Using Statistics in Charges of Discrimination Against City Hospitals

Under fiscal pressure, the New York City Health & Hospitals Corporation (HHC), which administers the city’s public hospitals, proposed new measures to reduce spending, including closing two municipal hospitals. In response, a class action suit alleging bias, filed in the U.S. District Court for the Southern District of New York, demanded an end to budget cuts and the restoration of services.

The suit argued that HHC’s actions had already negatively impacted blacks and Hispanics, and that further reductions would increase that impact. To support these demands, a statistics expert identified a “statistically significant” difference between the hospital services provided to minorities and others.  

Testifying on behalf of New York City, Dr. William Fairley of Analysis & Inference found that the plaintiff’s expert had incorrectly applied a “binomial” statistical model, which presented the closing of hundreds of beds as substantial evidence of discrimination. But in fact, only two decisions – to close two hospitals – could be the basis of discrimination. Applying a proper statistical test, Dr. Fairley showed conclusively that the choice of hospitals could have been the result of random selection, and not a discriminatory process.

The court (J. Abraham Sofaer) stated in its opinion: “Dr. Fairley made clear how little utility the plaintiff’s data possess... To utilize the data presented by plaintiffs on the issue of discriminatory purpose, Dr. Fairley explained, one would have to assume that the decision maker whose motivations were under scrutiny made individual, independent decisions to reduce beds in the system or in any hospital studied…”  

Judge Sofaer added: “No effort was made to controvert Dr. Fairley’s convincing analyses, all of which showed that the proposals in the Mayor’s Plan could reasonably have been reached by random selection, thereby tending to negate the hypothesis of discriminatory intent.”