Similar regressions have been reported by Belongia and by Grennes and Lapp, and each study supported neutrality of money growth after intervals of one year or less. Using annual data, Grennes and Lapp found that, once real demand and supply forces were accounted for, there appeared to be no effect of inflation on real agricultural prices. Quarterly data were used by Belongia to test the money neutrality hypothesis. He found that the difference in growth rates between the GNP deflator for farm products and that for industrial commodities was not affected by unanticipated money growth after two quarters.The agricultural sector is specified as a series of supply and demand equalions with market prices playing the key equilibrating role. The agricultural sector is composed of two blocks of crop equations and three blocks of livestock equations. As shown in Figure 1, these blocks are related to the international and macroeconomy sectors-through a number of linkages. These consist of variables from the non-agricultural components of the model, such as interest and exchange rates,greenhouse snap clamps that affect the agricultural sector. A more general treatment of the linkages between the sectors incorporates feedback effects as well.
These were not included in the simulation results reported in this paper, so the agricultural sector can be thought of as a satellite model. Each grain block includes behavioral equations for acreage planted, yield per planted acre, domestic utilization, and inventories. Production is computed as the product of acreage and yield. Domestic utilization is divided into two components: livestock and residual demand and industry or food demand. Inventories are either publicly controlled loans, and stocks in the farmer-owned reserve or, privately owned. The privately held stocks and inventories under CCC loans are aggregated into a single inventory position. This specification allows different rules governing the movement of the various types of stocks to be incorporated in policy experiments. Since the planting decision is tied to the discrete choice of participation in farm programs, an appropriate specification must incorporate the trade-oft between expected returns of all potential crop choices. Traditional acreage response equations included in past models do not fully incorporate these trade-offs. Acreage planted of each crop is presumed to depend on expected returns from noncompliance and compliance with acreage programs for the crop under consideration, the expected profitabilities form competing crops, and last years’ acreage planted. The final variable is included since acreage planted is modeled as a partial adjustment process. Crop production costs depend on inputs purchased from the non-farm sector.
Costs are a function of the wage rate paid for hired labor; the market interest rate paid for financing working _capital, machinery, and buildings; prices paid for energy and fertilizer; and an index of nonfood prices. This cost measure enters the expected profit calculations for wheat and feed grains and provides a direct link with conditions in the general economy. When farmers do not participate in government programs, profitability depends, among other variables. on anticipated output price. For estimation purposes, the expected output price was taken to be the iarch price for a September futures contract. For simulation purposes, these price expectations .I were assumed to be rational, and the March futures price used in the acreage and yield equations was set equal to the cash price observed in the third quarter of the simulation. Thus, the price “expectations” used in the simulations are those which bring forth a level of production just sufficient to create market conditions consistent with that price, and the need to simulate the relationship between cash and futures prices is avoided. Domestic consumption is divided into food consumption and feed and other uses with separate demand equations for each component. Since most wheat , that is fed goes to broilers, the feed demand for wheat is specified to be a function of own price and corn price, each relative to the price of broilers, and the number of broilers on feed. Domestic feed demand for feed grains is specified to be a function of the inventories of cattle on feed, pigs on feed, broilers on feed and the price of grains relative to the price of meat. As suggested by the theory of consumption, domestic per capita food demand for wheat is a function of the real price of wheat, an index of real food prices, and real per capita income.
Food and industrial use of feed grains is modeled as a function of real feed grain prices, ” trend variable representing technology and real income. Inventory equations are used to·complete the grains blocks and determine the price of each crop. As noted above, inventories are separated into three components. In general a measure of the expected profitability of holding stocks is the main determinant of private stock holding. The different specifications for the various public inventory positions reflect constraints imposed on relase and entry in the publicly controlled stocks and by other causal influences. Quantity demanded by the private sector for stocks by both producers and users is motivated by transactions and precautionary motives. A large part is also due to the seasonality of production and to speculative motives. Specula active demand is influenced by the farm price relative to expected farm price. It is also presumed that the difference between the farm price and the loan rate, and public stocks have an influence. The market stock equation was modeled in price-dependent form. Interest rates enter the stock holding equations in two ways– the real interest rate and the government interest rate subsidy are both included as explanatory variables. In the private stock equations, it is expected that increased interest rates should have a negative effect due to the increased opportunity cost of holding idle inventories. As real interest rates rise, prices of wheat and feed grains fall since the opportunity costs of holding grain inventories has increased. Demand for stocks from the private sector is modeled in price dependent form. Stocks demanded by the government sector include government-owned stocks and the farmer-owned reserve. To a large extent, government stocks are a residual with the government playing a passive role. Farmers place stocks with the government when the farm price is close to or below the loan price by defaulting on non-recourse loans. They redeem loans only as the farm price moves above the loan price. A:5 required by law, the government can only release its own stocks when prices are sut efficiently above the loan price. In the case of the farmer-owned reserve, stocks flow out whenever market price approaches or exceeds the release price. The livestock sector includes blocks of equations for beef, pork and broilers. The structure of each block in the meat sector is similar. The meats are disaggregated to reflect different consumption patterns over time,snap clamps for greenhouse different income elasticities, and different production processes . Per capita meat demand is modeled in price-dependent form as a function of own quantity, the price of substitute meats, income, and the price of 1 nonfood items. Prices and income are measured in constant dollars, and income is in per capita terms. Supply behavior in the cattle sector is disaggregated into equations explaining the closing inventory of cows, placements of cattle on feed, and production of beef. The cattle sector is disaggregated and the dynamics associated with biological production lags and interactions between beef, feed prices and interest costs are incorporated. Our model follows that described by Jarvis; Freebairn and Rausser; and by Arzac a.nd Wilkinson except that, for simplicity, we have only one beef price. The cattle breeder and fed cattle activities are treated as distinct operations with different decision makers. Because of the biological lags involved, a change in the current cow inventory reflects a history of decisions to retain or slaughter cows and sell heifers to feeder operators or retain them for breeding over a period of three years. These decisions are related to current and past beef prices relative to feed costs and current and past real interest rates .Placement of cattle on feed is expressed as a function of lagged cow inventories to reflect the availability of feeder calves and the expected profitability of cattle feeding. Profitability is influenced by the price of beef relative to feed costs.
Feed costs for beef cows depend on the cost of feed grains, measured by the farm price of corn. Production of beef comes from gross number of placements of cattle on feed in previous periods, cull cows, and other nonfed cattle slaughter. Cull cows and other nonfed slaughter are modeled as the change in lagged cow inventories. The price of beef and the feed cost for beef may have two effects. In the short term, they encourage feeding of animals to heavier weights and withholding of heifers to increase the breeding stock. This gives rise to a negatively sloped short-run supply curve. In the longer term, the supply curve will be upward sloping as placements on feed,from the higher breding inventories increase. As with the cattle sector, the representation of the hog sector is highly aggregated. It allows for cyclical responses of pork production to changes in the final product price and costs. On the supply side. equations are given for the closing inventory of breeding sows, pig crop and production of pork. As with beef, the decision to retain breeding sows or send them for slaughter represents a series of decisions to retain or slaughter breeding sows and to feed pigs for slaughter or retain them to add to’ the breeding stock. At each period, these decisions are based on a comparison of the current value of pork to the expected returns from the sale of hogs in the future. The closing inventory of breeding sows is positively related to the price of hogs. negatively related to feed costs, and negatively related to the interest cost of holding inventories. The pig crop is a function of lagged breeding hog inventory and anticipated profitability from producing pork. Production of pork depends on previous pig crops and on liquidation of breeding inventories, which is measured by the previous period’s change in the breeding inventory of hogs.Production of broilers is modeled similarly to the beef and pork sub-components. Equations with the same type of causal influences are specified for poultry production, broiler chicks hatched, and broiler hatchery supply flocks. Real interest rates enter the livestock equations as a measure of the opportunity cost of holding livestock inventories. An increase in real interest rates tends to decrease current breeding inventories. increase current slaughter and production of meat, and push prices down. The longer run effects of an increase in the real interest rate will be an increase in meat prices due to smaller herds. As is apparent from the discussion above, macroeconomic variables are incorporated into the agricultural sector in a number of places. Income and prices of nonfood items affect food demand. interest rates affect the willingness to hold stocks of either crops or livestock. and the exchange rate and rest-of world prices and income affect the exports of feed grains and wheat from the U.S. We turn now to the simulation results to examine the extent to which variations in these macroeconomic variables induced by monetary policy affects variables in the agricultural sector.The ad hoc regressions reported earlier provided some evidence concerning the effect of money growth on the rate of change of food and nonfood prices. However. this approach reveals nothing about the effects on real incomes m the agricultural sector, since agricultural output was not included and there is no evidence concerning consumption, inventory behavior, and exports. Enders and Falk and Huffman and Langley have estimated similar regressions with growth in output as dependent variables, focusing on whether unanticipated money growth has output effects. While this method does add some information to the price change regressions, it is still not amenable to policy analysis. To investigate the effect of monetary policy on agricultural sector prices. quantities. and incomes, as well as to indicate its effects on government outlays for the feed grains and wheat programs. w~ used estimated equations for the above model using data through 1983. The starting point of the sample used for estimation varied with the stability of the equation over time and also depended on whether the dependent variable was quarterly or annual. but no sample data from before the 1960s were used.