A grain of truth

Are grain farmers in Kazakhstan acting on bad information?


The grain market in Kazakhstan may be headed for the economic version of a train wreck. According to the business weekly Kursiv’, experts expect an increase in Chinese demand for Kazakhstani grain this year, accompanied by a shortage of grain carriers and granaries in the country as well as by a reduction in the harvest since last year’s record yield.

All of these factors presage an increase in grain prices here. But what is most striking about the market is its volatility. The grain yield in 2011 was more than twice that of 2010, which in turn was 40% below the yield in 2009, according to Kursiv’.

Cobweb, anyone?

The term refers to a market in which output and prices fluctuate because producers act upon mistaken expectations. Suppose that wheat prices are unusually low in Year 1. Farmers planning for next season’s market may assume that prices will remain anemic; and so they will sow little. As a result, the harvest will be small. This scarcity of wheat will send its prices soaring in Year 2. Now farmers, anticipating a bonanza, will sow a bumper crop…which will pull prices back down in Year 3. Economists call this predicament a “cobweb” because of the ever-growing oscillations of price and quantity when plotted on a market graph. Its vital feature is that the farmer always presumes that this year’s price will also be next year’s.

The grain market in Kazakhstan may be particularly vulnerable to cobwebs because nearly 90% of the country’s yield goes into the domestic market -- partly because of the lack of export facilities, although the country is a leading exporter of wheat and flour. In the domestic market, sharp changes in output may lead to sharp changes in prices.  On the other hand, wheat farmers in northern Kazakhstan may sow regardless of the expected price because they have few other uses for the land, suggest researchers at the United States Department of Agriculture


Cobwebs in the brain


A cobweb market has two causes. First, producers assume that current conditions will hold in the coming season. The government can address this misunderstanding by publicizing more sophisticated forecasts – or by developing the futures market, which rewards accurate forecasting.

Second, the farmer when planning his sowing assumes that his harvest will be too small to affect the national output and price. That’s rational, but if every farmer makes that assumption as well as the one of cobweb prices, then the national market may indeed gyrate.

One solution to this problem – organizing the farmers into a cartel that will make sowing decisions for all – is the sort of cure that kills the patient. Unfortunately, with its penchant for industrial planning, the government may choose essentially this treatment. A less intrusive policy would have the government – perhaps via KazAgro -- buy surplus grain in bumper years and release it in years when the harvest comes a cropper.

The cobweb model fails to explain why farmers repeat their errors when there is money on the table. Perhaps the market oscillations arise from rational expectations – which use all available information, not just past prices -- about shocks to the system. For example, a cattleman must decide whether to slaughter and sell his animal now or to breed it. A rise in beef demand that seems temporary may lead to more slaughters for a while. This affects the age distribution of the cattle, the birth and death rates, and ultimately the size of the stock. The cattle cycles generated may resemble cobwebs, but they stem from well-informed decisions.

At the University of Chicago, economists Sherwin Rosen, Kevin Murphy and José Scheinkman found that a rational-expectations model snugly fit the cattle cycles observed over 115 years when adjusted for trends in beef demand. They write: “Shocks to demand and supply have persistent long-term effects on future shocks by changing farmers’ incentives to carry breeding stock and altering the age composition and reproductive capacity of herds.”

Some studies have tried to observe cobweb expectations directly, by running a laboratory experiment. Participants in a simulated market are asked to predict the price in the next period. At the University of Arizona, Charissa Wellford found “little or no indication that cobweb or rational expectations are employed by the sellers.” (A pox on both houses.) Adaptive expectations, in which one forecasts the price partly by extrapolating past prices, “perform somewhat better.” This is not the first time that a cherished notion from economics has failed to find support in the lab. -- Leon Taylor tayloralmaty@gmail.com





Good reading

Mordecai Ezekiel. The cobweb theorem. Quarterly Journal of Economics 52: 255-80. February 1938.

United States Department of Agriculture, Production Estimates and Crop Assessment Division, Foreign Agricultural Service.  Kazakhstan Wheat Production: An Overview.  Online.


References

Sherwin Rosen, Kevin M. Murphy, and José A. Scheinkman. Cattle cycles. Journal of Political Economy 102: 468-92. June 1994.

Bakitzhan Toksharayev. Nazlo recordam. Kursiv’. March 20, 2012.  Page 1.

Charissa P. Wellford. A laboratory analysis of price dynamics and expectations in the cobweb model. University of Arizona Discussion Paper 89-15. 1989. Online.