Results from Experiment 2

The Fish Market with Shifting Supply Curves

Session 1

Figure 2.1 shows the demand and supply curves in Session 1 for a class with 47 students.    In a competitive market,  in this experiment, the supply of fish is fixed and fishermen are willing to supply their fish at any positive price.   Although demand and supply curves differ from class to class, the  competitive equilibrium price was always $15.
 
 

Figure 2.1


It is interesting to see that   in this session,  prices are  significantly different from the competitive equilibrium prices.
Figure 2.2 shows the distribution of classroom average prices in the first  round of  Session 1 and Figure 2.3 shows this distribution for the second round of Session 1. Figures 2.2 and 2.3 show the distribution of average prices observed in  each of  the classes in Rounds 1 and 2.   Average prices in Round 1 are seen to be well below the competitive equilibrium price of $15.   In round 2, these prices have risen slightly, but are still far from the competitive equilibrium prices.  We suspect that the reason for such slow movement toward equilibrium  can be seen by examining the supply and demand curves in Figure 2.1.  Notice that the vertical segment of the supply curve and the vertical segment of the demand curve are quite close together over the range of prices from $5 to $15.  This means that at any price between $5 and $15 there is little pressure for prices to change.  For example in the situation shown in Figure 2.1, if just two of the 21  buyers persistently  misunderstand their profit incentives or for some other reason fail to compete vigorously  to obtain a fish, the market could settle on any price between $5 and $15 with no pressure for change.

Figure 2.4 reports the complete distribution of all transaction prices in all sections. The purple line shows the histogram of  prices in Round 1 and the yellow line shows these prices for Round 2. From this figure it is evident that there is some tendency for prices to rise toward the competitive equilibrium price in Round 2, but the movement is slow.  It would have been interesting to see a third round.

Figure 2.2

Figure 2.3

Figure 2.4
 

Session 2

Supply and demand curves for Session 2 are shown in Figure  2.5.  The competitive equilibrium price is $5.
 
 

Figure 2.5
 







In the first round of this session, most classes had average prices closer to $9 than to the equilibrium value of $5.  In the second round, outcomes became more rather than less disperse across classes, though in many classes the price fell closer to the competitive equilibrium value.  We believe that the dispersion across classrooms is the result of a design defect in the experiment.   For some class sizes, the demand curve and supply curve would have been very close to each other over the vertical range from $5 to $15 and for other class sizes they would have been further apart.  Thus there would have been considerable variation in the amount of market pressure to move the price toward equilibrium.
 

Figure 2.8 shows the histograms for all prices in Rounds 1 and 2 of Session 2.  It appears that there is some downward movement in these price toward equilibrium, but this movement is very slow.

Figure 2.8

Session 3

In Session 3, the competitive equilibrium price is zero.  Although fishermen had to pay a fixed cost of $10 for gasoline, they have zero marginal cost and in competitive equilibrium are willing to supply at any positive price.  Each demander demands at most one fish and several more fish are available than the number of demanders. In the example shown below, there are 29 demanders and 40 fish are available.
Given the zero price, in competitive equilibrium, all fishermen suffer a net loss. Since each fisherman catches 2 fish, fishermen will lose money if the price falls below $5.

Figure 2.9

Figures 2.10, 2.11, and 2.12 show that the average price in classrooms tends to fall in each of the three rounds of this session.  By the third round, 10 of the 19 classes for which we have observations had   average prices well below $5 and so fishermen are in fact losing money.

Figure 2.10
 


 

Figure 2.11
 

Figure 2.12

We think that Figure 2.13 shows rather dramatic convergence toward competitive equilbrium prices over the three sessions.
 

Figure 2.13


Quantities in all Sessions

Figure 2.14 shows the distribution across all classes of the difference between the  number of transactions in the experiment and the competitive equilibrium quantity. Readers may be surprised to see that these results are strikingly different from the observation made by Edward Chamberlin (Journal of Political Economy, Vol. 56, 1948)  who pioneered the use of classroom markets.  Chamberlin observed that in his experiments the number of trades observed tended to exceed the competitive equilibrium quantity. This difference can be explained by the nature of supply and demand in this experiment.  In Sessions 1 and 2, competitive equilibrium requires that all  of the fish that are caught are sold to buyers.  The only way that the number of transactions can exceed the competitive equilibrium quantity is if somebody mistakenly (?) sells a fish that he or she does not have. This appears to have happened on just one occasion in all of the classes recorded.  In Session 3, in competitive equilibrium, all of the buyers will buy one fish.  Since buyers are not allowed to buy more than one fish, the only way that this can happen is if someone records
an illegitimate sale (or if there is an error in the reported number of buyers.)

The remarkable thing about these results is the frequency with which there were fewer sales than the equilibrium number, especially in Session 2.   This can happen only if at the end of trading there is some fisherman and some demander who could make mutual gains by making a trade.  In Session 2, some of the fishermen had 2 fish to sell, while in Session 1, they each had only 1.  For some reason, it appears that many fishermen didn't sell their second fish.  This helps to explain why the price was not forced down all the way to competitive equilibrium.  But why didn't they sell their second fish.  Can there be this many people who simply didn't understand the rules?  Another possibility is that there was some collusion among fishermen not to drive the price down by selling all of their fish.

Figure 2.14