• @wise_pancake
    link
    15
    edit-2
    2 months ago

    Working with the Van Spaendonck Groep, MKB Nederland analyzed the anonymized payslips of 1.4 million employees in the ten most common professions - pedagogical employee, pharmacy assistant, service employee, administrative employee, sales employee, account manager, driver, production employee, warehouse worker, and mechanic. The results were corrected for age and experience, so these do not explain the wage gap.

    Typically that means they threw the numbers into a regression with wage ~ sex + age + experience and looked at the sex component coefficients. With a regression model you can use the coefficients to estimate average baselines for each sex separately from age. You’d want to clean up the age values into categories or use a regression spline to allow non linear age effects. You’d probably want to log transform wages too.

    They could have done a really shitty analysis though, without the actual details I’m just guessing and I didn’t see any links in the article.