Small Menu Costs and Large Business Cycles: An Extension of Mankiw Model*

Hiranya K Nath†
Department of Economics and Intl. Business
Sam Houston State University

and

Robert Stretcher
Department of General Business and Finance
Sam Houston State University

November 2003

Abstract: Using a multi-period general equilibrium model, this paper extends the results
of Mankiw (1991) by showing that monopolistically competitive firms may require
‘relatively large’ menu costs to dissuade them from changing prices in response to an
aggregate demand shock that is perceived to be permanent. Thus, “small” menu costs
may be insufficient to contribute to large business cycles.

JEL Classification: E3

Keywords: Menu Costs; Nominal Rigidities; General Equilibrium

*
We thank Don Freeman for useful comments.
†
Corresponding author. Department of Economics and International Business, Sam Houston State
University, Huntsville TX 77341-2118; Phone: 936-294-4760; Fax: 936-294-3488; E-mail:
[email protected]
1. Introduction

It is by now a commonly accepted view among economists that nominal

rigiditities are the most apt characterization of the short run behavior of the economy.

However, the theories that have been proposed to explain sluggish adjustments of prices

and wages are varied and numerous1. One of the theories that gained popularity among a

section of economists in recent years suggests that firms are required to incur some costs

to change prices. These costs are often associated with printing menus, and therefore

referred to as ‘menu costs’. According to this menu costs theory, since changing prices is

costly, many firms do not respond immediately to a shock by changing prices, and as a

result, real variables such as output have to bear the brunt. Some economists, however,

cast doubts about this explanation because these menu costs are evidently small. Using

partial as well as general equilibrium models, Mankiw (1991) shows that these small

menu costs are in fact capable of producing large business cycles. Considering

monopolistically competitive firms that set prices, he shows that though menu costs may

be small, the incremental profits that result from price changes may be even smaller and,

therefore, firms are better off by not changing prices in response to a demand shock. In

Mankiw’s model the decision of the firm depends on a comparison between one-time

menu costs and the change in single-period profit. This paper argues that if the firms

consider changes in their future stream of profits that would result from the decision to

change price then ‘small menu costs’ may not be able to dissuade them from changing

prices. It essentially extends the results of Stretcher (2002), which presents a partial

equilibrium analysis of non-market clearing firm to show that introduction of the

1
For a comprehensive survey of these competing theories, see Blinder et al (1998) and Taylor (1998)

1
opportunity cost of capital to discount future incremental profits will reduce the ability of

‘small menu costs’ to generate large business cycles. In this paper, we build a general

equilibrium model which differs from the one in Mankiw (1991) in two ways: first, the

representative consumer maximizes life-time utility that involves inter-temporal transfer

of resources. Second and more importantly, the monopolistically competitive firm bases

its decision to change price on a comparison of the menu costs either with the change in

single-period profit, or with the discounted present value of the changes in all future

profits, depending upon whether it perceives the aggregate demand shock to be

temporary or permanent.

The rest of the paper is organized as follows. Section 2 presents a general

equilibrium model, with maximizing rules for consumers and firms. In section 3, we

introduce menu costs and discuss how they affect firms’ price setting behavior. This

section also includes the main propositions of this paper. Section 4 includes a few

concluding remarks.

2. A General Equilibrium Model with Monopolistically Competitive Firms

The economy consists of a continuum of monopolistically competitive firms,

distributed along the unit interval.

2.1. Consumers and Preferences

We assume that the economy is populated by a large number of identical

infinitely-lived consumers. The representative consumer has time-separable preferences

summarized by the following utility function:

∞ ⎛ 1 ⎛Md ⎞ ⎞
U = ∑ β t ⎜ ( 1 − φ ) −1 ∫ y i1,−t φ di + θ log ⎜⎜ t ⎟ − Lt ⎟
⎟ (1)
⎜ ⎟
t =0
⎝ 0 ⎝ Pt ⎠ ⎠

2
where 0<β<1 is the discount factor, yi,t is the quantity of good i she consumes in period t,

φ is the reciprocal of the elasticity of substitution between different goods produced by

the firms and 0<φ<1, M td is her money demand in period t, Pt is the general price level,

Lt is the labor supply2, and θ is the money demand parameter (θ > 0). The general price

level Pt is the geometric average of all Pi,ts, where Pi,t is the nominal price of the good

produced by firm i in period t, and is given as follows:

⎛1 ⎞
Pt = exp⎜ ∫ log Pi ,t di ⎟ (2)
⎝0 ⎠

The consumer earns wage income by supplying labor, and interest income from

lending in the previous period. She also receives money supply. In addition to spending

on consumption, the consumer lends. Thus the budget constraint for the representative

consumer is given by

∫ Pi ,t yi ,t di + Bt + M t = Wt Lt + Rt −1 Bt −1 + M t + Π t
d
(3)
0

where Wt is the nominal wage3 in period t, Bt is the amount lent in period t, Rt is the

interest rate in period t, Mt is the money supply and Πt is the total profits of the firms.

Note that Walras’s Law requires that the profits of the firms go to the individual. The

individual, however, considers profits as fixed in her utility maximization problem.

2.2. Firms and production

Each firm produces its output using labor only, and the technology is given by the

production function:

2
We may split this labor supply, by making the consumer decide the amount of labor she is willing to
supply to each firm. But since labor is perfectly mobile across firms this ‘twist’ in the model is
inconsequential. Also, the market clearing in the labor market requires that this labor supply is exactly
equal to the total demand for labor by the firms in the economy.
3
Since labor is mobile across firms, nominal wage rate is the same in all firms.

3
 yi,t = Li,t (4)

where Li,t is the labor input used by firm i in period t. Thus the cost function of the firm is

given by:

Ci,t = Wt Li,t = Wt yi,t (5)

The firm faces a demand function implied by the utility maximization and the firm

chooses yi,t and Pi,t in each period such that its profit is maximized.

2.3. Utility and Profit Maximization

The representative consumer maximizes her life-time utility given by equation (1)

subject to her budget constraint given by equation (3). The first-order conditions are

given below:

β t yi−,tφ − λt Pi ,t = 0 (6)

1 1
β tθ − λt = 0 (7)
M td Pt
Pt

− β t + λtWt = 0 (8)

λt − Et λt +1 Rt = 0 (9)

∫ Pi ,t yi ,t di + Bt + M t − Wt Lt − Rt −1 Bt −1 − M t − Π t = 0
d
(10)
0

Note that λt is the Lagrange multiplier for the budget constraint (3) in the consumer’s

utility maximization problem. Rearranging equation (8), we have

βt
λt = (11)
Wt

Substituting into equations (6), (7) and (9), and rearranging we obtain

4
 1
⎛W ⎞φ
y i ,t = ⎜⎜ t ⎟
⎟ (12)
⎝ Pi ,t ⎠

M td
Wt = (13)
θ

1 Wt +1
Rt = Et (14)
β Wt

Equilibrium in the money market implies that money supply equals money

demand. Thus,

M t = M td (15)

Substituting (15) into (13), we obtain:

Mt
Wt = (16)
θ

Then substituting (16) into (12) and (14),

1
⎛M ⎞φ
y i ,t = ⎜⎜ t ⎟
⎟ (17)
⎝ θPi ,t ⎠

1 M t +1
and Rt = Et (18)
β Mt

Rearranging equation (17)

Mt
Pi ,t = (19)
θyiφ,t

This is the inverse demand function faced by firm i in period t. Also, substituting for Wt

from (16) into the cost function (5), we obtain

Mt
Ci ,t = yi ,t (20)
θ

The implied profit function can be written as:

5
 π i ,t = ( y i1,t−φ − y i ,t )
Mt
(21)
θ

Firm i chooses yi,t in such a way that πi,t is maximized. The first-order condition of profit

maximization yields:

(( 1 − φ ) y −φ
i ,t )
−1 = 0

This implies

y *i ,t = (1 − φ )φ
1
(22)

where y*i ,t is the profit maximizing output of firm i in period t. Substituting for yi,t into

equation (19) we obtain the following profit-maximizing price for firm i in period t:

Mt
Pi*,t = (23)
θ (1 − φ )

As we can see from equations (22) and (23), a change in money supply does not affect

the profit-maximizing choice of output of firm i. It affects price only. Under ceteris

paribus, a one percent increase in money supply will increase the price of firm by one

percent. Thus, if all firms fully adjust prices in response to a monetary shock, then the

general price level will take the entire brunt of the shock leaving output unaltered.

3. Menu Costs and the Firm’s Decision to Change Price

Suppose the firm is required to incur a cost to change price. Following Mankiw

(1991), we assume that changing price involves a small labor input g. Thus, let the menu

cost of firm i be

Mt
zi,t =gt(i) Wt = gt(i) (24)
θ

6
The firm’s decision to change price depends on a comparison of these costs with potential

gains from such a change.

To start with, suppose the money supply is M0 in each period and each firm

chooses quantity and price according to equations (22) and (23), that maximize its profits.

Let y0 and P0 be the profit-maximizing quantity and price in each period corresponding to

this money supply. Suppose that suddenly the money supply is changed to M1 in period t.

If the firm decides to change its price, then the new price will be given by (23).

M0
Otherwise, it remains at P0 = . The nominal wage, however, changes from
θ (1 − φ )

M0 M1
W0 = to W1 = . Through product demand (equation (17)), output changes from y0
θ θ
1
⎛ M ⎞φ
to y1 = ⎜⎜ 1 ⎟⎟ y0 .
⎝ M0 ⎠

The firm’s decision to change price is based on whether the incremental profit

that results from the change in price outweighs the menu cost. However, it is important to

consider whether the firm perceives the shock to be transitory or permanent.

3.1. When the monetary shock is perceived to be transitory

If the firm perceives the change in money supply to be transitory, it will compare

the menu cost with the increment in profit in period t only. Because if the shock is

temporary then the money supply in the next periods will be M0, and y0 and P0 will still

be the profit-maximizing quantity and price. In that case, the marginal firm I that is

indifferent over changing price would be

⎛ ∆π i ,t ⎞
I = g −1 ⎜⎜ (( ) )
⎟⎟ = g −1 y01−φ − y11−φ − ( y0 − y1 ) (25)
⎝ W1 ⎠

7
If i<I, then the firm finds it profitable to change price even though it has to incur the

menu cost. If i>I, on the other hand, the firm leaves its price unaltered at P0 and produces

y1. Thus,

PROPOSITION 1: Following a monetary shock that is perceived to be transitory, if zi

(( ) )
> y01−φ − y11−φ − ( y0 − y1 ) W1 , then the firm does not change its price to P1. 4

3.2. When the monetary shock is perceived to be permanent

If the firm perceives the change in money supply to be permanent, on the other

hand, it will compare the menu cost with the discounted present value of all future

increments in profit in period t onwards. Because if the shock is permanent then the

money supply in all subsequent periods will remain at M1. If the firm does not change

price then y1 will be the output in period t and in all subsequent periods. In that case, the

marginal firm I that is indifferent over changing price would be

⎛∞ k
I = g −1 ⎜⎜ ∑ ⎛⎜ ∏ Rt +l −1 ⎞⎟
∆π t + k ⎞
([( ) ][ ])
⎟⎟ = g −1 y01−φ − y11−φ − ( y0 − y1 ) 1 + Rt−1 + (Rt Rt +1 )−1 + ...
⎝ k =0 ⎝ l =1 ⎠ W1 ⎠
(26)

From equation (18),

1
Rt +l = for all l = 0,1, 2, 3….. (27)
β

Thus, (26) becomes

⎛∞ k ∆π t + k ⎞
I = g −1 ⎜⎜ ∑ ⎛⎜ ∏ Rt + l −1 ⎞⎟ ([( ) ][ ])
⎟⎟ = g −1 y01−φ − y 11−φ − ( y 0 − y 1 ) 1 + β + β 2 + .......
⎝ k =0 ⎝ l =1 ⎠ W1 ⎠
(28)
[(
-1 ⎛
)
= g ⎜⎜ y 0 − y1 − ( y0 − y 1 )
1−φ 1−φ
] 1 ⎞
⎟
(1 − β ) ⎟⎠
⎝

4
If the shock is, in fact, temporary and the firm responds to the shock by changing its price to P1 then in the
next period it will have to change the price back to P0. In that case, the firm will incur the menu costs twice
and therefore will compare 2zi with the incremental profit in order to make a decision about price change. It
reinforces Mankiw’s (1991) result.

8
If i<I, then the firm changes price; otherwise, it leaves its price unchanged at P0. Thus,

PROPOSITION 2: Following a monetary shock that is perceived to be permanent, if zi

⎛
[( )
> ⎜⎜ y01−φ − y11−φ − ( y0 − y1 )
1 ⎞
] ⎟W , then the firm does not change its price to P1.
(1 − β ) ⎟⎠ 1
⎝

It is not difficult to show that [(

⎛ 1−φ
)
⎜⎜ y0 − y11−φ − ( y0 − y1 ) ] 1 ⎞
⎟>
(1 − β ) ⎟⎠
⎝

((y 1−φ
0 ) )
− y11−φ − ( y0 − y1 ) .5 Thus for given menu costs, the number of firms changing

prices in the latter case will be larger than in the former. In other words, if the firms

perceive the monetary shock to be permanent they will require relatively larger menu

costs to dissuade them from changing prices.

In both cases, total output is

1
Yt = ∫ y i ,t di = I y0 + ( 1 − I ) y1
0

The general price level is

⎛1 ⎞
Pt = exp⎜ ∫ log Pi ,t di ⎟ = exp(I log P1 + ( 1 − I ) log P0 )
⎝0 ⎠

When a monetary shock is perceived to be transitory, for given zis (even if it is small), I

will be closer to 0, and most firms will not change price. We will thus observe a

relatively larger effect of the monetary shock on output. On the other hand, if the

monetary shock is perceived to be permanent, I will be closer to 1 and most of the shock

will be absorbed by changes in prices. In that case, small menu costs may not be a likely

cause of large business cycles.

5
For example, for a value β = 0.95, the first term of this inequality is 20 times higher than the second term.

9
4. Concluding Remarks

Using a simple general equilibrium framework, this paper shows that if the firms

perceive the aggregate demand shock to be permanent they may require ‘not small’ but

‘relatively large’ menu costs to dissuade them from changing prices. In that case, their

decision to change prices will depend on a comparison between one-time menu costs and

discounted present value of all future incremental profits that would result from such

price changes.

10
References:

Blinder, Alan S, Elie E.D. Canetti, David E. Lebow and Jeremy B. Rudd, 1998, Asking
About Prices: A New Approach to Understanding Price Stickiness. New York:
Russell Sage Foundation

Mankiw, N. Gregory, 1991, Small Menu Costs and Large Business cycles: A
Macroeconomic Model of Monopoly, in N. Gregory Mankiw and David Romer (eds):
New Keynesian Economics, vol 1, Cambridge, MA: The MIT Press.

Stretcher, Robert, 2002, Discounting Price Rigidities, Journal of Economics and

Economic Education Research, vol. 3(2): 93-103.

Taylor, John B. 1998, Staggered Price and Wage Setting in Macroeconomics, NBER
Working Paper # 6754, Cambridge, MA.

11