Tuesday, June 21, 2011

Wal-Mart v. Dukes

Sociologists are upset.  The American Sociological Association filed an amicus brief in the case of Wal-Mart versus Dukes to defend their study against claims that it is not scientific.  The case itself, and the ASA stance, both deserve the close and deep investigation which they will receive over time.

As a criminiologist, I consider myself a sociologist.  Crime could be considered a psychological problem or a medical condition or a form of economic engagement or a kind of political science.  Sociology encompasses all of those and more.  To me the defining space is that alone on his island, Robinson Crusoe's opportunities for crime are limited.  Crime is the sine qua non problem of every society.  Crimes begin as harms; and the plaintiffs in Wal-Mart v. Dukes claimed that they were harmed, and that under law, this was criminal.

The Wal-Mart employees obtained from sociologist William Bielby expert support for their claim that the corporate culture is discriminatory.  (Dr. Bielby served as president of the ASA in 2002-2003.)  Bielby's work was rejected by the Supreme Court of the United States.


Bielby’s testimony does nothing to advance respondents’ case. “[W]hether 0.5 percent or 95 percent of the employment decisions at Wal-Mart might be determined by stereotyped thinking” is the essential question on which respondents’ theory of commonality depends. If Bielby admittedly has no answer to that question, we can safely disregard what he has to say.  It is worlds away from “significant proof” that Wal-Mart “operated under a general policy of discrimination.” Supreme Court of the United States, Syllabus, WAL-MART STORES, INC. v. DUKES ET AL.  CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT No. 10–277. Argued March 29, 2011—Decided June 20, 2011 
The ruling by the Appeals court brought up the Daubert Standard, questioning whether the sociological investigation met that measure.  In fact, fhe Daubert Standard applies to non-scientific expert testimony, as well.  Whether or not sociology is a science, it is a recognized field of academic study, no less dependent on the development of expertise than are music, theater, or art.

Whether sociology as a science must follow the forms and methods of physics is a different question entirely.  The sociology of humor might be entirely descriptive and not statistically validated - even if there is nothing funny about this case - and still be science.

Restraining sociology to the same  practices as physics may be invalid on the grounds that humans are more complicated than billiard balls.  This applies to the law, well.  Justice depends on the "reasonable man" (person) finding fact.   Last night, I attended a presentation on contract negotiation sponsored by Ann Arbor SPARK our entrepreneurship incubator.  The first speaker offered common sense as your reliable guide to know whether or not contract terms are essential or tangential to the relationship. He did not have a statistically validated body of empirical evidence.  But he was, apparently, a recognized expert in the field of contract law, and his testimony on that would be admissible in court.

That a single sociologist cannot offer empirical evidence to support his claims does not remove sociology from the family of sciences.  It may say something about the sociologist or his theory.  Numbers are easy to find.  What is the male-female distribution in the population from which Wal-Mart draws its employees?  What is the gender distribution of Wal-Mart employees?  What is the gender distribution of Wal-Mart managers?  How likely is that outcome?    Do any other factors mitigate, as in professional sports, the raising of children, or among religious orders, obvious, structural gender biases.

If the distribution of employees by gender is not random, then discrimination may be in place.  Note, however, that "random" does not mean "evenly distributed."  The number pi is accepted as an infinite string of random numbers.  Yet, within it are curious runs of apparent patterns. They being apparent to us who make patterns does not invalidate their randomness.

The plaintiffs alleged, and the Supreme Court acknowledged, that at Wal-Mart promotion into management depends on subjective criteria, not objective standards.

Wal-Mart permits store managers to apply their own subjective criteria when selecting candidates as “support managers,” which is the first step on the path to management.  Admission to Wal-Mart’s management training program, however, does require that a candidate meet certain objective criteria, including an above-average performance rating, at least one year’s tenure in the applicant’s current position, and a willingness to relocate. But except for those requirements, regional and district managers have discretion to use their own judgment when selecting candidates for management training. Promotion to higher office—e.g., assistant manager, co-manager, or store manager—is similarly at the discretion of the employee’s superiors after prescribed objective factors are satisfied.

Measurable, objective performance standards are known in sports because every game has a score; and other player statistics are gathered, as well.  Yet, on any day, any team can beat any other (any solo can win over any opponent).  Building a team depends on more than getting players with good numbers. So, too, with much of life.  It may be true that in a completely anonymized world, 101 out of 201 people in any and every set will be males.  It may happen somewhere on Earth that 53 out of 53 in a business will be men.  I once worked a project with five other Michaels and a Michele.  What are the odds?


Wal-Mart v. Dukes may prove to be the Plessy v. Ferguson of our century: the dissenters - Ginsburg, Sotomayor, and Kagan - may win the day, eventually, in the history books.  If that happens, it will signal a deep change in society, not only in the USA, but globally.  One indication of that will be the loss of women with management skills to Wal-Mart's competitors.  

No comments:

Post a Comment