|Print | Back||May 6, 2014|
The Dismal ScienceRidiculous Averaging
by Adam Smith
It is really difficult to sit and watch the television these days. So many of the contemporary shows have so much vulgarity that I do not enjoy them. That leaves me to turn to the news channels to see what people are talking about.
And what do I see? Our President talking about the pay disparity between men and Ö that that is how far I got before I went to see if there was actually history being shown on the History Channel (it should actually be called the Pawn Shop Channel).
Why, why, why, are we fourteen years into the 21st Century talking about the disparity between the pay between men and women? Even the The Washington Post thinks what the President was asserting was not accurate.
Here are the two sides of this issue. The President says that the average womenís salary is 77% of the average manís salary. This is a fact. The research done on the issue (you know, the science the President likes to say he endorses) indicates that once life differences (chosen profession, taking time off to have babies, etc.) are taken into account, then the difference is supposed to be 4%, not 23%.
People hear that last statement and kind of understand what it is saying, but what does it really mean? How do the economists know?
This article is going to, at a very high level, try to explain what economists do to understand the pay differences between men and women. Just a quick warning, this article is going to get very nerdy in a hurry.
Economist rarely talk about averages. They mean nothing. For example, if I told you that the average pay for someone in New York City was 300% higher than someone in Cozad, Nebraska (actual place we always stayed at when travelling across the country) would you be surprised? Not likely. Why, because you would immediately think of several rather important reasons for the disparity in wages.
So what do economists do? They do something they call econometrics. Using regression models, they try to find relationships between a dependent variable and several independent variables.
Reading the word ďregressionĒ is not a reason for your eyes to glaze over. All regression is doing is seeing how the dependent variable changes over time as the independent variables change.
In our example between New York and Cozad, the salary of the people is the dependent variable. What the different people get paid is determined by other independent variables. What would those independent variables be? Letís just pick two: level of education and chosen profession.
So our model (this is a terrible model but I am trying to make it easy to understand the concept not make the reader capable of doing econometrics) would look like the following:
Amount of Pay (Y) = (X1)level of education + (X2)chosen profession + e
We would start building our data by putting the information about individuals from New York and Cozad into the model. After the data is built, then we run our regression.
The regression is going to tell us a couple of important things. It will tell us whether our independent variables (education and profession) help explain differences in our dependent variable (wages).
When the independent variable is statistically significant then the answer is yes, the independent variable does help explain some of the change in the dependent variable. If the independent variable is not statistically significant, then it is not explanatory.
We learn more from the regression. The coefficients X1 and X2 tell us how much the independent variable changes the dependent variable. So the coefficient X1 would tell us how much pay changes as the level of education changes, and so forth.
The e on the end is the amount of change in pay not explained by our independent variables.
Letís get back the Presidentís statement. An average womenís salary is 77% of a manís salary. Based on our now expanded understanding of econometrics, what do we know about this statement?
The word ďaverageĒ is a red flag. When the word is used, it usually means somebody is telling you something that is entirely meaningless.
The statement does not hold up under the scrutiny of econometrics. Using the pay differential as the dependent variable, economists have used independent variables that have been shown to be statistically significant to help them determine what causes the pay differential.
Some of these independent variables that explain the pay difference are level of education, types of degrees earned, hours worked per week and time taken from careers for child rearing.
Since the pay gap can be explained by other influences, the charge of wide spread, systematic discrimination is not true.
There was a statement made by someone wondering why they always trotted out men to make the claim that there were other items influencing the pay differential. Where were the women economists? This is hard for some people to understand, but not all academic disciplines are about feelings.
The computer running the regression to determine what causes the pay difference does not care whether it is a woman or a man.
My assumption is that women economists are working on questions far more interesting than an issue to which 96% of the answer has already been determined.
|Copyright © 2021 by Adam Smith||Printed from NauvooTimes.com|