We’ve often written in this space about the Obama Administration’s very bad idea to take federal inspectors of the line at poultry processing plants, leaving the discovery of blood, guts, and feathers on the carcasses to overworked and underpaid line workers forced to process as many as 70 birds per minute at the average plant. The U.S. Department of Agriculture (USDA) is the architect of this proposal to “modernize” the food safety system without requiring a single additional test to make sure the birds are not infested with salmonella, campylobacter, and other bad bugs. Confirming the rule’s primary role as a windfall for the poultry industry, USDA’s initial cost-benefit analysis indicated that it would save companies like Holly Farm, Tyson’s, and Perdue $250 million annually. That windfall is attributable to the fact that under the proposal, the line speed will at least double, to as many as 175 birds per minute.
Today, GAO released a report that further discredits this bad idea, confirming the dire warnings of food safety experts: USDA is relying on junk data from a pilot program, raising the strong possibility that any final rule won’t survive judicial review. This rule is so controversial that it will almost certainly be challenged in court. When it is, it will be reviewed under the Administrative Procedure Act’s arbitrary and capricious standard, and USDA is running the risk that a judge will find the agency’s shoddy analysis fails to meet the standards of quality demanded by the Administrative Procedure Act (APA).
A dozen poultry slaughter facilities enrolled in that pilot program to test out the theory that removing USDA inspectors, handing over some of their inspection duties to plant employees, and speeding up the lines will somehow lead to chickens with fewer quality defects and microbial problems. GAO found that the data collected by the plants in the pilot program were inconsistently gathered and improperly aggregated. For example, GAO found that despite the fact plants in the pilot program collected sampling data for more than ten years, USDA cherry-picked numbers from just two, two-year snapshots to reach the conclusion that the pilot program has been a success. What’s more, GAO found that the economic data underpinning USDA’s cost-benefit analysis was old and potentially inaccurate. USDA’s numbers on labor costs were derived from data gathered between 1967 and 1992. And USDA’s estimates for the cost of redesigning slaughter lines to accommodate the proposed inspection system were based on a 2001 survey of pilot program participants – not necessarily a representative sample of the more than 300 plants that are potentially affected by the rule.
GAO slammed USDA for publishing the proposal without adequately disclosing the limitations of its data, and USDA has agreed to do a better job with that when it publishes the final rule. But simply acknowledging the data’s faults may not be enough at this point. GAO has identified such striking problems with USDA’s regulatory analysis that USDA ought to give serious consideration to the litigation risks of going forward with the rule. An argument could be made based on GAO’s analysis that the proposal is arbitrary, capricious, or not in accordance with USDA’s legal duty to establish rules that prevent adulterated poultry products from reaching consumers.
While GAO did a good job of independently auditing USDA’s food safety analyses, the congressional researchers could have done better on the occupational health and safety front. CPR and others have raised concerns about the ergonomic hazards posed by the fast line speeds that are already prevalent in the industry, and have criticized USDA for its proposal to allow those speeds to increase at the expense of workers’ health. We noted that the proposed rule failed to assess the significant costs to workers, employers, and insurance companies that arise from the musculoskeletal diseases that so many workers suffer. GAO briefly notes that USDA has promised to look at a NIOSH study of a single plant that is operating at increased line speeds and highlights some occupational health and safety concerns raised by other stakeholders, but the report could have gone much further and stated – as it did for the food safety analyses – that USDA’s underlying data is just too weak.
It’s no surprise that the occupational safety and health analysis behind the rule is so bad. USDA failed to consult with OSHA before the proposal was rushed out the door last year. Now that USDA has taken some heat for that failure, they’ve reached out to OSHA, but the request for help is a day late and a dollar short and to make matters worse, USDA officials have gone on record saying that they don’t think they have the authority to do anything to protect workers.
If this rule goes forward based on such poor data, we’ll basically be playing high-speed Russian roulette on the food safety front, and throwing plant workers to the wolves.