The connection between inflation and actual financial exercise has lengthy been central to debates in macroeconomics and financial coverage. On the core of this debate is the Phillips curve (PC), which measures how strongly inflation reacts to actions in financial situations. The steepness of this curve issues enormously for financial coverage: if the PC is steeper, inflation rises sooner throughout booms and falls sooner in recessions, which entails central banks having to behave extra forcefully in the event that they need to stabilize inflation round their goal. Prior evaluation discovered astonishingly small estimates of the slope of the PC, which means that the curve is “flat” (and even lifeless). On this put up, I current proof from coauthored analysis exhibiting that, opposite to the standard view, the Phillips curve is alive and steep, and it captures inflation volatility remarkably effectively as soon as actual marginal price is used as an alternative of normal actual financial exercise measures.
The Standard Formulation of the Phillips Curve
The Phillips curve hyperlinks inflation to expectations of future inflation and a measure of financial slack. Within the standard formulation of the PC, financial slack is often proxied by the output or unemployment gaps—the deviation of output or employment from its pure degree:
On this view, inflation rises when the financial system overheats and falls throughout slowdowns. The slope of the Phillips curve, Okay within the equation above, captures how delicate inflation is to those fluctuations. A big physique of analysis finds that the PC is kind of flat—the slope may be very small—implying that inflation hardly strikes in response to shifts in output or employment gaps. These findings have lengthy puzzled economists and fueled debate about how energetic financial coverage must be to steer inflation.
The Primitive (Price-Primarily based) Formulation of the Phillips Curve
There may be one other approach to consider inflation dynamics and its drivers. At its basis, the Phillips curve emerges from the aggregation of companies’ pricing selections in response to adjustments in manufacturing prices. Within the primitive formulation of the PC, the variable influencing inflation is actual marginal price in proportion deviation from pattern, somewhat than the output or unemployment hole:
On this view, inflation rises when financial forces improve companies’ manufacturing prices. The slope of the curve displays how strongly (and rapidly) these prices are handed by means of into output costs.
The Slope of the Price-Primarily based Phillips Curve
The associated fee-based formulation makes it simpler to know the important thing forces that decide the slope of the PC. In concept, if markets have been completely aggressive and frictionless, costs would transfer one-for-one with prices: any improve in wages or enter costs could be immediately mirrored in shopper costs. That’s, the slope could be one.
Actuality may be very completely different. Companies sometimes regulate costs occasionally, since doing so includes each direct prices (akin to relabeling or updating programs) and oblique prices (akin to complicated clients or shedding goodwill). As well as, companies typically set costs strategically, selecting to delay adjustments till they’re certain price pressures will final, or timing revisions to match opponents. Thus, frictions that result in rare value adjustments and strategic issues in value setting weaken the transmission of price shocks into costs. The extra companies deviate from the best of versatile, aggressive markets, the flatter the Phillips curve turns into.
Estimating the Slope of the Price-Primarily based Phillips Curve Utilizing Microdata
How steep is the cost-based PC within the information? Answering this query is notoriously troublesome, notably when the estimation is solely primarily based on mixture information through which many shocks affect inflation and actual exercise on the similar time. To deal with this drawback, we flip to firm-level proof. In a latest paper coauthored with Mark Gertler (New York College), Luca Gagliardone (Yale College), and Joris Tielens (Nationwide Financial institution of Belgium), we use detailed microdata on costs and prices to check how particular person companies regulate their costs in response to adjustments in manufacturing prices. This method permits us to quantify how nominal rigidities (rare value adjustments) and actual rigidities (strategic interactions amongst companies) dampen the response of costs to price shocks. With these estimates we get well the slope of the primitive type of the Phillips curve.
Our evaluation suggests a powerful hyperlink between inflation and actual financial situations, as captured by producers’ prices. On common, companies maintain costs mounted for 3 to 4 quarters, confirming a considerable diploma of nominal rigidity. We additionally discover sturdy proof of strategic complementarities: companies regulate much less aggressively as a result of they like to maneuver in keeping with opponents, which cuts the pass-through of price shocks roughly in half. Taking these frictions into consideration, we estimate the slope of the cost-based Phillips curve to be three to 10 instances bigger than the estimates for the output- or unemployment-based PC formulation. This suggests that the Phillips curve is steep, not flat, even in regular instances.
Accounting for Combination Inflation Dynamics
Price pass-through performs a dominant position in shaping mixture inflation. As an example this, we construct a price index by combining firm-level adjustments in labor and enter prices after which feed it into the cost-based Phillips curve. Utilizing the Belgian manufacturing sector as a case examine, we assemble a model-generated inflation collection by feeding information on prices into the Phillips curve. The outcomes, proven within the chart beneath, present that the anticipated inflation aligns very carefully with precise producer value inflation (PPI) in Belgium. The 2 collection are extremely correlated, with a correlation coefficient above 80 p.c; quantitatively, actions in manufacturing prices alone account for about 70 p.c of noticed inflation fluctuations, highlighting the central position of prices in driving inflation dynamics.
By way of the Lens of the Price-Primarily based Phillips Curve, Fluctuations in Manufacturing Prices Account Effectively for Inflation Volatility
Notes: The blue line represents the time collection of producing PPI in Belgium. The purple line is the model-implied manufacturing PPI obtained feeding an mixture price index to a cost-based PC.
Reconciling the Steep Price-Primarily based Phillips Curve with the Flat Output-Primarily based Phillips Curve
These outcomes increase a pure query: Why does the cost-based Phillips curve slope steeply whereas the output-based one seems flat? And are these findings at odds with earlier analysis?
Our analysis reveals that the “flatness” of the standard Phillips curve displays a weak hyperlink between output gaps and marginal prices. In different phrases, whereas manufacturing prices feed instantly into companies’ pricing selections, actions in output (or unemployment) bear solely a unfastened relationship to inflation.
From a conceptual standpoint, the standard PC holds provided that marginal price and the output hole transfer proportionally—an assumption that requires, amongst different issues, completely versatile wages. When these situations fail, the output hole could also be a poor proxy for actual marginal price, biasing estimates of PC downward. Furthermore, even when proportionality roughly holds, the output-based slope equals the cost-based slope scaled by the elasticity of marginal price with respect to output. If this elasticity is low, the slope of output-based PC shall be low as effectively, even when the slope of the cost-based PC is sizable.
These outcomes are confirmed within the information. Specializing in the pre-pandemic interval (1999–2019), we estimate a really low elasticity of marginal price with respect to output. This discovering helps clarify why the cost-based PC is steep whereas the standard PC seems flat.
Classes for the Publish-Pandemic Inflation Surge
Our findings present a powerful pass-through from marginal prices to costs, which explains why the cost-based Phillips curve matches inflation dynamics so effectively. The weak hyperlink between output and marginal price, then again, helps clarify why the standard output-gap model of the PC seems “flat.” In regular instances, two components drive this low elasticity: companies’ price schedules are inclined to show practically fixed short-run returns to scale, so marginal prices barely transfer with output; and wage rigidity additional dampens any suggestions from demand to prices.
The pandemic and its aftermath revealed how rapidly these relationships can change underneath stress. Extreme shocks—whether or not from labor market tightness or provide chain bottlenecks—pushed companies in opposition to capability limits, sending marginal prices sharply increased and fueling inflation. On the similar time, the slope of the Phillips curve itself can shift. Pre-pandemic information confirmed secure adjustment frequencies and an approximate linear relationship between inflation and (proportion) adjustments in actual marginal prices. Extra not too long ago, nevertheless, companies have been adjusting costs rather more typically, elevating the elasticity of inflation with respect to prices and producing nonlinear inflation dynamics. I’ll speak about nonlinear inflation dynamics—what it means, the way it works, and what it implies—in a companion put up on Liberty Avenue Economics.
Simone Lenzu is a monetary analysis economist within the Federal Reserve Financial institution of New York’s Analysis and Statistics Group.
The right way to cite this put up:
Simone Lenzu, “Anatomy (not Post-mortem) of the Phillips Curve,” Federal Reserve Financial institution of New York Liberty Avenue Economics, February 4, 2026, https://doi.org/10.59576/lse.20260204
BibTeX: View |
Disclaimer
The views expressed on this put up are these of the writer(s) and don’t essentially mirror the place of the Federal Reserve Financial institution of New York or the Federal Reserve System. Any errors or omissions are the accountability of the writer(s).
