Reviving Micro Real Rigidities: The Importance of Demand Shocks

Why Do Prices Resist Change?

Economists often assume that when demand changes, firms' profit margins stay the same. This paper shows that if demand isn't perfectly uniform, unexpected changes in demand can actually change how much profit a firm wants to make. This, in turn, changes their prices and makes monetary policy much more impactful on the real economy.

The central mystery involves "monetary non-neutrality"—the phenomenon where changes in the money supply actually alter real economic output rather than just changing the price level. For decades, researchers have known that prices do not adjust instantly to economic shocks. This "stickiness" creates a gap between nominal changes and real consequences. The precise mechanics of why firms hesitate to change prices remain a subject of intense debate.

The search for real rigidities

The authors investigate why firms do not immediately adjust their prices when faced with changes in aggregate demand. In macroeconomic theory, this resistance is categorized into two types of "real rigidities" (forces that dampen desired price changes). Macro real rigidities are aggregate-level frictions, such as sticky wages or rigidities in supply networks. Micro real rigidities, however, are rooted in the individual firm's decision-making process.

The specific question the authors address is whether the mathematical shape of consumer demand can serve as a powerful source of these micro real rigidities. If a firm's desired markup (the profit margin added to the cost of production) is sensitive to the specific circumstances of its demand, then a nominal shock might do more than change spending. It might fundamentally change how firms perceive their own profitability and optimal pricing.

Cracks in the constant elasticity assumption

Historically, quantitative models have relied heavily on Constant Elasticity of Substitution (CES) demand. Under CES, the price elasticity—a measure of how much the quantity demanded drops when the price rises—is assumed to be a fixed number. This simplification makes models easier to solve. However, it carries a heavy cost: it implies that a firm's optimal markup is always constant.

The authors note that this assumption fails to capture a widespread empirical reality. In practice, cost pass-through (the extent to which a firm passes cost changes to consumers) is rarely one-for-one. When a firm's costs go up, they rarely pass the full amount to the consumer. Instead, they pass through only a fraction, typically between 20% and 50%. Furthermore, the standard CES framework predicts that prices and productivity should be perfectly negatively correlated. However, empirical data shows a much weaker correlation, suggesting that something else is influencing the desired price.

Disciplining demand with firm dynamics

To bridge this gap, the researchers move away from the rigid CES framework. They instead embed a Kimball (1995) demand system into a quantitative monetary model. Unlike CES, the Kimball system allows for "variable elasticity." This means a firm's sensitivity to price changes depends on its effective market share. This allows the model to replicate the variable markups and incomplete cost pass-through observed in the real world.

The investigation hinges on a critical methodological choice regarding "demand shocks." The authors test two ways these shocks can enter a model. In a "level-only" placement, a shock simply shifts the volume of demand without changing the firm's market position. In an "effective-share" placement, the shock shifts the firm's effective market share. This shift changes its local elasticity and desired markup.

The authors use firm-level evidence from U.S. manufacturing data to identify which placement is correct. By targeting the correlation between prices and productivity, they demonstrate that only the effective-share placement can match the empirical data. As shown in, this placement ensures that demand shocks directly move desired markups and prices.

This creates a much more dynamic and responsive pricing environment.

Amplifying the impact of money

The results of this calibration reveal that accounting for demand curvature significantly changes the predicted impact of monetary policy. The authors report that their model generates cumulative output responses to a nominal expenditure shock that are approximately 34% larger than those produced by a standard CES model.

When a 0.2% unexpected increase in nominal expenditure occurs, the authors find an impact response of 75% in real output .

This suggests that micro real rigidities act as an amplifier for monetary policy. The paper also finds that the model successfully replicates several "untargeted" pricing moments. These include the average size of price adjustments and the distribution of markups [Figures 2, 3, and 4]. For instance, the model's implied cost pass-through rate of 43% sits squarely within the 20% to 50% range documented in existing literature.

Connecting micro motives to macro effects

The findings suggest that the "missing" piece in many monetary models is the granular structure of demand. If this mechanism generalizes, it implies that traditional models may be systematically underestimating how much monetary policy affects the real economy. By focusing on the firm level, the authors provide a "portable framework" that links industrial organization to broader central bank policy.

For modelers looking to implement these findings, the authors suggest that demand shocks should be modeled as shifting the operating point (effective share) rather than acting as mere level shifters. This ensures that shocks to demand can effectively move desired markups.

The paper does not explore how these effects might change in an environment with high levels of firm exit or entry. It also does not address how financial frictions, such as credit constraints, might interact with these demand-side rigidities. A logical next step would be to test whether these demand-driven rigidities persist during periods of extreme economic volatility.

Figures from the paper