Kessler David and Temin Peter (2007) “The organization of the grain trade in the early Roman Empire”, Economic History Review, 60/2, 313–332.
“Long-distance trade […] was beset by information problems”. Principal-agent issues have been studied for the medieval/early modern period, this article extends the scope to the Roman merchants. Who did they deal with asymmetric information? In particular how did the logistical nightmare that consisted of providing wheat to Rome’s inhabitants was overcome (313)? After the end of organized piracy in 67 BC, there was still a significant amount of uncertainty for the merchants. How did they managed their agents often located months away in distant provinces?
Historians studying other periods have often found that institutions had been ‘founded’ to deal with this information shortage. Thus they managed to keep the trade economically viable (314). Roman agriculture must have been quite productive as it succeeded in feeding a dozen cities with a population above 100,000, including Rome itself (at least a million). The metropolis’ inhabitants alone may have consumed some 300,000 metric tons of wheat a year. This required from 200 to 300 ships on the Mediterranean plus the barges on the Tibre (315).
From 15 to 30% of the grain consumed was given away by the government’s annona. But most of that was actually to private merchants through market mechanisms. Despite common reference in sources to “friends”, it is likely that the volume of trade was just too large to have been handled so informally. The government’s own exceptional attempts to regulate the prices indicate that a form of free market was at play (316).
Many of the grain trade actors were the rich and influential senators (although legally banned from doing business). Others were knights and wealthy freedmen (317). Most often merchants were grouped into companies to spread risks. It remains unclear whether these societas were permanent or not. Although, we know that a separation existed between ownership and management (called magistri). Some of these ‘companies’ may have had bureaux all over the Mediterranean.
The slowness of transports prevented an optimal decision-making by the top executives. To buy at the right time, one had to trust an agent in the provinces (319). But the quality and above all the trustworthiness an agent was next to impossible to estimate ex ante. Thus merchants often used kinsmen and friends as agents. Historians have observed that during other periods high wages, peer-monitoring and the oaths were used to reduce moral hazard by raising the costs of cheating (320). Evidently, being under one single rule of law made it easier to enforce contracts (321).
The public nature of the judicial system, and the circulation of writing afterwards, made the court a good deterrence for a wannabe contract-breaker. The annona office itself was very much able to destroy the reputation of those agents who had tried to defraud it. Thus the Prefect of the annona was an excellent information broker, providing insight about an agent’s reputation, market prices, stocks volume, etc. (322).
The principals’ caution
The ‘companies’ themselves were sophisticated information-sharing institution for its members (323). A complicated system of paperwork (triple bills of lading in particular) made it much more difficult for an agent to defraud his principal without being caught (324). A receipt was essential to conduct business. Labels and marks on the cargo were also commonly used. Samples of grains in sealed containers were also sent to the buyer so he could compare it to the rest of the cargo. Numerous other tricks were also used (325).
To make the numerous transactions necessary to bring the grain to Rome safer, the merchant sought to diminish the contact their agent had with the cargo. For instance, in Ostia, unloading the ships, storing and barging the grain to Rome was performed by guilds which – as sedentary organisations – were easier to monitor. Guilds functioned as self-enforcing institution by punishing themselves the misbehaviours of their members (326).
Reputation mechanisms worked very well in the small and close-knitted world of the Roman aristocracy, the importance of honour in the Roman culture may have carried an additional symbolic cost for the cheater (327). The fact that one agent may have several principals and one principal several agents made information travel even faster. “Various authors have presented an economy of friends as a substitute for a more formal market, but in fact they are complements” (328).
The same way as the annona was a public provider of information, grain merchant got physically together in a sort of hub on the Piazzale delle Corporazioni in Ostian increasing their chances to communicate news to one another. Wine dealers had the Forum Vinarium.
The initial calculations are really dodgy. The fact that all the grain was not produced locally does not mean that none was. The article is sometimes oddly written for instance, the fact that maritime loans (traiecticia pecunia) could include a clause stipulating that the borrower did not have to reimburse it in case of wreck, their interest rates were higher than usual amounting to a form of insurance in interesting but out of place.
More importantly, the whole article only slavishly applies a system developed elsewhere and brings little clearly new insight to the question. In it is not the author’s fault, sources ought to be very scarce at best and most likely inexistent. Still the result is a presentation of a way the Roman grain market may have functioned but with no guaranty that it did function accordingly. We have no idea of how effective the system was, we never see what actually happened when an agent did defraud his principal, basically this article does not bring much to its readers.