Gregory Clark

A Farewell to Alms

Hindu law books of the first to ninth centuries AD allow interest of 15 percent of the face amount of loans secured by pledges of property, and 24-30 percent of loans with only personal security. Inscriptions recording perpetual temple endowments from the tenth century AD in southern India show a typical income yield of 15 percent of the investment. The return on these temple investments in southern India was still at least 10 percent in 1535-47, much higher than European interest rates by this time. At Tirupati Temple at the time of the Vijayanagar Empire the temple invested in irrigation improvements at a 10 percent return to the object of the donor. But since the temple only collected 63 percent on average of the rent of the irrigated land, the social return from these investments was as high as 16 percent.

While the rates quoted above are high, those quoted for earlier agrarian economies are even higher. In Sumer, the precursor to ancient Babylonia, between 3000 and 1900 BC rates of interest on loans of silver (as opposed to grain) were 20-25 percent. In Babylonia between 1900 and 732 BC the normal rates of return on loans of silver were 10-25 percent. In the sixth century BC the average rate on a sample of loans in Babylonia was 16-20 percent, even though these loans were typically secured by houses and other property. In the Ottoman Empire in the sixteenth century debt cases brought to court revealed interest rates of 10-20 percent.

When we consider forager societies the evidence on rates of return becomes much more indirect. There is no explicit capital market, and lending may be subject to substantial default risks given the lack of fixed assets with which to secure loans. However, one important element underlying the existence of interest rates in any society is a behavior called time preference. Time preference is simply the idea that, everything else being equal, people prefer to consume now rather than later. The time preference rate measures the strength of this preference. It is the percentage by which the amount of consumption of a good next year must be higher than consumption this year for people to be indifferent between consuming now or later.

Time preference rates are very high in young children and decline as they age. Experiments suggest that American 6-year-olds have time preference rates on the order of 3 percent per day. That is, they will delay collecting a reward only if they are offered the equivalent of an interest rate of at least 3 percent per day, or a monthly interest rate of 150 percent. Time preference rates also vary across people within a society. They are higher among the poor and less educated. Children with high time preference rates in preschool in California did less well academically later and had lower SAT scores.

Anthropologists have devised ways to measure time preference rates in premarket societies. They look, for example, at the relative rewards of activities whose benefits occur at different times in the future: digging up wild tubers or fishing with an immediate reward, as opposed to trapping with a reward delayed by days, as opposed to clearing and planting with a reward months in the future, as opposed to animal rearing with a reward years in the future.

A recent study of Mikea forager-farmers in Madagascar found, for example, that the typical Mikea household planted less than half as much land as was needed to feed themselves. Yet the returns from shirting cultivation of maize were enormous. A typical yield was a minimum of 74,000 kilocalories per hour of work. Foraging for tubers, in comparison, yielded an average return of 1,800 kilocalories per hour. Despite this the Mikea rely on foraging for a large share of their food, consequently spending most of their time foraging. This implies extraordinarily high time preference rates.