Charles Diamond, Ph.D. a director in the New York office and Morgan Hrab, M.A.S. a senior consultant in the Washington, D.C. office were recently featured in Law360’s Expert Analysis Employment Section for their article, “Using The 'Natural Logarithm' For OFCCP Pay Equity Analysis.”  The authors explore how the OFCCP’s recent guidance on the use of natural logarithm transformations in the pay equity context corrects a flaw in its monitoring and enforcement activities.  Regardless of an organization’s underlying data distribution, the log-linear model is typically the correct specification. Decades of empirical analysis based on this theory have confirmed that in many if not most every circumstance, pay follows a log-normal distribution, and hence pay growth is only linear when using the natural log as the measure of pay.  As discussed, this particular data transformation has many benefits. Like any statistical analysis, a regression is best carried out with careful consideration to both the underlying data being assessed and the assumptions necessary for producing a reliable and valid statistical result. Human Capital theory is a rare example in economic science of theory leading to an empirical specification for statistical work.

 

 

https://www.law360.com/employment/articles/1137514/using-the-natural-logarithm-for-ofccp-pay-equity-analysis