Why women have to work harder to be promoted

Women have up to one-and-a-half year's extra education, and nearly a full year's extra workforce experience, than what is required for their job, research finds.

It took a Nobel Prize before Canadian physicist Donna Strickland got promoted to a full professorship. As anecdotal evidence that women have to prove themselves even more than men to earn a job promotion, her story is hard to beat.

Looking deeper, it's more complex than outright sexism. Strickland herself dismissed suggestions her career had ever been stymied by being treated differently to her male colleagues. Her explanation for why her intellect and achievements had not been recognised by promotion to full professor? "I never applied."
That she had to add a Nobel Prize to her CV before she considered applying for her promotion is likely to resonate among women. Many sense they need to do more than their male counterparts to prove their worth in the workplace.
    But is there any statistical evidence that women need higher credentials than men to be promoted and recognised in their own profession?
      Much research on gender gaps focuses on comparing absolute outcomes. This doesn't help with this specific question. We need a way to directly compare men and women on the same rung of the career ladder.
      To do this, I borrowed a methodology from productivity and efficiency analysis called stochastic frontier analysis. My findings: women have up to one-and-a-half year's extra education, and nearly a full year's extra workforce experience, than what is required for their job.

      Measuring capabilities

        This technique begins by measuring the overall capabilities of each worker. This is captured by their qualifications, years of experience, cognitive ability, language proficiency, health and traits that are shown to enhance productivity such as conscientiousness.
        Next, each worker's bundle of capabilities are benchmarked against workers at a similar rung on the career ladder. This position is captured by their workplace seniority, earnings and other entitlements (such as paid parental leave and flexible work arrangements).
        It is then possible to identify those workers who possess a higher bundle of capabilities than the minimum required to reach their current position on the career ladder.
        The Household, Income and Labour Dynamics in Australia (HILDA) survey provides a broad sample of about 5,000 workers to study.
        Comparing everyone else to the "most economical" performers -- those who have achieved their position with the lowest capabilities to their name -- I found women "over-invest" more than men in their skills and capabilities.
        Men over-invest by up to 4%. Women by up to 11% -- equivalent to a year-and-a-half extra education and nearly a full year's extra workforce experience.
        These figures, using stochastic frontier methodology, avoid the distortions arising from simpler statistical methods like averages. We're looking at a distribution curve for capability "overinvestment" -- a curve that is not neatly symmetrical. At one end of the curve are the "most economical" performers. At the other end, in the tail of the curve, are the workers who are furthest away from the minimum requirement and have over-accumulated

        Uncovering implicit biases

        This technique allows us to test what factors are driving women's greater accumulation of credentials. Tellingly, this over-investment isn't directly connected to children and care responsibilities. Nor is it due to women's lower confidence, a hypothesis that I tested by including a variable called "achievement motivation".
        Existing evidence steers us towards implicit biases wov