Special to Tuesday’s Horse
Written by CAROLYN M. BETTS, PhD Economics
The press release circulated by Representative Kristi Noem (R-SD) “GAO Confirms Problems With Horse Processing Restrictions” makes two claims based on the GAO report. The elimination of domestic slaughter plants has (a) reduced domestic horse auction prices and (b) increased rates of domestic equine abandonment and neglect. Her claims are unsurprising, since the GAO report itself argues that both are true; yet neither can be substantiated based on the GAO’s own data and analysis.
The thrust of the GAO’s “theory” relating the cessation of domestic slaughter to horse prices, neglect and abandonment is as follows.
The cessation of US horse slaughter has resulted in higher costs of live horse transportation to slaughter plants, now located further away in Canada and Mexico. This has reduced profit margins of slaughter buyers in the US, and caused them to demand fewer horses for slaughter at auction at any price – it has shifted down the slaughter demand for auction horses.
This in turn has reduced average auction prices for slaughter horses and auction house commissions, which are partly proportionate to prices. The contraction in prices and commissions has caused closures of “low end” horse auctions nationwide since 2007. And these price effects and closures have reduced the rewards and opportunities available for owners wishing to discard their horses, causing them to neglect and abandon their animals instead.
To test this theory, at the very least one would need to (a) isolate the demand driven impact of the plant closures for auction prices of horses bound for slaughter, (b) isolate the effect of reduced demand and prices for sale volumes and auction closures, and (c) establish, quantitatively, a connection between reduced slaughter sale rewards and opportunities for owners, and increased abandonment and neglect.
Sadly, the GAO’s analysis accomplishes none of these things.
The GAO’s econometric model of equilibrium auction prices at three sales in the US is neither designed nor specified in a way such that the impact for prices of increased transport costs for slaughter buyers can be isolated from a myriad of other factors that shift demand and supply over time. A theory that recreational horse buyers had deserted auctions in favor of freely available “for sale” listings on Craig’s List since 2007 would account for the GAO’s econometric estimates of price effects as well as a theory that plant closures mattered.
Further, the GAO not only fails to conduct an analysis of sale volumes at auction, but also presents no data whatsoever on either auction closures or declining sales. Assuming such data exist, however, they also would be consistent with a myriad of alternative factors, including viral growth in free internet horse marketing options for cash-strapped owners.
But most devastating for the GAO’s “theory” is its failure to establish any credible link of auction price and availability with rates of abandonment and neglect.
A natural consequence of a decline in slaughter demand for auction horses that reduces prices by more than any other relevant factor (e.g. “the economy”), as the GAO argues on the basis of its econometric results, would be a decline in the number of horses slaughtered. And, a decline in the number of horses slaughtered would certainly suggest that some owners had effectively lost the “auction to slaughter” option for disposing of their horse.
Yet, the closure of U.S. slaughter plants has been associated with no significant change in the number of U.S. horses slaughtered annually.
As the GAO reports, federal trade and production data show the same number of US horses – approximately 138,000 – were slaughtered in the U.S., Canada, and Mexico in 2006 prior to the closure of the U.S. slaughter plants as were slaughtered in Canada and Mexico in 2010 following those closures. Even smoothing year to year fluctuations, and eliminating the transition year 2007, the number of U.S. horses slaughtered in the three years 2004 through 2006 prior to the closures was 346,835 – or an annual average of 115,612 – which is negligibly different from the number slaughtered in the three years 2008 through 2010 after the closures which was 346,520 – an annual average of 115,506.
There is, by definition, no correlation between something that stays roughly constant over time – the number of horses slaughtered – and something that the GAO claims has gone up significantly over the same time period – the number of horses abandoned and neglected. In the absence of an observable correlation, it is nothing short of heroic for the GAO to assume a causal relation from a proximate constant to a variable that it argues has increased.
This assumes, of course, that the very limited state and local data and the “anecdotes” (the GAO’s language, not mine) carefully recorded and analyzed from a small sample of 17 state veterinarians on neglect and abandonment utilized in the GAO study are to be taken seriously. If so, a more plausible explanation for any increase in neglected and abandoned horses is that it derives from increased economic hardship; reassuringly, for the data driven theorist, the GAO does at least assert that there is a positive correlation between them.
Caroline M. Betts received her Ph.D. from University of British Columbia. She is currently an Associate Professor of Economics at the University of Southern California Prof. Betts teaches in International Macroeconomics, International Finance and Macroeconomic Theory. Her current research interests are real exchange rates and relative prices, tradability of goods, market segmentation, pricing-to-market, patterns of trade and specialization, debt default and depression.