Hispanic farmers seek class action status for discrimination suit

Although the U.S. Department of Agriculture has been compensating Black farmers who sued for past discrimination in USDA programs over many years, Hispanic farmers who allege similar discrimination have not been certified as a legal "class" that can jointly bring a lawsuit. The Hispanic farmers are still permitted to bring individual lawsuits alleging discrimination, but these are expensive and have little chance of success.

NPR's All Things Considered yesterday afternoon emphasized similarities between the two legal disputes, suggesting that the Hispanic farmers are being treated unfairly by comparison to Black farmers in similar circumstances.

Soon after President Reagan took office in the early 1980s, the USDA's civil rights division was quietly dismantled. Nevertheless, the agency continued to tell farmers that if they felt they weren't getting loans because of their color or gender, they should file a complaint.

But for the next 14 years, those complaints were put into an empty government office and never investigated. By the 1990s, black farmers filed a lawsuit — Pigford v. Glickman. Because the USDA failed to investigate years of discrimination complaints, U.S. District Judge Paul Friedman certified the black farmers' case as a class action. And with that ruling, rather than risk a trial, the federal government settled with 15,000 black farmers for $1 billion.

The next year, Hispanic farmers filed their lawsuit. And although their discrimination complaints had been thrown into the same empty USDA office, the judge in their case decided the Hispanic farmers would not be allowed to sue as a class.

I looked up the 2006 decision by the U.S. Court of Appeals, to see if the court mentioned any differences between the class-action petition of the Hispanic farmers and the earlier lawsuit by Black farmers. That appeals court decision centered on a debate about statistics. Because the USDA program rules on their face seemed to be non-discriminatory, the lawyers for the Hispanic farmers needed to show that there was systematic discrimination in practice.

First, an econometrician, Jerry Hausman, showed that Hispanic farmers seemed to get USDA program support at lower rates than non-Hispanic farmers. But, opposing attorneys and their experts pointed out that many of the Hispanic farmers may have never applied for support, in which case USDA could not be blamed if the farmers did not receive report. Second, the farmers presented another statistical analysis that appeared to show differences in USDA data for Hispanic and non-Hispanic loan applicants. But, opposing attorneys argued that some of the Hispanic applicants might not have been citizens, or they may have had deficient applications in other respects that were not "controlled" using regression analysis.

However, details such as citizenship information were not provided in the USDA data that were available for the second analysis. At one point, the author of this analysis, Karl Pavlovic, gave vent to his frustration (.pdf):

Dr. Freedman [the opposing expert] is simply using the myriad deficiencies in the databases produced by USDA to play a speculative game of 'gotcha' against the simple analyses that can be performed with the limited data produced. USDA produced some boards and a few nails. Dr. Freedman then criticizes my analyses for being a serviceable raft and not the Queen Mary.

As with many policy arguments, the question turns on where one places the burden of proof when absolutely clear answers cannot be found. The USDA and the court of appeals place the full burden on the plaintiffs to provide proofs that rule out alternative explanations for the disparate treatment of Hispanic farmers by USDA programs.

Because the chosen standard of proof is essentially impossible, this approach would mean that a farmer who faced real discrimination would not in fact be able to pursue a remedy through USDA or the courts.