Saturday, August 11, 2007

Class certified in Novartis gender discrimination suit

Bad news for Novartis Pharmaceuticals.

On July 31, 2007 Judge Gerald Lynch of the Southern District of New York granted class certification status to "[a]ll women who are currently holding, or have held, a sales-related job position with [Novartis] during the time period July 15, 2002 through the present."

The plaintiffs are seeking $200 million in compensatory, nominal, and punitive damages, claiming that Novartis discriminates against women in a variety of ways, including compensation, promotions, performance appraisals, and adverse treatment of women who take pregnancy leave.

The case in instructive for us because of how the judge viewed expert opinion in this case. One of the plaintiffs' experts noted that Novartis' performance evaluation system was flawed because ratings were subject to modification by higher-level supervisors and because ratings had to fit into a forced distribution. In addition, appeals by employees went to either the manager who made the original rating or an HR person with no real authority to change ratings.

Another plaintiffs' expert noted that male sales employees are 4.9 times more likely to get promoted to first-line manager than female sales employees. In addition, 15.2% of male employees were selected to be in the management development program compared to only 9.1% of eligible female employees--a difference of 6.0 standard deviations.

What these statistics really signify and whether the plaintiffs end up ultimately winning the suit is anyone's guess. The important thing here is to keep in mind that what you may think is a logical way to make promotion decisions may look "subjective" to others and riddled with potential for bias to enter the equation.

Bias (and risk) can be reduced by implementing practices such as:

1 - Having raters undergo intensive training, including a discussion of potential biases and several "dry runs" of the process.

2 - Having a standardized rating form with clear benchmarks based on an analysis of job requirements.

3 - Considering carefully the use of a "forced distribution" system. If you do use one, make sure raters and ratees alike understand why--and how--this is being done.

4 - Making performance in the current job only part of the promotional criteria--give applicants a chance to show their stuff through ability tests, work sample tests, personality tests, and the like.

5 - Taking complaints seriously. If someone believes there is an opportunity for abuse of the system, investigate.

6 - Track, track, track those applicant flow statistics, including selection into training programs. Uncover discrepancies before they uncover you.

7 - Get HR involved--not just as gatekeepers but as partners. Hold HR accountable for providing best practices.

8 - If you have something like a management academy, make the criteria for entry transparent and have a point person for questions.

You can read the order here, and read more analysis of the case here.

No comments: