We now discuss alternative hypotheses about the nature of the trading mechanism. Consider
rst the hypothesis of perfect collusion among traders. As we discussed in Section 4, if traders
colluded with each other perfectly, they would force the farmer down to his reservation price for
potatoes, and whether the farmer knew the realization of would not matter for the outcome.
The information interventions would then have no eect on the equilibrium. So this hypothesis
cannot explain the heterogenous treatment eects we observe. Similarly, if all traders engaged
in simultaneous price competition, and the farmer did not have an option to sell directly in the
wholesale market, the farmer's response to price oers of various traders would be independent of
his prior information about the wholesale market price. The equilibrium would then be unaected
by the information interventions.