The number of manufacturing jobs hit an all-time high in 1979. I have wondered for years why people keep saying there was stagflation in the 70’s. Marcus Nunes made my day with his comment! Stagflation is like the Emperor’s New Clothes, not really there.
Robert Mundell was the economist who identified the economic condition of the late 70’s as “stagflation”. By some coincidence he had a way to fix it, Supply Side Economics. In other words the snake oil salesman diagnosed the disease that he said he could cure. His student Art Laffer cooked up a curve to show that tax cuts create economic growth so that the tax cuts would pay for themselves.
Reagan’s tax cuts did not produce any extra growth and the tax cuts did not pay for themselves (see “The Economic Consequences of Major Tax Cuts for the Rich” by Hope and Limberg). The tax cuts for the rich were paid for with public debt. From 1933 to 1981 high top tax rates effectively imposed a cap on top incomes. Reagan’s tax cuts removed that cap and allowed unlimited incomes. This enabled the top 0.1% to take the $47 trillion that the bottom 90% didn’t get.
In summary, a fake “stagflation” boogeyman was used as a justification for using the bogus Phillips Curve as rationale to suppress wages and the bogus Laffer Curve as rationale to cut top tax rates. That’s how trillions of dollars in income were redistributed from the wage earners to the rich.
In the 70s I was a young exec. leading a small team on public policy issues in the midwest and Rocky Mtn. region. The land was awash with high anxiety over commodity shortages, hyper inflation, the pragmatic value of taking on huger personal debt, and screwball investment schemes. It seemed like emotions drove most decisions, even as speakers rolled through town promoting rational choice economics. It just didn’t pass the smell test, but we were young, junior, and doubted our own suspicions in the face of senior expertise.
It suggests that wage setting conventions play a large role in determining inflation. In the 60s and 70s wage increases to account for inflation would often be baked into employment contracts. Econ admits inflation is hard to estimate; it’s likely also hard for employers and workers to estimate, so they have to rely on conventions to combat wage erosion from a force everyone knows is there but no one knows the size of for sure. In the dark ages of manufacturing, yearly cost increases used to be estimated by taking last year’s cost and adding a set percentage. A similar approach to inflation embedded in wage negotiations could lead inflation dynamics to be unmoored from the usual factors Econ considers.
Excelent post Professor Krugmam. I am Jose Luis Oreiro, professor of economics at University of Brasília. My field of research is on macroeconomics and develepment economics. I had a research group named Structuralist Development Macroeconomics with researchers from US (Peter Skott from UMass), Italy, Spain and Brazil. Peter wrote an interesting paper on the nature of inflation in developing or dual economies in Lewis' s sense. If you want to read it please visit our website www.sdmrg.com.br. With my best regards.
There were at least three very large shocks in the early 1970’s that affected the price level and may gave jointly destabilized inflation expectations: the release of wage and price controls, the Russian Grain Deal and associated food price shock, and the oil price shock. Plus Nixon put great pressure on Arthur Burns and the Fed to remain accommodative in order to not hurt his reelection chances in 1972.
"How do you ask someone to be the last worker to be unemployed for a mistake?"
Your paraphrasing game is reaching new heights! The original is arguably John Kerry's greatest contribution to American history.
I've missed your NYT blog, glad to see this one! Twitter has its place but I think blogging lends itself better to more careful analysis and writing.
Is "stagflation" the combination of higher than normal inflation with higher than normal unemployment? Did that actually occur in the 1970's, and how do you square that with the fact that total employment rose dramatically for demographic reasons? If there was high unemployment, how much was that a matter of so many more people looking for work? Perhaps the 70's should be heralded as the last great decade of real growth, combined with more jobs created than ever--a success--and not the failure Friedman suggested. The failure began afterwards, largely due to his influence.
I think of myself as a political economist, and not a classically trained mathematical economist. (As a graduate student in Sociology, I did take several economics courses and audited a few others, but it is not the same as the rigor of a sequence of courses. Okay. I've given you my caveat and background).
I remember the 1970s - I was able to put my babysitting money in certificates of deposit that earned 6 to 7% interest. At the time, in high school, I vaguely understood that the reason for higher interest rates were to reduce inflation.
But part of the inflation in the 1970s was due to a unique event in a certain time period and economy. Essentially the Arab oil cartel dramatically increased oil prices and produced the inflationary response. (The rational response would have been for oil companies to actually pay countries a reasonable price for their raw materials and help with education and technological investment, like decent human beings. But I digress).
But in the early 80s, several trends happened: 1) increased innovation, especially in computing and telecommunications,
2) outsourcing of work in the United States, partly because of the first trend. (Of course, there were many other factors, I was just thinking about these two in particular).
Innovation in computing and other technology reduced prices for consumer goods for positive good reasons: production methods and much better computer processing decreased the price of consumer goods, thus decreasing inflationary pressure on prices.
In the early 1980s, another part of the inflationary pressure was wages. In response, United States companies, with the full support of the Reagan administration - remember the Carribean Basin Initiative - outsourced work to lower wage countries.
It is most likely that wages were less inflationary after the early 1980s because companies saved so much money outsourcing work to other countries and paying much lower wages overseas. This also undercut unions, reduced their bargaining power, and resulted in decades of underpayment of United States workers.
As a political economist, I am interested in events in certain times and regions. Mathematical modeling is very useful for many problems. But sometimes it is used to dazzle rather than illuminate.
I get the impression that at least over in the MMT-influenced chunk of the economics discourse I inhabit the narrative you are moving towards here has been the clear consensus for quite a long time: https://neweconomicperspectives.org/2013/04/and-the-last-shall-be-first-it-was-the-peanut-farmer-not-the-tall-guy-or-the-iron-lady.html
Paul, if you wanted to talk about the Phillips curve problems you should have gone to the time when it was first “discovered”. Bill Phillips always regretted having written his famous paper and Samuelson and Solow did not estimate it their famous version for the United States but they drew by hand. Several papers show that neither Phillips nor Samuelson and Solow data support the existence of a trade off between wage or price inflation and the unemployment rate. In other words, the Phillips curve never existed in the first place so it is not a surprise that Nakamura and Steinsson found it has been flat which is observationally equivalent to say it does not exist. You seem to have had doubts since the missing inflation episode after the financial crisis and you wrote that explaining inflation was not the core objective of Keynesian macro. As you point out here, even MS require a leap of faith for their results to make sense and that is the case of any other paper that claims to have rescued the Phillips curve. Prominent economists such as Robert Hall and Roger Farmer have already dump it. Friedman had the chance to kill for good in his famous 1968 speech by reexamining Samuelson and Solow data but instead he decided to believe them and give the famous trade off a new life as a short run event. The spirals of inflation and unemployment you show in your chart are also explainable by saying the Phillips curve never existed and this Okham razor type argument also explains all the other episodes of this famous modelas failures.
Sorry, but the price of oil can NOT be blamed for the stagflation of the late 70's/early 80's though it contributed a little bit. The core CPI number (which subtracts out the price of food and oil) was very high. Adding the price of oil didn't add that much to prices. I take Krugmen's point though that we didn't see the kind of economic growth in the 1970's that should have led to the runaway inflation that occurred. It also can't be blamed on government spending. Government spending as a percentage of GDP was not as high in the 70's as it was after the great recession or even now in 2021.
I'm not an economist but it seems to me that after nearly 50 years we OUGHT to be able to come to a consensus regarding what caused the stagflation of this period. It changed the course of American history. Yet I've heard so many conflicting theories and seen no evidence of a "smoking gun" as of yet. The most convincing argument I've heard was that it was the result of the "Nixon shock" and the end of Bretton Woods. The value of the dollar is said to have fallen by a THIRD in the 1970's. I don't know enough about economics though to know if this is enough to be the major causal factor.
Not as wonkish as feared. Just keep the equations to a minimum and I’ll be fine. Thank you.
The pattern looks like an example of hysteresis due to "memory" or lag in a non-linear system. I'm not that well-versed in the concept, but it seems reasonable that if expectations were at odds with the actual future values (which could be triggered by events such as unexpected "shocks") it could lead to hysteresis.
Inflation is expected to accelerate when aggregate monetary demand exceeds potential supply. The problem with the NAIRU is in assuming that a single labor statistic, the unemployment rate, is a reliable proxy for the true output gap. That's just too simplistic. Potential output is not only limited by the supply of labor but the composition of labor (skills needed vs. skills available) as well as limits on the supply of basic commodities (for example, oil) and the propensity of business to make investments that boost productivity. Indeed, it's possible and even likely that a low unemployment rate leads to increased potential because businesses will increase their investments in productivity boosting technology in order to save on labor costs. The NAIRU idea then is an oversimplification of the complex relationship between aggregate demand and potential supply. It may appear to be workable over in some periods only to fail in others, just as history shows.
Is it possible that the natural rate of unemployment is just significantly lower than 40 years ago. The communications revolution over the past 40 years may have allowed the job market to operate more efficiently and reduce the natural rate of employment. It is certainly easier to get in contact in contact with applicants. No more phone tag. No more running around town looking for job applications.