In the Feb. 22 edition of North American Market Preview, I addressed the findings of the recently released Livestock and Meat Marketing Study by USDA's Grain Inspection, Packers and Stockyards Administration (GIPSA). The report has been a topic of widespread discussion since its release, but a number of those discussions have misinterpreted the results.
A perfect example came up in a discussion with University of Missouri agricultural economist Ron Plain earlier this week. Both Ron and I believe it is important enough to warrant attention in this week's column.
One of the findings highlighted in the study's executive summary was this:
". . . the estimated elasticities of industry-derived demand indicate a 1% increase in contract hog quantities causes the spot market price to decrease by 0.88%, and a 1% increase in packer-owned hog quantities causes the spot market price to decrease by 0.28%."
As background, an elasticity is a measure of the responsiveness of one economic variable to a change in another variable. Elasticities are expressed as percentages in order to remove the effect of the units of measurement (i.e. data using pounds has bigger numbers than data using tons).
These measures say that if the number of contracted hogs or packer-owned hogs changes by 1%, the spot market price will move 0.88% and 0.28%, respectively, in the opposite direction.
Some have interpreted that to mean: "If we reduce the proportion of hogs sold under contracts by 1%, we will increase spot market prices by 0.88%." Or, similarly, "If we reduce the proportion of hogs owned by packers by 1%, spot market prices will increase by 0.28%." By that logic, if we just eliminate the 25% or so of hogs owned by packers, we will increase spot market prices by 0.25 x 0.28 = 0.07 or 7%. Better yet, if we just outlaw the contracts used to sell 60-65% of all hogs, we would increase prices by 0.60 x 0.88 = 0.528 or nearly 53%. Wouldn't that be great?
But those numbers are wrong. The research does not address changing the proportion of hogs sold under these arrangements. It used the number of hogs sold through the various methods. So, a change in the number of hogs sold under contracts would affect the total supply of hogs. No wonder the effect is negative -- it is a demand elasticity that measures the change in total quantity on price.
In addition, each elasticity applies only to the amount of hogs sold under those methods. So, a 1% change in the number of hogs sold under marketing contracts is 1% of roughly 60% or 0.6% of the total supply. Similarly, a 1% change in the number of hogs owned by packers amounts to 1% of about 25% = 0.25% of total supply.
GIPSA's data indicated that, for the study period of October 2002 through March 2005, 59% of hogs were sold through contracts, 11% were sold through negotiated trades, and 30% of hogs were packer-owned. Given these shares, the correct way to use the estimated elasticities is:
- Contract supplies. A 1% change in the number of hogs sold under contract would amount to a 0.59% change in total supply and that change in supply would be associated with a 0.88% price change in the opposite direction. The price flexibility for total supply would be -0.88/0.59 = -1.49. Meaning that prices move 1.49% for each 1% change in total supply, if the change is sold through contracts.
- Spot market hogs. GIPSA found that the price elasticity with respect to spot market supplies was -0.27. A 1% change in the number of spot market hogs would amount to .01 x .11 = .0011 or 0.11% change in total supply. The price flexibility for that change in total supply would be -0.27/0.11 = -2.45. Each 1% change in total supply drives prices 2.45% in the opposite direction, if the change is sold in spot markets.
- Packer-owned hogs. The price elasticity with respect to packer-owned pigs was 0.28. A 1% change in the number of packer-owned pigs would amount to 0.01 x 0.30 = 0.3% change in total hog supply. The price flexibility for that change would be -0.28/0.3 = -0.923. That is, each 1% change in the total supply drives prices 0.92% in the opposite direction, if the change is comprised of packer-owned hogs.
Those impacts are dramatically different than the raw numbers. In fact, packer-owned pigs have the smallest impact on spot market prices and spot-market-sold pigs have the largest. If you think about it for a moment, that makes sense. Wouldn't the pigs sold at the price being measured have the biggest impact?
Whether you like these numbers or not, we need to use them according to how they were derived and what they actually mean -- not according to what we want them to mean. In addition, we need to remember that these numbers are only accurate near the level at which they were estimated. Doubling the number of hogs sold under spot markets would not have a -0.27% impact for each 1%. The -0.27 only applies near the number of hogs that represent the 11% market share in the data set.
None of this means this study is perfect. It is not. The huge amount of data and vast variation in the way producers and packers provided data appears to have caused some problems. But let's take time to make sure we are using the numbers correctly and, if need be, take another look at the data and the research methods and improve on them.
Click to view graphs.
Steve R. Meyer, Ph.D.
Paragon Economics, Inc.