Testing Doux Commerce in the Lab

In a famous priming experiment it was shown that changing the name of a prisoner’s dilemma type problem from “The Community Game” to “The Wall Street Game” reduced the amount of cooperation. The suggestion is that Wall Street evokes in the mind concepts of exploitation and self-regarding behavior thus making these behaviors more likely. Wall Street is a very particular aspect of capitalism, however, what about the idea of markets and trade more generally? Montesquieu famously noted that

Commerce is a cure for the most destructive prejudices; for it is almost a general rule, that wherever we find agreeable manners, there commerce flourishes; and that wherever there is commerce, there we meet with agreeable manners.

In fact, market economies are associated with greater levels of trust and cooperation, so might we not expect markets and trust to be associated in the mind? Al-Ubaydli, Houser, Nye, Paganelli, and Pan (the list includes several GMU colleagues) prime experimentees with words associated with markets and then have them play a trust game; they find evidence in support of the hypothesis:

Using randomized control, we find evidence that priming markets leaves people more optimistic about the trustworthiness of anonymous strangers and therefore increases trusting decisions and, in turn, social efficiency. Given the general mechanisms by which priming affects behavior–that an individual’s mental representation of markets is the result of the individual’s experiences with markets–we can interpret our results as evidence in favor of the hypothesis that market participation increases trust.

…Absent markets, economic interactions with strangers tend to be negative. Market proliferation allows good things to happen when interacting with strangers, thus encouraging optimism and leading to more trusting behaviors. Participation in markets, rather than making people suspicious, makes people more likely to trust anonymous strangers. Our results seem therefore to corroborate the idea of doux commerce….We stress, however, that this is cautious evidence; a wider array of evidence is necessary for the solidification of this conclusion.


Game theory can offer hints about what kinds of market behavior maximizes trust. For example, long term rootedness helps. The ideal situation might be when all participants in the economic market also participate as families in the local marriage market, so business people have motivations to preserve the good name of the family in order to help their children make good marriages.

Here's a fascinating case study of a situation in which those conditions didn't apply:

Postville: A Clash of Cultures in Heartland America by Stephen G. Bloom


Rootedness and trust inarguably go together. Interesting that Stalin-era Russians used the phrase "rootless cosmopolitans" to suggest a certain untrustworthiness.

The central rule seems to be "only do business with people you can hurt if you find they have ripped you off". The hurt can be reputational or physical or legal - but if people are highly mobile this cannot hold.

Interesting. I hadn't personally thought much about it in those terms before but it does seem fairly logical.

After all, "markets" presuppose well defined property rights and, more generally, the rule of law. These two elements sure make interactions with strangers a lot more trustworthy (and safe!) than the law of the jungle.

I wonder if the priming example contains a problem.

If you prime with the word "Community Game", what you implicitly indicate is that there will be a recurring interaction with an individual, who is part of the community, such that there can either be future games of cooperation, or games of retribution, dependent on your first move. In other words, "Community Game" signals repeat games with the same person, same sets of persons, monitoring, and repeatedness.

Whereas, "Wall Street Game" implies a one shot game, not repeated with the same person, low likelihood of an opportunity for future cooperation, etc. So, I think you are missing something with your statement: "market economies are associated with greater levels of trust and cooperation" They are IF there is repeatability, monitoring, likelihood of future interaction, rules negotiated prior to trade, etc. Simply selling on an exchange without rules, nope.

I would argue your game depends on the term "market economy". So, for the example above, involving transactions on impersonal exchanges...no.

But, if there is repeated contact where you can monitor, where there are opportunities for future exchange, where there are situations where persons can be situationally disadvantaged and be subject to hold up...then, yes, a market game can be a cooperation game, and really, a community game. Furthermore, I would argue that "markets" often have rules--that you have to clear by such a date, you have to have so much cash in your account to trade, etc.--so that the rules provide for the "trust", and those rules are created by a trading community prior to the trade. So, the term "market with rules" may be the same as "community".

By the way, let me illustrate why future interaction is a key element in community games, and why, if you simply had a market without rules, you would have no exchange.

I have students play the Ultimatum game...where someone can allocate a fixed pie, provided the other person accepts, in which case they get the split determined by the dictator. Time and again, if there is a perceived unfair division--usually less than 20%--there is no allocation, or one party gets angry because they view it as "unfair". Then, I play the game again, or rather another game, where the party who received the unfair offer, in their mind, can choose or not choose to play in a cooperation game with another person, or where there is a split opportunity, this time with the split option on the other party. The result is that people don't cooperate, and the get revenge. I use this model in teaching about manufacturer/distributor/joint venture agreements to illustrate that sometimes you act fairly, in your best interest, because you may be at the short end of the stick during some part of the relationship, or may require unspecifiable cooperation from someone later. It's really fun to watch students try to negotiate later with someone they treated unfairly earlier.

The other reason I am a little skeptical of the word market as a prime (if it is not associated with rules or with monitoring) is from some results of E-Bay experiments involving the sale of baseball cards. Baseball cards where you could see the reputation of the seller sold at a higher price that cards where you couldn't see the reputation of the seller. If the term "market" connoted trust you wouldn't see these results. E-Bay is a market. If you just had the "market", that would be sufficient, but where there is "reputation" and reputation risk of the seller on future sales, you get a market and you get a higher price, and trades that you would not have otherwise been able to execute.

Looking at the paper, they didn't simply use the word "market." They derived a frequency list of words associated with commerce from the priming literature and then apparently validated the list with entirely different students (?) or people from somewhere else. I see these included words like "buy," "exchange," "deal," and "transactions."

Those words still don't make it in my mind, nor do they address empirical results such as the baseball card example. I wonder if there is implicit in the collection of terms a sense that there are rules that govern the market.

You could test this by using the term: "Wild West Market" and see what degree of trust you would find.

I think that's exactly the point. These words conjure images of all sorts of things and that's why priming has some effect.

Isn't most of the priming lit next to impossible to replicate? We know associative effects are real, and they can affect behavior and vice versa. So there's nothing special about priming. It just seems like there's been an explosion in these types of experiments because they're both easy to run and they make provocative pop-sci headlines.

Priming is pretty much the same thing as advertising, and the oldest joke in the marketing research business is "We know half our advertising budget is wasted, we just don't know which half." But that also implies that half of our advertising actually does manipulate people into buying your product.

It's not real hard to clue college students into how you want them to behave. If you call the game "Rape & Pillage" they'll assume you want them to behave differently than if you call the game "Mother Teresa." That "Wall Street" has connotations to students closer to "Rape & Pillage" is not the fault of the researchers.

This is perhaps the most obvious post in recent memory.

Forget Jared Diamond, look to British values. Not PC? Well too bad.

This is a bit of a tangent, but I recently watched a video over at Reason.com by Jonathan Haidt on disgust responses and morality. One of his points was that "liberals" along with libertarians have lower disgust responses than conservatives. My initial reaction was that, no, liberals have all sorts of disgust responses, say to GMOs, non"organic" food, etc.

But upon further consideration, I think there's a more important one: commerce. There's a general perception in our culture (not just the left part either) that there's something degrading about commercial activity. The presumed ideal, which underlies the appeal of communism, is a society where there are no commercial transactions, and all resources are shared. This is contrasted with the presumably degrading situation of paying people, or worse, being paid to do things that are useful for others. To a libertarian this sounds completely irrational. All other things being equal it's objectively better to get paid than to just do it for free, right?

From the perspective of evolutional biology though, it makes a lot of sense for there to be a digust reaction associated with commerce. Money (traditionally coin) changes hands frequently, and thus is a likely disease vector. Historically, disease spread along trade routes, including the bubonic plague, famously. Thus things that came into contact with money would be more likely to carry disease, and people who were heavily engaged in trade would be more likely to contract diseases. Eventually that may have evolved into a general aversion towards commerce.

What does this have to do with the OP? Nothing really, except maybe if ou're trying to convince people that markets are good for society and improve cooperation and so forth, it would probably help to recognize that people's aversion to markets and commerce isn't based in a rational appraisal of the benefits, but an irrational morality mechanism rooted in some of the brains deepest organs. No amount of reasoning with a Muslim that pork is ok to eat is going to convince him short of a wholesale religious conversion.

I'd like to see if this is reproducible with non-GMU students.

Where in the linked paper is there support for the claim "market economies are associated with greater levels of trust and cooperation?" They did look at what they called "indicators of formal institutions," which included some things that are necessary for a market economy like property rights and contract enforcement, but if that's all you're going on, I'm pretty unimpressed.

It's an interesting result and clearly we need to think about the market settings that engender those trust relationships and those that create barriers to them.

In the context of a different blog I was thinking about how the face to face market interaction I have with local shops, or private trades, and how even in the cases where the trade doesn't occur good information is learned by both sides. That exchange of information helps facilitate positive results in the next round of exchanges. I then thought about how I react to all marketing and survey requests. I simply don't trust them and don't participate because I don't think they goal is to better understand my needs -- and it's clearly not a bi-lateral exchange it's a unilateral exchange -- but rather is purely an attempt to gain information in order to capture additional consumer surplus from me.

The other aspect I like about the study is it seems to home in on a view of the Prisoner's Dilemma game that I always wondered about. What is preventing the players from establishing additional rules that allow the high payoff outcome to be the expected outcome. Granted, the Prisoner's Dilemma was created to induce defection because the underlying assumption is that stealing is wrong and society is better off when thieves are punished so defection is desirable. But often the game is applied to more neutral situations and then viewed as a market failure requiring government intervention. There's no standing back and asking if this is just an adjustment period or if there are other restrictions on the players addressing the problem themselves in a more organic way than government.

I don't think the PD was created to induce defection. The fact that it's framed as a criminal cooperation dilemma is just something the authors invented. Possibly to "unbias" the reader from assuming that cooperation is the "good" outcome.
IIRC, it was invented as a model problem to describe certain situations in international diplomacy during the cold war. (Defection = War, Cooperation = peace).

Comments for this post are closed