RIP Phillips Curve
David Fields
Economist, State of Utah
Fellow, The Monetary Policy Institute
Louis-Philippe Rochon
Full Professor, Laurentian University
Editor-in-Chief, Review of Political Economy
Monetary Policy Institute Blog #116
“In essence, the Phillips curve is fiction, which for decades has unwaveringly functioned as a master narrative of macroeconomic policy. It’s ad hoc, a unique descriptive statistic from the 1950’s and 1960’s, which does not amount to an historical empirical regularity. It is pseudoscientific, dangerous, and must be put to rest.”
The Phillips Curve is at the core of monetary policy today. Central banks rely on it in an effort to bring inflation back to target. Without it, the whole justification for monetary austerity crumbles.
The Phillips curve rests on the idea that inflation is demand-determined, and that only decreases in aggregate demand, via the labour market and unemployment, can successfully bring the inflation back down. The curve not only plots the variables, but explicitly assumes a causality. Hence, the only possible policy to lower inflation is to raise unemployment and, in the process, lower aggregate demand. This was the precise position of Lawrence Summers, for instance, who called on more unemployment: :We need five years of unemployment above 5% to contain inflation — in other words, we need two years of 7.5% unemployment or five years of 6% unemployment or one year of 10% unemployment.”
It is in this sense that this @monetaryblog has repeatedly argued against austerity, and called it a policy of fighting inflation on the backs of workers.
With the recent discussion over Team Transitory, for which the @monetaryblog was an early advocate, discussion over the fate of the Phillips curve has now come back into fashion — although post-Keynesians and heterodox economists have been arguing for a few decades now that the curve has become unreliable to guide monetary policy because it has considerably flattened. It is in this sense that Rochon has talked about the ‘ineffectiveness of monetary policy’
There is no doubt mainstream models including the Phillips Curve did very poorly in predicting the recent inflation surge, as well as its decline. The reason is quite simple, as Claudio Borio stated, “the response of inflation to a measure of labour market slack has tended to decline and become statistically indistinguishable from zero. In other words, inflation no longer appears to be sufficiently responsive to tightness in labour markets”. Even Janet Yellen as admitted as much: “The slope of the Phillips curve — a measure of the responsiveness of inflation to a decline in labor market slack — has diminished very significantly since the 1960s. In other words, the Phillips curve appears to have become quite flat.” This stands true even in a post-pandemic world.
Granted, these quotes were pre-COVID and some have argued that the Phillips curve has reinstated itself. But there are two reasons why we are skeptical: 1) first, the onus of proof is on those who defend its use today; second, there is just enough data to make such a statement. In other words, it’s too soon to argue this. And let us add a third: the relationship between unemployment and inflation post-COVID shows no sign of a traditional Phillips-type argument. After all, as unemployment has fallen, so has inflation! The divergence between reality and the Phillips Curve model forecasts are stark; by early 2023 the data revealed that CPI de-accelerated substantially as unemployment remained low, empirically validating a glaring pitfall of using the Phillips Curve to guide monetary policy. In July of 2023, in fact, the ignominious Larry Summers was forced to admit that “given how hot the economy is — the inflation performance at this point is better than I think many standard models would have predicted”.
Here, we must say that those who defend it have not been very honest. In 2018, for instance, Jerome Powell decided to ignore all the empirical evidence against the Phillips curve, and simply stated that “the Phillips curve continues to be meaningful for monetary policy”. It wasn’t dead, it was simply “hibernating”, as Fred Mishkin insists. Of course, he would have said that. Admitting anything else would have revealed the weakness of monetary policy.
Recently, Paul Krugman warned us about being warry of economists who refuse to admit they were wrong on inflation. It’s also time they admit once and for all how they were wrong about the Phillips Curve. Let’s admit it once and for all and relegate it to dustbin of history.
Post-Keynesians have argued that interest rates are foremost a distributive variable, and as such, its transmission mechanism works through its effect on income distribution, and hence aggregate demand. This has been a core belief of the Monetary Policy Institute Blog since it was created, back in June 2022. Monetary policy is about income distribution, as in the work of Marc Lavoie, Mario Seccareccia, Louis-Philippe Rochon, Matias Vernengo, John Smithin, Randy Wray, Colin Rogers, and more. Indeed, on January 19, Eckard Hein presented a summary of the post-Keynesian position on this topic, by arguing that inflation is always and everywhere a conflict phenomenon, in which income distribution is central.
Criticism over the relevance of the Phillips is not new, although it is safe to say that heterodox economists have consistently argued against it.
This said, because of its relevance for monetary policy, the profession is not ready to abandon it quite yet. As McLeay and Tenreyro argue, “The Phillips Curve is one of the building blocks of the standard macroeconomic models used for forecasting and policy advice in central banks.”, A sentiment echoed by Bullard (2019): “U.S. monetary policymakers and financial market participants have long relied on the Phillips curve — the correlation between labor market outcomes and inflation — to guide monetary policy.” Without it, as stated above, there is no justification for monetary austerity. No wonder the mainstream is scrambling to reinstate it at all costs, as the implications of its demise are simply too great.
As the Economist recently stated, “the covid-19 pandemic has given economists another chance to learn from their mistakes.” This is one of these instances. Let’s learn how to better model inflation, and it starts by considering the supply side of the economy foremost and to monitor conflict, as is done in post-Keynesian economics, but also in some more mainstream work, as in the work of Olivier Blanchard.
In the end, the Phillips curve is an idealistic model that suggests that low unemployment is the effect of workers forgoing the choice to not work, a sign of a strong labor market, which fuels strong demand and therefore high inflation. Ipso facto, low inflation constitutes workers not working, which means weak demand, and therefore low inflation.
The reductionist takeaway is that the primary policy tool to fight inflation is monetary austerity, to forcefully have workers choose leisure, rather, have no other alternative than to experience unemployment due to a manufactured recession. Inflation, it is assumed, is anywhere and everyday the result of a level of unemployment that is below some perceived natural level, or in other words caused by excess aggregate demand. Supply-side shocks may eventually cause inflation too, but those are seen as of secondary importance. The implication is that absence of imperfections, “rational” individuals and an optimal market structure would produce full utilization of resources at a level that ensures a stable rate of inflation. In the aggregate, if policy, both monetary and fiscal, moves the economy towards full employment, wages can increase excessively, which can ensue runaway inflation. A condition of a non accelerating inflation rate of unemployment (NAIRU) is present to ensure policy maintains a level of employment that prevents the economy from overheating.
This account, however, mystifies an inherent class conflict. The reality is that as the economy is growing rapidly, the reserve army of unemployed is depleted, which gives workers stronger bargaining power to raise wages, gain more control of the workplace, and shift the distribution of total income in their favor, potentially squeezing corporate profit shares. To prevent this, monetary (and fiscal) austerity is encouraged, as prescribed by the so-called NAIRU, in order to facilitate a reduction in investment spending, which leads to a decline in job creation and, thus, higher unemployment, maximizing corporate power over labour power. The Phillips Curve constitutes an insidious conceptual frame.
In essence, the Phillips curve is fiction, which for decades has unwaveringly functioned as a master narrative of macroeconomic policy. It’s ad hoc, a unique descriptive statistic from the 1950’s and 1960’s, which does not amount to an historical empirical regularity. It is pseudoscientific, dangerous, and must be put to rest.
