‘Controlling for’ — a methodological urban legend

6 Dec, 2018 at 18:42 | Posted in Statistics & Econometrics | 9 Comments

Trying to reduce the risk of having established only ‘spurious relations’ when dealing with observational data, statisticians and econometricians standardly add control variables. The hope is that one thereby will be able to make more reliable causal inferences. But — as Keynes showed already back in the 1930s when criticizing statistical-econometric applications of regression analysis — if you do not manage to get hold of all potential confounding factors, the model risks producing estimates of the variable of interest that are even worse than models without any control variables at all. Conclusion: think twice before you simply include ‘control variables’ in your models!

The gender pay gap is a fact that, sad to say, to a non-negligible extent is the result of discrimination. And even though many women are not deliberately discriminated against, but rather self-select into lower-wage jobs, this in no way magically explains away the discrimination gap. As decades of socialization research has shown, women may be ‘structural’ victims of impersonal social mechanisms that in different ways aggrieve them. Wage discrimination is unacceptable. Wage discrimination is a shame.

You see it all the time in studies. “We controlled for…” paperAn example is research around the gender wage gap, which tries to control for so many things that it ends up controlling for the thing it’s trying to measure. As my colleague Matt Yglesias wrote:

“Take hours worked, which is a standard control in some of the more sophisticated wage gap studies. Women tend to work fewer hours than men. If you control for hours worked, then some of the gender wage gap vanishes. As Yglesias wrote, it’s “silly to act like this is just some crazy coincidence. Women work shorter hours because as a society we hold women to a higher standard of housekeeping, and because they tend to be assigned the bulk of childcare responsibilities.”

Controlling for hours worked, in other words, is at least partly controlling for how gender works in our society. It’s controlling for the thing that you’re trying to isolate.

Ezra Klein


  1. So, why do women choose to be discriminated?

    • Do they choose to be discriminated against, or do the jobs available to them start getting paid less

      • No, the salarys in typical female ockupations e.g. care or teaching that is within the public sector has been approxiamtely the same for decades so when the females decided on education the information was availble.

        • In the US, care and education have been dominated by women for at least a century. Further women are often socialized away from the high paying jobs to the traditional women’s occupations, that were better paying until women began to dominate them

          • Many occupations have wages trending lower, in Sweden relative wages in the public sector, teaching and care as compared with the private sector trend lower and as well as compared to enginering, IT, engineering, accounting, finance. It’s probably not causally dependent on gender. Why should it be? Initially when women entered the jobmarket there might have been supply exceding demand preasuring down wages, but thatt situation chenged during the 80’s in Sweden when women and men participated to the same degree.

            Still, I tried to point to some structural problems regarding sevices below and you as well point to inherant female sociocultural values as I do. Or do you mean that ceratin carrear paths are closed to women in the US depending on the power situation? Isn’t it also a fact that the distribution of income in the US get ever more unequal? And also, I hear at least, that college premium as regards salaries tend to drop? So the top employees takes more of the cake?

      • In Sweden the private sector has been leading salary setting in all sectors, the export industries in particular, but salaries tend to adjust in a capitalist economy between sectors. Still, the service sektor such as the public sector have less productivity gains as compared to massproduction of goods and services, so the public sector is squized as possible tax limits were reached in the 1980’s and as GDP per capita is less than 1 % per year since the year 2000 depending on mass imigration and low participation in the job market the possibilities for the public sector to increase wages are very slim.

      • There is no difference in education possibilities and participation in Sweden since the 1970:s regarding females and males. Sweden was egalitarian in this respect back then and participation by women in the job market was huge and the existence of public day care for children was supplied. Taxes was changed from family based to individual. Net income after taxes and public income support made wage differnece net of taxes very similar between al income brackets. In the public sector the proportions of male and female managers are the same. In politics the proportions are the same.

        The salary differences are about 6 % for comparable occupations according to the Swedish authority with arguably the best statistics.

        In my opinion the possibilities on the job market are similar, so discrepances must be sought elsewhere and probably the feminist rethoric neglects the basic female interest in caring especially for their children and this takes stime and energy, but is a joy I presume, but can’t be undone or left to the men.

        I suppose it’s as well a stress factor for women keeping the family going as well as work and probably itäs not all that can cope 100 %. Sick leave for women in age Groups under 50 is twise as high as for men.

        Another factor is probably cultural inkluding both men and women, but particulary the women. The traditional female role probably collide with the demands of working life and adaptation may take some generations so even if the playing field is leveled. It’s not patriarchical structures that devalue women today.

        Personally I think it’s a joint problem for men and women how to manage life and family, but I do believe the feminsit rethoric and problem description in Sweden need to change into a practical analysis. To perform in working life qualifications are needed and most women have that e.g. in the health care services, but not in politics and not, it seems, in universities at some institutions if ypu look on e.g. the Genus studies, but of cource in economics, natural science, technolodgy and such subjects.

        Generally it seems to me that in Sweden, in all areas, there is problems with how responsibility is handed out and followed up on and how rewardsystems ar implemented. The rethoric on disrimination and equal rights, thou they already exists, only Clouds the real problems in Sweden and for some it makes a carear!

        • Sorry about spelling errors, Should have used Word and not this editor.

      • Looking outside the Scandinavian countries the situation is different ofcourse. Africa, the Middle East, Afghanistan, Pakistan, India, South America, South East Asia, It’s a shame how women is treated in some of theses countries and problems regarding treament of women are imported to Sweden as well with the huge immigration from countries outside the EU. Cultural differences doesn’t disappear when you cross a border and get a new passport.

Sorry, the comment form is closed at this time.

Blog at WordPress.com.
Entries and comments feeds.