Microdynamics • 227

competition. Finally, we consider how the structure of an industry emerges from the
competitive interplay of its member firms.

MICRODYNAMICS
We use the term microdynamics to refer to the unfolding of competition, over time,
among a small number of firms. This contrasts with macrodynamics, a term we use to
describe the evolution of overall market structure. Chapter 5 discussed two important
models of competition among small numbers of firms—the Cournot model of quan-
tity competition and the Bertrand model of price competition. Both of these models
were static; firms made decisions simultaneously. Though unrealistic, the models did
provide insights into important strategic concepts such as the revenue destruction
effect, and the impact of capacity constraints and consumer loyalty on competition.
But the static nature of the models clearly limits their ability to help us understand
strategic decision making in the real world, where strategies unfold over time. In the
first part of this chapter we explore how adding a time dimension affects strategic
options, by focusing on the following aspects of microdynamics:
• The Strategic Benefits of Commitment
• The Informational Benefits of Flexibility
• Competitive Discipline

Strategic Commitment
A strategic commitment alters the strategic decisions of rivals.1 As such, it must involve
an irreversible decision that is visible, understandable, and credible. The commitment
must be irreversible or it carries no commitment weight: the firm can back down if
the commitment does not have the desired strategic effect. It must be visible and
understandable or rivals will have nothing to react to. It must be credible so that rivals
believe the firm will actually carry out the commitment.2
The famous example of Hernán Cortés’s conquest of the Aztec Empire in Mexico
illustrates these concepts. When he landed in Mexico in 1518, Cortés ordered his men
to burn all but one of his ships. What appeared to be a suicidal act was in fact a move
that was purposeful and calculated: by eliminating their only method of retreat,
Cortés committed his men to the battle. According to Bernal Diaz del Castillo, who
chronicled Cortés’s conquest of the Aztecs, “Cortés said that we could look for no help
or assistance except from God for we now had no ships in which to return to Cuba.
Therefore we must rely on our own good swords and stout hearts.”3
To explore commitment in the context of models of competition, we shall revisit
the Cournot model of quantity competition described in Chapter 5. Recall that the
basic facts in that model are as follows: there are two firms (1 and 2) with identical
cost functions: TC1 5 10Q1 and TC2 5 10Q2. Market demand is given by P 5 100 2
(Q1 1 Q2). Each firm chooses its output simultaneously and treats its rival’s output
choice as fixed. We calculated that the resulting equilibrium quantities, prices, and
profits are Q1 5 Q2 5 30; P1 5 P2 5 40; and ␲1 5 ␲2 5 $900.
Suppose that instead of choosing quantities simultaneously, firm 1 can commit to
Q1 before firm 2 selects Q2. This could occur if firm 1 builds a new factory or signs
contracts with workers and suppliers prior to firm 2 taking similar actions. In this
228 • Chapter 7 • Dynamics: Competing Across Time

situation, known as a Stackelberg model, firm 1’s choice of Q1 can influence firm 2’s
choice of Q2. To see why, recall that firm 2 chooses Q2 according to the reaction func-
tion: Q2 5 45 2 0.5Q1. (See Chapter 5 for the derivation of the reaction function.)
The important difference between the Stackelberg model and the Cournot model is
that firm 2 does not have to guess the value of Q1. By building its factory first, firm 1
has committed to Q1 and firm 2 knows it.
Because firm 1 can compute firm 2’s reaction function, it knows exactly how much
firm 2 will produce in response to any choice of Q1. In other words, firm 1’s initial
choice of Q1 completely determines total quantity and the market price. This is
enough to allow firm 1 to compute its profits for any choice of Q1. In particular, firm
1 knows that the market price and profits will be:
Price: P 5 100 2 (Q1 1 Q2 ) 5 100 2 (Q1 1 (45 2 0.5Q1 )) 5 55 2 0.5Q1
Profits: ␲1 5 Revenue 2 Cost 5 PQ1 2 10Q1 5 (55 2 0.5Q1 ) ? Q1 2 10Q1
Some calculus reveals that the profit-maximizing value of Q1 5 45.4 In response,
firm 2 chooses Q2 5 22.5 and the market price is 32.5. Profits are ␲1 5 $1,012.5 and
␲2 5 $506.25. Firm 1 is doing much better than in the Cournot simultaneous choice
model, while firm 2 is doing much, much worse.
By committing to produce 45 units of output instead of 30, firm 1 has forced its
rival to cut back production to 22.5; this prevents price from falling too rapidly and
makes expansion more profitable for firm 1 than it was in the Cournot model, where
firm 2’s output was fixed.
As with the Cournot model, it is unrealistic to expect firms to compute such
detailed equations and perform the required calculus. But it is completely believable
that firm 1 would anticipate that its commitment to expand output would lead firm 2
to cut back production, providing exactly the incentive for expansion that the formal
model demonstrates.

Strategic Substitutes and Strategic Complements

In the Stackelberg game, firm 1’s decision to expand output caused firm 2 to contract
output. When one firm chooses more of some action, such as an output decision, and its
rival firm cuts back on the same action, we say that the actions are strategic substitutes.5
Quantities in the Stackelberg game are strategic substitutes. When one firm chooses more
of an action and its rival chooses more as well, the actions are strategic complements. Prices
are usually strategic complements; when one firm raises its price, its rivals may respond
by raising theirs. Certainly when one firm lowers its price, we expect its rivals to do so as
well. The concepts do not just apply to prices and quantities. If Burger King launches an
ad campaign and McDonald’s responds in kind, then advertising is a strategic comple-
ment. If Glaxo increases R&D investments in cardiovascular products and Merck scales
back its cardio R&D spending in response, then R&D is a strategic substitute.
To formalize the concepts of strategic complements and substitutes, we return to
the Cournot model of quantity setting and the Bertrand model of price setting. Recall
that in the Cournot model it was convenient to represent the equilibrium using reaction
functions. In a two-firm Cournot industry, a firm’s reaction function shows its profit-
maximizing quantity as a function of the quantity chosen by its competitor. In the
Cournot model, reaction functions are downward sloping, as Figure 7.1a shows. Reac-
tion functions in the Bertrand model with horizontally differentiated products are
defined analogously.6 In this case, however, the reaction functions are upward sloping,
as in Figure 7.1b.
 Microdynamics • 229

FIGURE 7.1
Strategic Substitutes and Complements

q2 p2
R1

R2

R2

R1
q1 p1
(a) (b)

Panel (a) shows the relation functions in a Cournot market. The reaction functions R1 and R2
slope downward, indicating that quantities are strategic substitutes. Panel (b) shows the reac-
tion functions in a Bertrand market with differentiated products. The reaction functions slope
upward, indicating that prices are strategic complements.

In general, when reaction functions are upward sloping, the firm’s actions (e.g.,
prices) are strategic complements. When actions are strategic complements, the more
of the action one firm chooses, the more of the action the other firm will also opti-
mally choose. In the Bertrand model, prices are strategic complements because when
one firm reduces prices, the other firm finds it profitable to reduce prices as well.
When reaction functions are downward sloping, the actions are strategic substitutes.
When actions are strategic substitutes, the more of the action one firm takes, the less
of the action the other firm optimally chooses. In the Cournot model, quantities are
strategic substitutes because when one firm increases its quantity, the other firm finds
it profitable to also increase quantity.

The Strategic Effect of Commitments

Commitments have both a direct and a strategic effect on a firm’s profitability. The
direct effect of the commitment is its impact on the present value of the firm’s profits
if the competitor’s behavior does not change. This is analogous to thinking about
quantity and price choices in the static Cournot and Bertrand models. For example, if
Nucor invests in a process that reduces the average variable cost of producing sheet
steel, the direct effect of the investment is the present value of the increase in Nucor’s
profit due to the reduction in its average variable costs, less the upfront cost of the
investment. The increase in profit would come not only from cost savings on existing
units produced, but also from any benefits Nucor gets from lowering its price or
increasing its output.
The strategic effect takes into account the competitive side effects of the commit-
ment. How does the commitment alter the tactical decisions of rivals and, ultimately,
the market equilibrium? In the Stackelberg game, the increase in production by firm 1
caused its rival to scale back production, which helped support pricing and increase
firm 1’s profits. Nucor’s investment would have a strategic effect if it caused rivals to
adjust their investment plans (or any other business decisions, for that matter). If a firm
takes the long view when making its commitment decision, as we believe it should, then
it must take into account how the commitment alters the nature of the equilibrium.