Great Experiment, but I want more

Of all the experiments we have done this semester, I enjoyed this week’s one the most.

Of particular interest to me are the prevailing prices for intra-village trade. In my village, prices weren’t determined in terms of supply and demand, at least initially. Instead, prices were determined by each party involved making a partially verifiable representation about their payoff functions, and then seeing the trading partners try to match announced income levels. I say partially verifiable because, though you do see the quantities that each person uses to produce in each round, you don’t see their actual income. Though taking advantage of this would be difficult and require some weird moves, I think it could be done, at least under the income-equality pricing that I observed.

One of the cool things about this experiment is that it provides an outside option. Every person in the experiment has at least 2 potential trading partners, which should give participants the ability to punish cheating partners by negotiating with their other counterpart. For example, in my village, I was a red producer, trading with the blue merchant. At least in the initial rounds before the merchants realized they could trade with other villages, it would have made sense for me to change with the red producer in my village. However, that producer didn’t start specializing until late in the experiment, meaning that if the blue merchant had wanted to take advantage of his unique ability to produce blue units, I would have had no outside option and would have been forced to either accept his offer or simply produce nothing over the course of the experiment. However, this never really happened. My trading partner and I roughly matched income in spite of his market power. I find the fact that no one in our experiment managed to take advantage of this situation rather surprising, but I guess it falls in line with the greater narrative of unfairness-aversion within groups.

As an aside, it has been a little frustrating in this class, that people seem so willing to cooperate with each other, but it is to be expected when people go through so many experiments with each other; it is almost as if this entire class is a trust game. The most interesting part of this is that this Monday saw our final experiment, and the last period of our 4 month-long trust game went by without any collapse in the internal trust.

Back to the experiment however, once the merchant realized he could trade with people, we quickly ironed out how many pinks we would try to obtain and how much red/blue the pair of us were willing to give up to obtain it. This leads me to disagree with part of the analysis that was in the resulting paper. On page 290 of the Moral Markets book, we read “Notice all the unspoken trust in the red-blue­ village...Person 1, without producing or discussion, begins the day by immediately moving 20 reds to person 4”. I understand that there is no speaking prior to trading on the aforementioned days of the experiment, and I guess this does constitute trust. However, this is observed in day 15, when trading/producing habits are largely set in stone. Additionally, at least in my experiment, the norm was to discuss things in the latter half of the day, after the merchant had returned from the market as opposed to beforehand. So while it was true that I wouldn’t begin the day by bartering with my trading partner, that is in fact how I would end the previous day, making the silent exchange in the morning no more trusting than if I had hammered out the details in the moments following production.

In the last few rounds of my experiment, my merchant finally realized he had market power and began to charge higher prices for his services. As he began this process, I slowly began to feel out the other merchant to see if I could contract with him as a punishment/profit-making mechanism, and to see if it would go above and beyond my income from my habitual partner. This is the reason I am such a huge fan of this experiment. It allows for all of these levels of complexity in such a simple, basic environment.

I would love to see this done again, but with a couple of changes; namely the allowance of private bilateral communication, which would facilitate distrust and make it easier to change trading partners, and the exclusion of merchants from production, since that would more closely replicate real life. It is true that the merchants produced smaller quantities than producers, but due to their Min[2*blue,1*red] payoff function, it wasn’t as big of an issue.