The proposed rule establishes a Class III price for milk used for cheese, Class IV price for milk used for butter and nonfat dry milk, and a weighted moving average of the Class III and Class IV price (whichever is higher) to be used as a Class I price mover. In so doing, the basic formula price (BFP) is effectively eliminated.
The purpose of this report is to evaluate the proposed rule for pricing manufactured products. It uses the same statistical procedures as in the previous Agricultural and Food Policy Center reports.(1) As explained in the University Study Committee report, these procedures place emphasis on the extent to which price movements reflect changes in supply and demand conditions, the extent to which they reflect movements in products prices and the extent to which they generate prices that are relatively stable. They do not evaluate the level of prices generated by the proposed rule. However, some observations will be made on the price level issue in the concluding section of this report.
Under the proposed rule, the Class III price is established by the sum of the following component price formulas:
The rationale for this formula is complex but is based on standard milk composition relationships. The calculation for the butterfat component uses the NASS AA butter survey price minus a make allowance of $0.079 per pound of butter divided by a moisture adjuster of 0.82 meaning that butter is 82 percent fat. The other solids component is equally simple utilizing the NASS survey price for whey minus a $0.10 make allowance for drying a pound of whey and adjusting for 3.2 percent moisture by dividing by 0.968.
The protein price formula is more complex. The first component subtracts the $0.127 per pound make allowance for a pound of cheese from the NASS cheese block survey price and multiplies the result by 1.32 which is the number of pounds of cheese made from an additional pound of protein. The second component represents the additional value of butterfat in cheese. It is included in the protein price to avoid the problem presented by having two prices for butterfat used in manufactured dairy products. The 1.582 is the number of pounds of cheese from an additional pound of fat. The additional value of butterfat is then multiplied by 1.20, which means that each pound of protein also creates 1.20 pounds of added value of milkfat in cheese, compared to butter.
The Class IV price is established by the sum of the following butter-powder formula components standardized to 3.5 butterfat and 8.7 percent nonfat solids:
The rationale for this formula is much more simple and easy to explain than was the case for cheese. The butterfat component is the same as for cheese. The nonfat solids component is the NASS survey price for nonfat dry milk minus the make allowance of $0.125 per pound and adjusting for 4 percent moisture by dividing by 0.96.
The Class I price mover is a weighted formula using the higher of the Class III and Class IV prices. Its rationale is based on a perceived need for greater price stability and the maintenance of a fixed differential between the Class I price and the higher of the two prices for manufactured products. The use of the higher of the two prices presumably is based on the need to be able to provide a constant level of monetary incentives to move milk from manufacturing plants into fluid use.
It may be perceived that it is not legitimate to analyze movements in the Class I price by the same criteria as for manufactured product prices. However, most economists believe that movements in Class I prices should reflect changes in manufactured product prices and changes in stocks of manufactured products because the Class I price affects the quantity of milk available for manufactured products. A Class I price that did not correspond with supply-demand conditions for manufactured products would be expected to generate greater Class III and/or Class IV price instability.
Table 1 indicates the Class III price, Class IV price and Class I price mover that would have been generated by the proposed rule for the period January 1994 to August 1997 as published by USDA. It is important to note that these are not the prices that would have existed had the proposed rule been in effect during this period. In addition, due to the fact that spot market rather than survey prices were used in computing these Class prices, if these prices had been in effect, then resulting supply and demand conditions would have been different thus generating different product prices. In the absence of market-generated prices, this analysis relied on the data generated by USDA from existing product prices.
Figure 1 provides a comparison between the Class III price and the BFP over the indicated time period. It may be noted from Table 1 and Figure 1 that the Class III price tends to lead the BFP - one would expect commodity market prices to lead competitive farm prices. Figure 2 compares the Class IV price with the BFP. Here the tendency to lead is less clear, perhaps because of the dominant influence of cheese in determining the price of milk in the Minnesota-Wisconsin region. Figure 3 compares the Class I price mover with the BFP. Note that because it uses the higher of the Class III and IV prices, its level is almost always higher than the BFP -averaging $0.78 per cwt higher.
The University Study Committee developed three criteria for quantitatively evaluating the options for setting manufacturing milk prices, including:
The statistical technique used to make those quantitative determinations is vector autoregression (VAR) time series analysis. VAR was utilized to analyze the impacts of changes in the price of milk on changes in product prices. It is a particularly useful technique in that feedback effects between milk prices and product prices are considered. At the same time, it allowed analyses of the relationship between milk prices and stocks -- a prime measure of responsiveness to changes in supply and demand conditions.
The VAR technique was utilized to evaluate the impact of a one-time 248 million pound milk equivalent change in stocks (one standard deviation) on the Class III price, Class IV price, Class I price mover under the proposed ruled and the BFP. The results are presented in Table 2 and may be summarized as follows:
Overall the options contained in the proposed rule compared favorably with those analyzed previously. Importantly, they represented a substantial improvement over the BFP.
When product prices change, the BFP should adjust to reflect both the magnitude of change in product prices and the share of products in the sales mix of manufactured products. VAR was used to measure the proportion of price variation in each proposed Class price option and the BFP that is explained by the prices of cheese, butter and NDM. The results are presented in Table 3 indicate:
Overall the options contained in the proposed rule compared favorably with those analyzed previously and represented a substantial improvement over the BFP.
Milk price instability has become a major producer concern. Economists are most concerned about price variability that cannot be explained by economic factors. The standard deviation of the VAR model, as reported in the price stability at 6 months, indicates the amount of price variation that cannot be explained by either product prices or stocks -- the two economic factors influencing price that were considered in our studies (Table 4). The results of these analysis indicate:
Once again the results for the options proposed compare favorably with those analyzed previously but also reflect the instability that is inherent in product prices where demands and supplies are highly inelastic.
Table 5 provides a rank ordering of the options included in the proposed rule compared with the BFP. Except from a stability perspective, the options consistently outperformed the BFP, and, therefore, represent an improvement over the current antiquated Minnesota-Wisconsin Grade B milk price survey.
While the VAR method of analysis is a powerful analytical tool, it does not give adequate attention to the level of price. It might be legitimately asserted that the proposed formulas generate prices that are too high relative to the BFP. The higher price is a result of the use of spot market prices as opposed to NASS survey prices, the assumed relatively low make allowance levels and the inclusion of dry whey as a by-product in the Class III price. If California, the number one milk-producing state, comes into the Federal Order System, these higher prices do not represent a significant concern. This is the case because if milk prices are too high, milk production will rise and, through supply-demand forces, will eventually lower the level of Class III, Class IV and the Class I mover prices. The opposite also is true.
In the absence of California becoming a part of the Federal Order System, conformity needs to be achieved in the pricing of milk used for manufacturing. Otherwise, major competitive problems could result. Over the period studied the California equivalent to the proposed rule Class III price would appear to be about $1.20 per cwt lower than the Class III price. The California equivalent of the proposed rule Class IV price would appear to be $0.70 lower.
These differences are sufficiently large to prompt a need to reevaluate the price levels generated by the proposed rule formulas. It is conceivable that some middle-ground, economically-sound compromise could be achieved between the two systems. In any event, these changes need to be known and agreed upon in advance to make informed decisions and avoid production and marketing disruptions. If such changes cannot be known in advance, an economic study of the implications of administering such disjointed milk pricing systems would appear to be a high priority.