Data that show women earn less than men has been around for decades, and has often been explained away with a myriad of reasons: Different jobs, different education, different career path, and so on.
Researchers at GMI Ratings, a U.S. corporate governance firm, have done a study that suggests a glass ceiling exists: Pay for female chief financial officers of U.S. companies is at least 10 per cent below that of male CFOs.
Ah, skeptics might say: Although it's the same job, the female CFOs might work for smaller companies, or have shorter tenures. Tellingly, the GMI Ratings staff controlled for these factors – and still came up with its results.
The choice of examining the CFO position, GMI Ratings says, allowed for a statistically meaningful sample (there are far fewer female chief executive officers) and a population of executives who have roughly similar job responsibilities and financial backgrounds.
GMI Ratings found hardly any female CFOs at companies with market capitalizations under $100-million or above $25-billion, so it restricted the study to companies in that size range – and found the male CFOs made an average of $1.56-million, versus $1.35-million for female CFOs.
The conventional way to analyze the data, GMI Ratings said, would be to look at the overall differences in pay between female and male CFOs, and then break down the data to control for other variables. Instead, GMI Ratings built the model in reverse, trying to predict a CFO's gender based on pay.
"If, in fact, there is no glass ceiling, and all differences in pay between men and women are explained by other variables, then compensation should be of no use in predicting the gender of a CFO," write the report's authors, Tom White, GMI Ratings' director of quantitative research, and Kimberly Gladman, its director of research and risk analytics.
Alas, their results showed otherwise. Their first run at the data, including company size, found that compensation is a "marginally significant" predictor of gender. GMI Ratings then included tenure as a CFO, age, and the number of outside corporate boards the CFO sat on.
A CFO's age was a statistically significant predictor, as the average age of a male CFO was 51, about 18 months older than the female CFOs. Board service was also significant, as female CFOs were about 50-per-cent more likely to serve on an outside board.
With these factors accounted for, however, gender remained statistically significant. "Our model indicates that the more a CFO is paid, the less likely the CFO is to be female."
Further tests showed that 62 per cent of the time, the model correctly predicted whether a CFO was female, based on pay. "Given that any model would randomly be correct 50 per cent of the time, 62 per cent is not necessarily an impressive result," GMI Ratings acknowledges.
"However, we weren't trying to build a perfect model to predict CFO gender – we were simply trying to demonstrate that a model that includes CFO compensation is predictive of CFO gender. If CFO pay predicts CFO gender, then it becomes very difficult to argue that pay equity exists across genders for CFOs."
Based on its model, GMI Ratings predicts that all other things being equal, the average female CFO would earn about $215,000 more if she were male – $1.54 million versus $1.32-million.
Special to The Globe and Mail