Lecture 4: Coordination Games: Multiplicity of Nash

Equilibria
Vera Sharunova

1 Coordination Games: Multiplicity of Nash Equi-

libria
In this lecture, we will consider a class of games that illustrate the fact that games
can have multiple Nash equilibria. We can always cook up an example of a game with
multiple Nash equilibria; unlike “cooked-up” examples, coordination games actually cap-
ture interesting aspects of some real-life and economic situations. Namely, coordination
games represent strategic situations in which players may have completely or partially
aligned interests, which make them prefer a particular Nash equilibrium out of all possi-
ble Nash equilibria in a game. We will consider three examples of coordination games: (i)
Game 1 with perfectly aligned interests; (ii) Game 2 with partially aligned interests; and
(iii) Game 3 with misaligned interests, inducing players to want to avoid coordination.

Game 1: Advertising Game

In 1970, Congress was considering banning cigarette advertising on radio and TV. When
they held hearings on the issue, no one from the cigarette industry showed up to argue
against the legislation. How so? Let us see how game theory sheds light onto this para-
dox.

Tobacco is addictive, so it is safe to assume that people will continue smoking even if
they are not exposed to cigarette advertising. However, advertisements are not useless
as they can shape brand loyalty among existing smokers. Let us capture these two ideas
by setting up a toy model of the advertising game, assuming there are only two identical
firms in the market.
Firm 2
Advertise Don’t Advertise
+
Advertise $400m $400m $700m, $300m
Firm 1
Don’t Advertise $300m, $700m $800m, $800m+

Q: Why do we call this situation an equilibrium? What does ‘equilibrium’ mean? Equi-
librium means that no player wants to deviate from their course of action, taking the

1
ECON3308.01: Game Theory in Economics Summer 2021

actions of the competitor as given.

This game has two Nash equilibria (Advertise, Advertise) and (Don’t Advertise, Don’t
Advertise). Prior to the ban, tobacco companies were in the (Advertise, Advertise)
equilibrium. They could not communicate and agree on not advertising, fearing being
caught in anticompetitive behavior. We can now explain why no one from the cigarette
industry showed up to argue against the legislation. Eliminating advertising from the set
of strategies allowed companies to coordinate on the “good” Nash equilibrium (Don’t
Advertise, Don’t Advertise). In fact, as noted in a 1972 paper published by James
Hamilton soon after the ban: “In the first year following the ban, advertising expen-
ditures were 20-30 per cent smaller than in 1970, and for the first six months of 1971
industry earnings were 30 per cent larger than for the same period in 1970.”

Q: What was the idea behind the ban? The idea was to prevent new consumers from
picking up smoking as a result of exposure to advertising.

Q: Is our toy model missing an important effect of advertising? Yes, advertising may
potentially expand the market for cigarettes through attracting new smokers. Did this
effect it matter in reality? Apparently not, as cigarette sales were primarily impacted
by health scare around tobacco consumption, largely created through subsidized anti-
smoking advertising on TV. The Public Health Cigarette Smoking Act of 1970 not only
proposed to ban radio and TV cigarette advertising, but also to eliminate free broad-
cast time subsidy given to antismoking advertisements, underestimating its effect on
smoking deterrence. As pointed out by Hamilton (1972): “Cigarette companies, unlike
ban advocates, apparently were not confused about the trade-off... As an indication of
their perceptions, prior to the Congressional ban the cigarette companies volunteered to
Congress a private broadcast advertising ban, if in exchange Congress gave the compa-
nies’ agreement antitrust immunity and all federal agencies dropped their proposals for
stronger health warnings.”

This example of a coordination game describes a situation in which all players’ interests
are completely aligned; everyone gets $800m profit in the “good” equilibrium outcome
and everyone gets $400m in the “bad” equilibrium outcome. Next, we will consider an
example of a coordination game where interests are partially, but not perfectly, aligned.
The risk of coordination failure in such games is a lot higher.

Game 2: Meeting Your Crush at a Dining Hall

This game is known in the field as the battle of the sexes. The name derives from
the story, cooked up by game theorists of the 1950s, to represent this payoff structure.
A husband and a wife have to silently coordinate on either going to see a ballet or a
boxing match with the payoffs assigned based on conventional gender preferences at the
time. We will consider a more palatable (in terms of equality of the sexes and lack of

2
ECON3308.01: Game Theory in Economics Summer 2021

communication) version of the game.

Imagine that you have a crush on someone at Boston College. You think that they like
you back, but you are not entirely sure. You want to play it cool by not confessing
your feelings first and just hope to “randomly” bump into your crush during lunch. The
problem is: Boston College has two dining halls. Assume also that you prefer going to
McElroy dining hall to get a nice salad at the Eagle’s Nest. You also once overheard that
your crush likes the Screaming Eagle sandwich, so you think they might go to Corcoran.
These ideas can be represented by the following payoff matrix

Your Crush
McElroy Corcoran
McElroy 3, 2+ 1, 1
You
Corcoran 0, 0 2, 3+

Q: What are the Nash equilibria of this game? There are two Nash equilibria (McElroy,
McElroy) and (Corcoran, Corcoran). Notice also that no equilibrium outcome is either
good or bad for both of the players.

Q: Is it possible to coordinate in this case without communication? Both players may

want to act tough and stick to their preferred dining hall. Both players may want to
act nice and go to the dining hall preferred by their crush and end up in an even worse
outcome.

Game 3: Chicken
The story comes from the game that was supposedly played by American teenagers in
the 1950s. Two teenagers take their cars to opposite ends of a street late at night and
start to drive toward each other. The one who swerves to avoid collision is the “chicken”,
and the one who keeps going straight is the winner. If both maintain a straight course,
there is a collision in which both cars are damaged and both players are injured. This
game is still a coordination game, even though the two players have to coordinate on
not choosing the same action.

Assuming that both teenagers, James and Dean, are reasonable people and consider
head-on collision a very bad outcome, the payoff matrix may look something like this

Dean
Swerve Straight
Swerve 0, 0 -1, 1+
James +
Straight 1, -1 -2, -2

This game has two Nash equilibria (Swerve, Straight) and (Straight, Swerve).

3
ECON3308.01: Game Theory in Economics Summer 2021

Q: How would you ensure that you win this game? You have to establish a reputation for
being a tough guy and convince your opponent that Straight is your dominant strategy
by modifying your payoff from (Straight, Straight). We call such players commitment
types. An example of a game with James maintaining a reputation of a “tough guy”:

Dean
Swerve Straight
Swerve 0, 0 -1, 1|
James
Straight 1, -1+ 1, -2−

References
[1] Hamilton, J. (1972). The Demand for Cigarettes: Advertising, the Health Scare,
and the Cigarette Advertising Ban. The Review of Economics and Statistics, 54(4),
pp.401-411.