Economic Classroom Experiments/Bertrand Competition

Bertrand (1883) modelled firms competing on price (as oppossed to quantity). The experiment demonstrates the Bertrand model and tests the equilibrium prediction.

Theory

Examine the case with two firms where both firms choose prices simultaneously and have constant marginal cost $c$. Firm one chooses $p_{1}$. Firm two chooses $p_{2}$. Consumers buy from the lowest price firm. (If $p_{1}=p_{2}$, each firm gets half the consumers.) An equilibrium is a choice of prices $p_{1}$ and $p_{2}$ such that firm 1 wouldn’t want to change his price given $p_{2}$ and firm 2 wouldn’t want to change her price given $p_{1}$.

Take firm 1’s decision if $p_{2}$ is strictly bigger than c: If he sets $p_{1}>p_{2}$, then he earns 0. If he sets $p_{1}=p_{2}$, then he earns $1/2*D(p_{2})*(p_{2}-c).$ If he sets $p_{1}$ such that $c<p_{1}<p_{2}$ he earns $D(p_{1})*(p_{1}-c).$ For a large enough $p_{1}$ that is still less than $p_{2}$, we have: $D(p_{1})*(p_{1}-c)>1/2*D(p_{2})*(p_{2}-c)$. From this we see that each has incentive to slightly undercut the other. Thus, an equilibrium is that both firms charge $p_{1}=p_{2}=c$. Note with three or more firms, in a (pure strategy) equilibrium, we need only two of the firms to set price equal to $c$, the others can charge a higher price.

Hand Run

Andreas Ortmann (2003) has published a hand-run version. On varying the number of firms in experiments on Bertrand competition see Martin Dufwenberg and Uri Gneezy (2000)

There are some elements to Bertrand competion in the Twenty-pound auction.

Computerized Version

There is a computerized version of this experiment available on both Veconlab and Exeter games site. The main advantage of the Exeter games version is the graphing feature which allows graphing of the average sale price in addition to average price. Looking at the graph of the data, this makes the results much clearer (since for three or more firms average price is not a reflection of proximity to equilibrium).

You can quickly log in as a subject to try out this group participation experiment, by pretending to be one of the original participants in a real session. You may also find the sample instructions helpful.

Results

This experiment has fairly consistent results. At Exeter, it has been run on sixth form (high school) students and Korean executives (from Korean Gas Corporation). When there are two firms in the market, prices are above marginal cost. This persists even when matching is random. Once we increase the number to 4 or 5 firms and random matching, the sales price drops to marginal cost. Note that the average price does not always do so. Three firms and random matching usually also goes to marginal cost.

Student Quotes

These quotes are from Exeter microeconomics students.

In reference to the two firm, Bertrand competition experiment:

“I learnt that collusion can take place in a competitive market even without any actual meeting taking place between the two parties.”

In reference to the four firm, random matching, Bertrand competition experiment:

“Some people are undercutting bastards!!! Seriously though, it was interesting to see how the theory is shown in practise.”

Discussion

We see cooperation (collusion) between firms in setting prices in practice as well. This can occur without direct communication.

Cooperation in Bertrand Competition

Case: The New York Post vs. the New York Daily News

time	Post Price	Daily News Price
January 1994	40¢	40¢
February 1994	50¢	40¢
March 1994	25¢ (in Staten Island)	40¢
July 1994	50¢	50¢

Until Feb 1994 both papers were sold at 40¢. Then the Post raised its price to 50¢ but the News held to 40¢ (since it was used to being the first mover). So in March the Post dropped its Staten Island price to 25¢ but kept its price elsewhere at 50¢, until News raised its price to 50¢ in July, having lost market share in Staten Island to the Post. No longer leader. So both were now priced at 50¢ everywhere in NYC.

Theoretical Cooperation

If firms get together to set prices or limit quantities, what would they choose?

As in your experiment, demand for the goods is $D(p)=15-p$ and each firm has the cost function of $c(q)=3q$ (marginal cost equal to 3).

Price Cooperation

Let us say the firms try to coordinate on a price that maximizes their profits. Here the firms' joint problem is $\max _{p}(p-3)\cdot (15-p)$. What is the choice of p? It is $p=9$. (The expression $(p-3)(15-p)$ is an upside-down parabola that starts at 3 and ends at 15 reaching its peak midway.) This is the monopoly price and D(p) is the monopoly quantity!

Quantity Cooperation

The joint firms' problem is $\max _{q_{1},q_{2}}(15-q_{1}-q_{2})\cdot (q_{1}+q_{2})-3(q_{1}+q_{2})$. All that matters in the solution is that $q_{1}+q_{2}=6$. (We can call $q:=q_{1}+q_{2}$ and then the problem becomes $\max _{q_{1},q_{2}}(12-q)\cdot (q)$.) This is the monopoly quantity and $D^{-1}(6)$ is the monopoly price.

[Slides] for teaching Bertrand Competition.

Footnotes

This page was first created by Toddkaplan 07:46, 26 April 2007 (UTC).