Menu Cost

Menu cost is a core concept in business and economics theory and New Keynesian macroeconomics denoting any fixed real resource expenditure, managerial deliberation cost, or customer-relations friction incurred by a price-setting firm whenever it alters its nominal product prices. Originating as a literal metaphor for the expense a restaurant incurs when printing new paper menus—a term formalized in continuous-time optimal pricing models by Eytan Sheshinski and Yoram Weiss (1977) and placed at the center of business-cycle theory by N. Gregory Mankiw (1985), George Akerlof and Janet Yellen (1985), and Olivier Blanchard and Nobuhiro Kiyotaki (1987)—menu costs explain why profit-maximizing firms under monopolistic competition or oligopoly leave nominal prices unchanged for months at a time despite continuous fluctuations in aggregate supply and demand and monetary policy. Most importantly, through the application of the envelope theorem to imperfectly competitive markets, second-order private menu costs at the firm level can generate first-order aggregate welfare fluctuations across the macroeconomy.

Empirical studies of corporate price-setting—including Alan Blinder's (1998) direct surveys of chief executives and Daniel Levy, Mark Bergen, Shantanu Dutta, and Robert Venable's (1997) ethnographic audits of multi-store supermarket chains—demonstrate that physical repricing is only one component of a broader taxonomy of price-adjustment frictions:

1. Physical Repricing Costs: Printing and distributing updated catalogs, replacing shelf price tags across retail aisles, reprogramming point-of-sale barcode databases, and re-labeling inventory. Levy et al. (1997) documented that in large U.S. supermarket chains, physical menu costs averaged 0.70% of total revenue and 35.2% of net profit margins, consuming thousands of labor hours per store each year.
2. Managerial Information and Deliberation Costs (Zbaracki et al. 2004): Gathering market demand data, conducting competitive price elasticity audits, convening executive pricing committees, and retraining sales personnel. In industrial business-to-business manufacturing, Mark Zbaracki and colleagues (2004) found that managerial and sales-force decision costs are roughly six times larger than physical catalog-printing costs, totaling 1.22% of annual company revenue.
3. Customer-Anger and Implicit Contract Costs (Rotemberg 1982; Okun 1981): Frequent or opportunistic price increases after demand surges can alienate repeat customers, trigger search for rival suppliers, and damage brand reputation in long-term relational contracts.

Why Small Menu Costs Produce Large Business Cycles

Before the mid-1980s, New Classical macroeconomists argued that menu costs were far too small (fractions of a percent of revenue) to explain large recessions or the real output effects of central bank money-supply shocks. In independent 1985 papers, N. Gregory Mankiw ("Small Menu Costs and Large Business Cycles") and George Akerlof and Janet Yellen ("Can Small Deviations from Rationality Make Significant Differences to Economic Welfare?") resolved this paradox using the envelope theorem in a monopolistically competitive economy.

Consider a firm with market power facing a downward-sloping inverse demand curve P(Q) = M · D(Q), where M represents nominal aggregate demand (proportional to the money stock). At the unconstrained profit-maximizing price P^*, the derivative of the firm's profit function Π(P) with respect to its own price is zero:

Because the profit function is flat at its peak (P^*), a Taylor series expansion of Π(P) around the optimal price P^* has zero first-order term and begins only with the negative second-order curvature term:

Suppose nominal aggregate demand M falls by a small percentage ΔM (for example, a 2% monetary contraction). Because the private profit loss from leaving the old price P₀ unchanged is of second order—proportional to (ΔM)²—even a tiny fixed menu cost z > ΔΠ_private makes it privately optimal for the firm to leave its nominal price stuck at P₀.

First-Order Aggregate Externality Under Imperfect Competition

By contrast, in an economy with monopolistic competition or monopoly markups, the initial equilibrium output Q₀ is already inefficiently low because price exceeds marginal cost (P₀ > MC). When the firm refuses to cut its nominal price P₀ in response to a monetary contraction ΔM < 0, real money balances M / P₀ fall and real quantity sold drops by ΔQ < 0. The resulting loss in total social surplus W (the sum of consumer surplus and producer profit) is proportional to the pre-existing markup wedge (P₀ − MC) multiplied by the first-order quantity contraction ΔQ:

Because ΔW_social is of first order (proportional to ΔM) while the firm's private incentive to adjust price ΔΠ_private is of second order (proportional to (ΔM)²), the ratio of social welfare loss to private menu cost grows arbitrarily large for small monetary shocks. Blanchard and Kiyotaki (1987) formalized this gap as an aggregate demand externality: when one firm cuts its price, it slightly lowers the aggregate price index P, raising real money balances M/P and shifting outward the demand curves of all other firms in general equilibrium theory—a macroeconomic benefit that no individual price-setter internalizes.

The Sheshinski–Weiss (S, s) Band and the Selection Effect

In dynamic settings with ongoing trend inflation π > 0 and stochastic shocks, a firm facing a fixed menu cost z follows a state-dependent (S, s) pricing policy (analogous to inventory control models): the firm defines an inaction band [s, S] around its target real price p^*. So long as its current real price p_t = ln(P_i,t / P_t) remains inside the inaction interval (s, S), the firm pays zero menu cost and leaves its nominal price unchanged. The moment inflation or a cost shock pushes p_t outside the boundary s or S, the firm pays the menu cost z and resets its real price to the interior return point p^*. By the principle of smooth pasting under Brownian motion (Dixit 1991), the width of the inaction band scales with the cube root or fourth root of the menu cost (z^1/3 under deterministic drift; z^1/4 under pure Wiener diffusion), meaning a miniature menu cost of z = 0.0001 creates a wide inaction band of z^1/4 = 0.10 (10%).

In 1987, Andrew Caplin and Daniel Spulber proved a striking aggregation theorem: if monetary shocks are purely positive and the cross-sectional distribution of firms' real prices is uniform over [s, S], then a monetary injection ΔM > 0 pushes a fraction ΔM / (S − s) of firms across the lower threshold s, and each adjusting firm raises its price by the large step (S − s). In the aggregate, the product [ΔM / (S − s)] × (S − s) = ΔM, causing the aggregate price level P to jump one-for-one with M and rendering money completely neutral despite individual price stickiness.

Mikhail Golosov and Robert E. Lucas Jr. (2007) extended this insight to an economy with large idiosyncratic firm-level productivity shocks, calibrating the model to U.S. Consumer Price Index (CPI) micro-data compiled by Mark Bils and Peter Klenow (2004). Golosov and Lucas showed that state-dependent menu cost models exhibit a powerful selection effect: the specific firms that choose to pay the menu cost after an aggregate monetary shock are not a random sample of the economy (as assumed in Guillermo Calvo's 1983 time-dependent staggered pricing model), but rather the firms whose prices are furthest from optimum and who therefore make very large price adjustments, muting the real output effect of monetary shocks.

Subsequent multi-product menu cost models by Virgiliu Midrigan (2011) and Fernando Alvarez and Francesco Lippi (2014) demonstrated that when a retail or manufacturing firm sells multiple products and faces economies of scope in menu costs (paying a single fixed cost z allows the firm to reprice its entire catalog simultaneously), the selection effect weakens substantially: firms repricing due to a large shock on Product A simultaneously make small adjustments on Products B, C, and D, matching the empirically observed leptokurtic (fat-tailed) distribution of micro-level price changes and restoring strong real effects of monetary policy.

How can tiny menu costs cause large recessions if they cost less than 1% of revenue?

By the envelope theorem, a profit-maximizing firm's profit curve is flat at its optimal price P^*. Therefore, failing to adjust price in response to a small drop in aggregate demand causes only a second-order loss in private firm profit (proportional to (ΔM)²), so even a tiny menu cost prevents the firm from cutting its price. However, when all firms keep nominal prices rigid, real purchasing power falls and aggregate output contracts, creating a first-order loss in social welfare (proportional to ΔM) because imperfectly competitive prices exceed marginal cost.

What is the difference between state-dependent and time-dependent pricing?

In time-dependent pricing (such as the Calvo 1983 or Taylor 1980 staggered contract models), the probability or timing of a firm's price change is fixed exogenously by time regardless of economic conditions. In state-dependent pricing (menu cost models), the timing of price changes is endogenous: firms monitor the state of their real price and pay the menu cost to reprice whenever inflation or cost shocks push their price outside an optimal (S, s) inaction band.

What is the selection effect in Golosov and Lucas (2007)?

In a state-dependent menu cost model, after an expansionary monetary shock, the firms that cross the (S, s) threshold and choose to change their prices are specifically those whose old prices were already eroded near the lower boundary s. Because these selected firms increase their prices by a much larger amount than the average firm would in a random Calvo draw, the aggregate price level adjusts much faster and the real output effect of money is smaller.

1. Mankiw, N. G. (1985). "Small Menu Costs and Large Business Cycles: A Macroeconomic Model of Monopoly." The Quarterly Journal of Economics, 100(2), 529–537. https://doi.org/10.2307/1885395
2. Akerlof, G. A., & Yellen, J. L. (1985). "Can Small Deviations from Rationality Make Significant Differences to Economic Welfare?" The American Economic Review, 75(4), 708–720. https://www.jstor.org/stable/1821349
3. Caplin, A. S., & Spulber, D. F. (1987). "Menu Costs and the Neutrality of Money." The Quarterly Journal of Economics, 102(4), 703–725. https://doi.org/10.2307/1884277
4. Levy, D., Bergen, M., Dutta, S., & Venable, R. (1997). "The Magnitude of Menu Costs: Direct Evidence from Large U. S. Supermarket Chains." The Quarterly Journal of Economics, 112(3), 791–825. https://doi.org/10.1162/003355397555352
5. Golosov, M., & Lucas, R. E. (2007). "Menu Costs and Phillips Curves." Journal of Political Economy, 115(2), 171–199. https://doi.org/10.1086/512625