Gender Pay Gap – a statistical MESS
The White House continues to launch press statements claiming that women earn 78 cents on the dollar compared to men. Relatively the same statistic used 40 years ago. Has nothing changed?
As I have said numerous times in my commentaries and blogs, statistics can easily be manipulated to support a false premise, a very dangerous course of action. What the White House number fails to address in their presentation of discrimination is more important than what it does address.
The White House number is a simple gender average. It covers all jobs, all employment, without factoring education, experience, of industry. In fact, it factors nothing but gender which has always been a skewed number. But it isn’t just the White House arguing these numbers, there are multiple organizations that project a false reality of the discrepancy – and there is a discrepancy. But the answer is that only a small portion can be attributable to actual discrimination while the remainder has to attack all, ALL, relative factors that comprise the whole.
Factors to include: Industry/occupation, education, experience, age, location, attitude, ethnicity, hours per week, and marriage status and children. If women choose careers that typically have a lower pay scale, then there is no discrimination, no gender pay gap. It is a matter of choice. Which is why industry should be the number one factor when creating a statistical analysis. Football players… -0- women, 100% men. The sports world would have to be set aside, taken out of the equation completely. But comparing firemen to newscasters or financial advisors to school teachers has no relativity and completely distorts the picture. Hey, I want to make what Robert DeNiro makes… but that doesn’t mean we have a comparable gender gap in wages reflecting discrimination.
Education is the second factor that would have to be added to the pool of factors. In the real business world, when choosing between a high school graduate and a college graduate, one is going to make more than the other, otherwise we wouldn’t have education.
Location is a factor when creating an analysis. For example, if it were found that in New York state the pay grade was skewed, what if the demographics of city vs rural found a skew in male and female percentages? For example, Colorado is 49.9% female compared to the national rate of 50.8%. Not a huge difference, but when you look at Vail, the percentage changes to 42% female, and Steamboat Springs is 45%. These factors change the skew.
Age, Ethnicity, and Marriage Status and Children. Age relates to possible experience, or it can relate to ‘too old and not worth putting in the effort’. Ethnicity can help distinguish if it is a discriminatory factor or not. Marriage status and children can account for the emotional skew which is to say that many women choose lower paying jobs, or even turn down promotions in anticipation of being more flexible for a home life balance.
Nothing is so simplistic as the White House rhetoric which states: “On average, full-time working women earn just 78 cents for every dollar a man earns. This significant gap is more than a statistic — it has real life consequences. When women, who make up nearly half the workforce, bring home less money each day, it means they have less for the everyday needs of their families, and over a lifetime of work, far less savings for retirement.”
In a study in which hours worked per week and industry/occupation were inserted into the equation as variables, the gap lessened to 91 cents. Imagine factoring in choice, location, education and experience. IS there a gap at all?
In a study done by the American Association of University Women, they found a 7 cent discrepancy in pay among male and female graduates one year out of school, when controlling for occupation and college major. Imagine how much more closed the gap would be if controlling for ALL factors instead of cherry picking a select few?
A report given to the White House by CONSAD Research Group, adjusted for one factor, industry, and found the gap narrowed to 5 cents and 1.2 cents for mining to real estate. The one ‘industry’ not mentioned in the report was ‘government’. A curious omission. I wonder what those statistics would reveal…
The Census Bureau is the source for much of the information, but even on their site they tend to compare apples to oranges. In Colorado, within the same data field they will use 2007 as the latest source AND 2013, AND 2009…because their data is not ‘up to date’. In addition, the numbers don’t add: when giving the total number of businesses in Colorado as 547,770, the percentage of minority and non-minority owned added up to 514,702… If you are not a minority or a non-minority, then what are you? An alien? When given the number owned by men vs women, the numbers added to 424,372. So if a man doesn’t own the business and a woman doesn’t, we are left with animals and children… Huh. Obviously the statistics are less than complete.
It’s like a poll, when the variance is +/- 8% and the result states that candidate A has 40% and candidate B has 34%, the real number may be that A has 32% and B has 42%. The picture changes considerably when considering ALL the relative data.
Is their a gender gap in wages? We don’t really know because no one is willing to actually consider all the factors, adjust for the proper comparisons, and put an emotion factor into the skew. Why? Because, sometimes Truth just doesn’t sell as well.