The Original Phillips Curve
If you have studied economics at some point, you have likely come across the Phillips Curve, a simple graph that purports to show the relationship between unemployment and inflation in an economy.
The Phillips curve is named after New Zealand economist A.W. Phillips who examined the unemployment levels and average wages in the U.K. between 1861 and 1957 and published his findings in a paper in 1958. It showed, on average, an inverse relationship between the level of unemployment and the rate of change in wages.
The intuition is simple: a falling unemployment rate means that the demand for labour is rising, and as labour supply gets more scarce, it pushes up wages faster, causing wage inflation; on the other hand, a rising unemployment rate signals a falling demand for labour, which pushes down wages.
On a simple diagram, with the unemployment rates given on the horizontal axis and the inflation rates given on the vertical line, the relationships would be represented by a curve negatively sloped from left to right. The curved nature of the relationship shows the differing elasticities at different levels of unemployment and inflation. Wages grow faster at lower unemployment rates and vice versa. Economists then extended the wage inflation to price inflation, since the two are closely associated — changes in wages are reflected as changes in prices in the same direction. Additionally, fiscal policies that target unemployment affect the demand for products directly which is also reflected in their prices.
This relationship showed a trade-off between the two variables, which had serious policy implications. An economy cannot have low inflation and low unemployment at the same time — any attempts to lower unemployment would push up wages and prices. The Phillips curve and the associated elasticities were calculated for developed economies and economists were able to use these to formulate optimal targets and policies. This was how the inflation-unemployment relationship looked like in the US in the 1960s:
The 1960s saw the Phillips Curve grow in popularity, as both economists and the government of developed economies applied its reasoning in their economic policymaking. If the economy was at a high unemployment rate, they could afford to stimulate the economy and raise prices a little to lower unemployment, but an economy already at a very low unemployment rate would have to face very high inflation if they attempted to reduce unemployment further.
Modifications and Criticisms
However, there were several modifications and criticisms made of the Phillips curve, both theoretical and empirical, in the subsequent years and decades, to the point where its very relevance and significance have come under question.
In the late 1960s, economists Milton Friedman and Edmund Phelps employed the role of expectations to modify the Phillips curve conclusions. Friedman argues that employees are subject to a temporary “money illusion”, in that, when prices rise, workers do not immediately realise their real wages have fallen and so will be ready to work at the same nominal wage rates, allowing firms to hire more workers since they would get higher profits for the same given nominal wages. But eventually, as workers realise that their real wages have fallen and demand higher pay, firms will fire some workers, causing the unemployment rate to be back at the original rate. Friedman named this unemployment rate that is maintained in the long run as the “natural rate of unemployment”. Any attempts to bring unemployment below this natural rate would result in only a short-term reduction in unemployment because of the “money illusion”. But once they insist on higher wages, the demand for labour will fall and the unemployment rate will be back at the natural rate. But the inflation continues, putting the economy in a worse position than it started at. Thus, unemployment below the natural rate would generate accelerating inflation, and unemployment above it would cause accelerating deflation. This logic was employed to argue that governments should not intervene in the economy to drive unemployment below the natural rate. It also lent credibility to Friedman’s famous proposition that inflation was a monetary phenomenon, not caused by real factors.
Edmund Phelps and the new classical economists, who believed in the rational expectations concept, while accepting the natural rate of unemployment and the vertical long-run Phillips curve, rejected the possibility of money illusion and contented that rational workers would only react to real wages, and not nominal wages.
The natural rate theory was proved in the 1970s when inflation and unemployment rose over the same period, resulting in what is known as ‘stagflation.’ Here is how the inflation-unemployment relationship looked like at various points in time between the 1970s and 2019 in the US:
As you can see, the neat negative slope of the original Phillips curve has broken down. Instead, in the long run, unemployment seems to have averaged at a certain rate — around 5.5% — 6%.
And so it turned out that the trade-off interpreted from the original Phillips curve was temporary; it was a short-run phenomenon. In the long run, the curve transformed into a vertical line at the natural unemployment rate. This illustration is called the expectations-augmented Phillips Curve:
New classical economists have come to accept the existence of a Non-Accelerating Inflation Rate of Unemployment, or a NAIRU, a rate of unemployment that, if maintained, would keep the inflation rate stable, that is, keep the rate of change in inflation zero (it was not called the natural rate because economists recognised that there is no socially optimal or unchanging rate of unemployment). The NAIRU can be thought of as the empirical counterpart of the theoretical natural rate of unemployment. In the US the NAIRU was calculated to be 6% on average between 1976 and 2002 and such a rate has been assumed to be relatively constant over time.
The New Keynesians’ Phillips Curve, in whose models prices are sticky in the short-run, there is a negative relation between the rate of inflation and the rate of unemployment (derived from the positive relationship between inflation rate and demand level). This implies that an increase in inflation can lower unemployment temporarily, but not permanently, similar to Friedman’s hypothesis. However, they reject the thesis that the unemployment rate always returns to the NAIRU. Instead, they opine that as the unemployment rate rises and persists over a long period, the NAIRU also rises. Therefore, the NAIRU is dependent on the actual unemployment rate — this is called the hysteresis hypothesis (this is attributed to the complicated impacts of unionisation — for more, see suggested readings)
However, such models have also been found lacking as they could not account for the fact that low levels of inflation have been experienced alongside a wide range of unemployment levels — if a unique NAIRU did exist, inflation should have decelerated significantly when unemployment was above that rate and accelerated when unemployment was below that rate. Estimates of the NAIRU have been very imprecise and difficult to pin down. Additionally, many of Friedman’s assumptions employed in his models, such as the full flexibility of price and wages, have since also come under heavy criticism. The rational expectations hypothesis is heavily contested, to boot.
At the turn of the millennium, economists including George Akerlof suggested a moderate trade-off between low levels of inflation and unemployment — if inflation is reduced from two to zero per cent, unemployment will be permanently increased by 1.5 per cent. This is because workers generally have a higher tolerance for real wage cuts than nominal ones (a worker will more likely accept a wage increase of two per cent when inflation is three per cent, than a wage cut of one per cent when the inflation rate is zero).
Heterodox critics point out that imperfect markets make for uneven bargaining power for employees vis-a-vis employers which hinders the possibility for them to act on their perceptions or expectations. One could also reason, as many of the earliest classical political economists have in the past, that it is the lower unemployment which allows workers to amass bargaining power and push for higher wages, which in turn pushes firms to raise prices.
The aftermath of the Great Recession of 2008 saw inflation in the advanced economies deviate from the conventional Phillips curve descriptions — actual inflation during 2009–2010 was higher than expected in a phenomenon described as “missing deflation.” On the other hand, the unemployment rate fell below 5 per cent in the US by 2016, close to its natural rate, but inflation turned out to be dormant, confounding expectations of inflationary pressures. The latter could very well be chalked up to the inflation-targeting that was initiated by the US Fed in 2012. Economists desire some level of inflation to be present to keep the economy growing, and so aim to achieve a certain inflation rate over a period. Therefore, even as the US unemployment rates fell consistently since 2010 (until the pandemic), inflation has remained low, rendering the Phillips curve more or less flat.
The “missing deflation” episode and flattening of the Phillips curve have been given many explanations — that well-anchored inflation expectations have made the relationship static; that the short-term unemployment rate, which has declined to a much lesser extent, matters more for inflation than the overall unemployment rate; that the inflation expectations of households, which are more volatile and inflated, are what matter and not the relatively stable expectations of financial markets; that the Phillips curve might be convex, in that the response of inflation to demand conditions might be quite tempered during recessions as opposed to its response during expansions; and that not only domestic but global factors need to be factored in appropriately.
A different version of the Phillips curve which measures the relationship between the inflation rate and the output gap has been employed to explain these new peculiarities. The output gap is the difference between actual output and potential output. A low unemployment rate signals a growing economy where potential output is also rising, which abates the pressure on prices and prevents inflation from rising too fast.
In fact, it has been found that inflation fluctuations are now influenced less by expectations and more by changes in the output gap. So while there might be no defining relationship between unemployment and inflation anymore, the modern Phillips curve indicates that the relationship between inflation and economic activity persists, which is moderated by inflation-targeting policies. Though this is quite different from the original Phillips curve, some have explained this as “the return of the original Phillips curve”.
Furthermore, what’s to say that it is indeed high employment levels causing inflation? Inflation could be the result of exogenous shocks or a rise in other business costs. It could also very well be the case that firms are unable to raise prices in very competitive markets despite a rise in demand or wages, in which case there would be no relation between inflation and unemployment. While these may seem to be secondary considerations to the overarching general framework of the Phillips curve, these are relevant and significant to a world economy as dynamic and imperfect as today’s.
As we’ve seen, the interpretations derived from the Phillips curve influence economic policy. The original relationship has broken down and new monetary policy frameworks have even made it almost irrelevant in the eyes of many economists. These debates on the Phillips curve hinge on expectations and the behaviour of economic agents and will continue as the research on the relationship between economic activity, expectations and inflation rage on.
The nature of changes and situating them in a context is however crucial when choosing to act on policies. The recent flattening of the curve has emboldened reserve banks in advanced economies to confidently pursue inflation targeting. But such a policy can also be dangerous if inflation pressures are caused by exogenous shocks such as the one caused by the ongoing Russian invasion of Ukraine.
It should also be noted these theories were formulated in the context of advanced developed economies and are not necessarily universal. It has been suggested that the Phillips curve is reversed in poor countries where agriculture is the dominant sector and that there is no trade-off between inflation and unemployment rates in many Middle Eastern and North African countries. (In India, an RBI paper found the presence of a Phillips Curve with respect to consumer price inflation over the high-inflation period of 2007–2016).
What seems to be evident is that there are no simple explanations for the associations between macroeconomic variables. Economies are in constant flux and as they change structurally, the extent and nature of these relationships also change. This also makes the questions on these relationships (or lack thereof) more important because the prevailing consensus on them directly affects the masses through their influence on economic policy. Modified versions of the Phillips curve are still used by central banks for forecasting inflation and understandings its macro-impacts. Inflation and employment are two of the most crucial economic variables that directly concern the population, especially the poor, working class and marginalised communities of various societies, and so it is significant for us to understand how they interact.
To understand the theoretical underpinning of mainstream theories:
- Phillips, A.W. “The Relation Between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861–1957.” Economica, November 1958, Vol. 25, Issue 100, pp. 283–99.
- Samuelson, Paul A., and Robert M. Solow. “Analytical Aspects of Anti-inflation Policy.” American Economic Review 50, no. 2 (1960): 177–194.
- Phelps, Edmund S. “Phillips Curves, Expectations of Inflation and Optimal Employment over Time.” Economica, n.s., 34, no. 3 (1967): 254–281.
- Friedman, Milton. “The Role of Monetary Policy.” American Economic Review 58, no. 1 (1968): 1–17.
- Friedman, Milton. Nobel Lecture: “Inflation and Unemployment”. Journal of Political Economy 85.3 (1977), 451–472.
- Galí, Jordi. (2015). “Monetary Policy, Inflation, and the Business Cycle: An Introduction to the New Keynesian Framework and Its Applications”, second edition, Princeton University Press.
- Laurence Ball, Joern Onken. (2021), “Hysteresis in unemployment: evidence from OECD estimates of the natural rate”. European Central Bank Working Paper Series No 2625.
- Akerlof, G. A., Dickens, W. T., Perry, G. L., Bewley, T. F., & Blinder, A. S. (2000). “Near-Rational Wage and Price Setting and the Long-Run Phillips Curve”. Brookings Papers on Economic Activity, 2000(1), 1–60.
- James Forder. (2014). Macroeconomics and the Phillips Curve Myth. Oxford University Press.
- Gordon, Robert J. (2011). “The History of the Phillips Curve: Consensus and Bifurcation”. Economica. 78 (309): 10–50. doi:10.1111/j.1468–0335.2009.00815.x. S2CID 759217
- Herbener, Jeffrey M. (1992). “The Fallacy of the Phillips Curve”. Dissent on Keynes: A Critical Appraisal of Keynesian Economics. New York: Praeger. pp. 51–71. ISBN 978–0–275–93778–2.
- Ramaa Vasudevan. (2006). “What’s the relationship between inflation and unemployment?” Dollars & Sense
- Charles Oliver. (1999). Phillips Curve. Ludwig von Mises Institute.
For more heterodox resources, check out Rethinking Economics India’s Pluralist Resources Database. The project aims to build a public repository of resources to support students and scholars in learning pluralist economics. The database is constantly updated by our team by collating resources of various forms from several sources and organised across 21 categories.
This blog is authored by Lintha Saleem. Lintha is a volunteer with the Project Pluralist Resources at the Rethinking Economics India Network and looks at the curation and dissemination endeavours. Lintha is currently studying for her Masters at the Centre for Economic Studies and Planning at the Jawaharlal Nehru University.
- Kevin D. Hoover. Phillips Curve. Econlib
- Kristie M. Engemann. (2020). What Is the Phillips Curve (and Why Has It Flattened)? Federal Reserve Bank of St. Louis
- Peter Lihn Jørgensen and Kevin J. Lansing. (2020). Return of the Original Phillips Curve. Federal Reserve Bank of San Francisco
- Filippo Occhino. (2019). The Flattening of the Phillips Curve: Policy Implications Depend on the Cause. Federal Reserve Bank of Cleveland
- The Economist. (2017). The Phillips curve may be broken for good
- Akerlof, G. A., Dickens, W. T., Perry, G. L., Bewley, T. F., & Blinder, A. S. (2000). Near-Rational Wage and Price Setting and the Long-Run Phillips Curve. Brookings Papers on Economic Activity, 2000(1), 1–60.
- Phillips Curve. (2022). Wikipedia.
- Michael Bleaney, Manuela Francisco. (2018). Is The Phillips Curve Different in Poor Countries. Bulletin of Economic Research
- Muhammad Azam, Rasheed Khan, Saleem Khan. (2022). Does the Phillips Curve Exist? Evidence from the Middle East and North African Countries. Journal of Central Banking Theory and Practice, 2022, 3, pp. 59–78
- Harendra Behera, Garima Wahi, Muneesh Kapur. (2017). RBI Working Paper Series №08: Phillips Curve Relationship in India: Evidence from State-Level Analysis.