Danish love of gathering big data and the gender wage gap

“Mothers in Denmark Are Less Productive at Work, Study Finds, Partially Explaining Gender-Wage Gap” (WSJ) looks at an analysis of a comprehensive data set on Danish workers:

Young mothers are less productive at work than their male counterparts and women without children, according to a new study of Danish workers, a finding with important implications for gender-pay gaps.

Productivity is measured as output per hours worked, using Danish government records that tie workers’ demographics to output data from individual firms.

For example, mothers between 30 and 32 years old were about 87% as productive as similar childless men … Women without children between 30 and 32 years old were 101% as productive as similar men

An interesting article and study (“Motherhood and the Gender Productivity Gap”) for those passionate about big data sets and those passionate about gender issues (which I hope all of us are!).

[Related analysis made possible by the Danish passion for data… See the discussion of “Parental Responses to Child Support Obligations: Causal Evidence from Administrative Data” (Rossin-Slater and Wust 2014) within “Litigation, Alimony, and Child Support in the U.S. Economy”.

Using a comprehensive data set for all Danish adults, the authors found that for every 1,000DKK ($141 at 2017 exchange rates) additional that a father is supposed to pay under the Danish [child support] formula, the following behavioral changes are observed:

  • a 2.1 percent reduction in the likelihood that the father and child will ever live together
  • a 3.2 percent increase in the likelihood that the mother will have an additional child with a different man, either while married or not
  • a 3.7 percent increase in the likelihood that the father will have a subsequent child with a different woman (“higher obligations may lead to less time spent with existing children, freeing up time available to invest in future children”)
  • a reduction in his labor force participation rate by 0.2 percent (with larger effects for higher earners)
  • only about $70 additional actually paid, as fathers reduced their voluntary contributions]

On a casual reading, it sounds as though markets are rational. Women with children are 87 percent as productive as men without children and earn 85 percent as much. But then it seems that markets are not rational. Women without children were actually more productive than men (101 percent) and yet earned only 89 percent as much. On the third hand, perhaps this is an example of Gresham’s Law: “bad money drives out good”. It may not be possible for an employer to determine if a woman has children or is going to have children and therefore the economically sensible course of action is to pay her as though she does have children.

Readers: What do you think? In an age of fluid gender identity, can this kind of data set be of any value? If so, what can we learn from the data-driven Danes?


5 thoughts on “Danish love of gathering big data and the gender wage gap

  1. I wouldn’t trust this result. The age range of 30-32 is too narrow as if they are nit-piking to make a point. Not only that, they don’t tell us the age of the men they compared this data to but let us assume it is 30-32 too.

    A much useful data would to be cover a 10 year range window across the age range of 20 to 70.

  2. @George A, you can trust the results. The data age range was from 25-64. Reading the working paper will provide a better overall understanding of the study instead of drawing conclusions from an abbreviated summary.

  3. I’ve skimmed the study.

    What bothers me and have bothered for a while with research of this sort is that productivity, the crucial metric, is practically unmeasurable except in simple cases, e.g. bricklaying or drywall hanging. It seems that the dominant assumption in the study is that males and females of “equal” qualification (how do you measure that ?) are “perfect substitutes” in microeconomic sense, i.e. their productivity is the same. That assumption may or may not be true but again since one cannot measure this critical phenomenon, how do you really know ?

    Now, to your other question: “In an age of fluid gender identity, can this kind of data set be of any value?”. In the era of gender fluidity, the data is likely of no value. If you audacious enough, you can easily fix the wage gap by legally converting from a man to a “woman without children”. Just like the smart Argentinian gentle[wo]man who legally changed his/her gender to retire 5 years earlier:


  4. These sources may also be trustworthy.

    Where I live changing your gender is pretty easy: apply online, enjoy the 6 month grace period. Then get a new social security number, a new drivers license and a new passport.

    I haven’t really found a use for it (I’m not competing for a position at a university or eligible for early retirement). But my son may use it to avoid the draft (military conscription).

Comments are closed.