Profit-Led Inflation: Trivial or Wrong?

The economist Christopher Conlon has now offered a more formal treatment in a forthcoming paper in the International Journal of Industrial Organization. He comes to the same conclusions.

First, a reminder. What is the theory of “profit-led inflation” for the recent inflation surge? Per Conlon:

One problem with this is as a theory is that the outcomes it predicts for prices and profits are really no different from the alternative story that excessive monetary and fiscal stimulus drove up demand levels across the economy, pushing higher spending up against supply constraints.

In this textbook theory of excess money meeting a constrained supply, output rises, but so do prices and profits (temporarily). Why? Because when nominal spending surges into a world of limited short-run supply, buyers become less price sensitive. Firms don’t need a new “permission slip” to raise prices. They just face a demand curve that has shifted up, and they move prices up to match.

So if “profit-led inflation” is just a colourful way of saying “in a hot economy with surging spending and constrained supply, firms with some market power will raise prices and margins,” then: yes, that’s true. But trivially so.

Yet that is not the spirit of the narrative that profit-led inflation theorists were pushing. Their stronger claim implied that cost increases themselves changed industries’ competitive conduct—i.e. that it created conditions that moved firms from more competitive to more cooperative equilibria—so prices and profits could rise without a corresponding increase in demand. Cost increases, in other words, greased the wheels for tacit or implicit collusion to raise prices aggressively.

Indeed, the theory’s proponents made clear that they were making a causal claim about inflation, not a descriptive one. This is most obvious in policy terms, where profit-led inflation advocates typically downplayed or opposed the need for monetary tightening to choke off inflation. Instead, they recommended going after firm profitability directly via “expanded antitrust enforcement, laws preventing, ‘price gouging,’ a tax on ‘excess’ or ‘windfall’ profits, and price caps.” They saw inflation, in other words, as being borne of undesirable microeconomic firm decisions, not macroeconomic policies.

Conlon’s Theoretical Insights

This is where Conlon’s industrial organization economics is useful, because it forces you to say what would be different in the data if the profit-led inflation story were true.

If we set aside firms setting prices expecting future changes in conditions, Conlon uses microeconomic theory to show that there are essentially three reasons a profit-maximizing firm raises prices:

(1) demand increases—consumers become less price sensitive;

(2) marginal costs increase—input costs rise or supply constraints bite;

(3) conduct or the nature of competition changes, at least temporarily.

But these different reasons have different implications for price, output and industry profits, as his Table 1 shows.

Take the cost shock thesis. Greedflationists say that costs rose; firms raised prices more than costs; profits rose; and this is supposed to be evidence of “cost-push profiteering.” But under standard models, a pure industry-wide cost shock does not raise industry profits. Indeed, if higher costs by themselves reliably raised profits, Conlon says, firms would lobby for excise taxes on their own industry. They typically do not.

Testing the stronger claim about profit-led inflation thus really requires differentiating between the remaining demand story and the competition/​conduct story. At a macroeconomic level, real output growth was strong during the period where prices surged, providing evidence that this was a demand-led story.

Theoretically, there are ways you could try to distinguish causes at the microeconomic level too. ou need industry-level data where you (a) estimate demand elasticities, (b) measure or credibly proxy marginal costs, and then (c) to test whether pricing behaviour fits a more competitive model or a more cooperative one in different periods.

Even then, identification is hard. Conlon warns against treating “conduct” as a free parameter to estimate because simultaneous cost increases can look like a simultaneous softening of competition, given this also raises prices and lowers output. The practical way forward is testing models using moment conditions and, crucially, instruments: you need variables that shift demand (or “rotate marginal revenue”) but not marginal costs. When you have those tools, it is in principle possible to separate supply from demand and select a model of firm conduct.

That is what the strong claim commits you to show. You need a demonstrable break in conduct around 2021–22 across a wide swathe of industries, strong enough to move the aggregate price level, and not explainable by demand or cost changes. But greedflationists haven’t done that sort of work.

The evidence greedflationists have provided…

Now compare that standard of evidence to the evidence the greedflation proponents have actually offered. Conlon reviews their papers comprehensively, and the results are not flattering. To summarize, they have taken as evidence:

1. Periods where CPI was growing quicker than the Producer Price Index, implying that the difference is higher profit. Yet this is based on a misconception that the PPI is an index of input costs, when it actually measures prices received by producers “from the perspective of the seller.” Differences between those two price indices largely occurs because their baskets and weights are wildly different. The differential doesn’t track profitability. Even PPI sub-indices that do track the price of intermediate goods aren’t synonymous with production costs either, because they don’t include other inputs like labour, energy and transportation.
2. “Profits caused 50+% of inflation” national accounts decompositions. These papers break down “value added” into labor, non-labor costs, and “profits,” treating the identity like a causal equation to then claim profits going up can explain most of the price rise in specific periods. But these studies don’t illuminate the underlying causes of changes to these factor shares, and tend to be cherry-picked to short periods where the result holds to avoid longer periods where it looks like workers’ wages are “driving” inflation.
3. Markup and margin charts. Yes, some industries show rising markups during the recent inflation. But rising markups and rising prices are consistent with both increased demand or more collusion; that’s exactly the identification problem. Add to this that markup measurement is messy in practice, that firms are hard to assign to a single “industry,” accounting measures don’t map cleanly to marginal cost, and the timing often doesn’t line up neatly with popular narratives, and you’re left with weak suggestive evidence.
4. Earnings calls: “pricing power” as smoking gun. Again, announcements of price increases “in excess of cost increases” on earnings calls do not distinguish between “strong demand” from “changes in conduct” as the underlying cause. And several of the precise quotes held up seem to either just reflect executives preferring industry-wide cost shocks to firm specific ones, or confirm that they think demand has shifted. Proctor and Gamble’s CFO, for example, said the firm raised prices because consumers showed a “lower reaction…in terms of price elasticity than what we would have seen in the past.” That is not a confession of collusion or using cost increases as an excuse. It is telling you a demand change was the underlying driver!

The bottom line

If “profit-led” inflation means that strong demand in a supply-constrained economy produced temporarily higher profits in some sectors, then that’s true. But if it means inflation was driven primarily by a widespread shift to more cooperative or tacitly colluding conduct—prices and profits rising without demand—then it is a bold hypothesis that certainly hasn’t been tested, let alone proven.

Conlon is doing yeoman’s work, treating this with an economic seriousness it really doesn’t deserve. But even he is mainly just considering the microeconomic side. Even if there were some tacit collusion or a weakening of competitive conduct in some industries, that should only really be a big enough effect to alter relative prices, not the aggregate price level.

And that’s what I keep coming back to. You can’t have a sustained, broad-based rise in the overall price level unless households and firms are able to pay those higher prices. Where does that ability come from? Higher nominal spending! In which case, we are right back to surging demand from excessive stimulus being the cause of the inflation.