Over the past few weeks I explored some core concepts in development economics. Last week, I discussed the concept of nonseparability. The week before, I discussed heterogeneity. In both cases, my goal was to show that those two phenomena, though they are not completely absent in developed economies, are important in developing countries in that they can cause persistent poverty.
This week, I wanted to talk about nonanonymity. In the stylized Walrasian model, a consumer is a consumer is a consumer and a producer is a producer is a producer, and provided we are talking about a single good, it does not matter who buys what from whom–the price will be the same whether consumer 1 or consumer 2 buys from producer 1 or producer 2. “Anonymity” is a bit of a misnomer here, but it refers to the fact that the identity of the parties to a transaction do not affect the price of that transaction.
In developing countries, however, it is often the case that one’s identity does affect the price of a transaction. For example, in many African cities, it is not uncommon for taxi rides to have two prices: one for locals, and one for foreigners.
In a chapter* titled “Rational Peasants, Efficient Institutions, and a Theory of Rural Organization: Methodological Remarks for Development Economics,” which was published in Pranab Bardhan’s Economic Theory of Agrarian Institutions, Stiglitz (1989) points out two important ways in which transactions are nonanonymous:
- People tend to repeatedly interact with one another, and
- There exist distinct power relationships between individuals.
Let’s look at those two sources of nonanonymity in turn.
(Again, the point of this series of posts is not to talk about phenomena that are exclusive to developing countries–it is certainly the case that people interact with one another and that there exist power relationships between individuals in OECD countries–but to talk about phenomena that tend to drive economic outcomes significantly more in developing countries than they do in developed countries.)
Repeated Interactions
Markets in rural areas of developing countries often lack simple enforcement mechanisms, or mechanisms that allow determining whether a new trading partner has a good reputation.
The way people manage to sustain business and large transactions is by repeatedly trading with the same people, starting with small transactions of little consequence, only to increase transaction size as the number of successful transactions increases. That is, trust allows sustaining transactions of increasing importance and is a function of the number of successful transactions (i.e., transactions where no party got screwed) in the past. Both Jean-Philippe Platteau and Marcel Fafchamps have written extensively about this: Platteau in a two-part article in the Journal of Development Studies (part 1 and part 2), and Fafchamps in his 2004 monograph on market institutions.
Power Relationships
In feudal Europe, it was not uncommon for peasants to have to “sell” some of their crops to the local lord for a price equal to zero. Likewise, several countries have caste systems which dictate and codify social and economic interactions–the best example of which being India.
One example of the way in which power relationships can dictate economic outcomes is via triadic interactions, discussed by Basu (1986). The gist of the theory of triadic interactions is this: whereas most relevant economic interactions are dyadic (i.e., they depend only on two parties), some economic interactions are triadic in that they depend on the interactions between three parties.
For example, if I lease in land from a landlord in exchange for a share of the crop, I might not be able to get credit from the local moneylender until I have paid my landlord the crop I owe him in cases where they know each other, or the moneylender might put pressure on my landlord for me to repay the money I borrowed from the moneylender. This can also relate to patron-client relationships (about which I remember Kaushik Basu recommending in class that we read anthropologist Jan Breman’s Patronage and Exploitation).
I realize that nonanonymity and the concepts explored here will seem basic to anthropologists and most social scientists outside of economics. There was a time, however, where economists simply did not concern themselves with those notions, perhaps in a misguided attempt to stylize economic theory as much as possible. Yet, the relative importance of nonanonymous transactions in developing countries can clearly limit people’s welfare, maintain some in persistent poverty, and contribute to economic underdevelopment.
* I like to think of Stiglitz’s most important contribution to social science to be that of showing that people in developing countries are just as rational as people in developed countries. Early on in his career, Stiglitz would look at situations where seemingly dysfunctional institutions, inefficient outcomes, or irrational behavior would persist in equilibrium and masterfully explain that they were the outcome of rational decision making, usually on the basis of asymmetric information between the parties involved. See examples here, here, and here. In my opinion, though I cannot claim to have read everything Stiglitz has ever written, nowhere does he summarize his rationalist and humanist position as well as in this chapter.