While Huey Lewis may croon about "The Power of Love", I choose to praise "The Power of the Internet" today. 

Case in point: On Friday, the Hartford Courant came out with a provocatively titled article: "Gender Wage Gap in Connecticut  is Higher Than National Average."  The article raised a lot more questions than it answered and it’s through the power of the Internet that we can dig deeper. 

And so, with all the talk about the Paycheck Fairness Act this week, you’d probably think that the article concludes that gender discrimination is the main culprit in this statistical disparity.  

And you’d be wrong.

Indeed, according to a statistician for the U.S. Census, employment discrimination isn’t a factor:

Jennifer Day, a statistician with the U.S. Census, the source for all this data, detailed the gaps within jobs. The lingering effects of employment discrimination 35 and 40 years ago isn’t the answer — the proportion of women among the older workers in a job did not determine the relative pay, she found.

So, if employment discrimination isn’t the answer, what is? For that, I sent a tweet to Stephanie Thomas, Ph.D., who runs The Proactive Employer blog in her spare time as an economic and statistical expert.  Stephanie has written before about these statistics and why using these numbers to support the Paycheck Fairness Act is inappropriate. To Stephanie, it’s an issue of using numbers for advocacy without understanding their meaning.

Stephanie posted her response to my tweet this morning.  As she notes, the pay gap issue isn’t as simple as comparing "the typical earnings of men and women and subtracting." Multiple factors are at work, including "industry, occupation, education, work experience, union status, hours worked, and the choices made by individuals…." 

So where might  the answer lie? As she notes, "The answer to Connecticut’s larger-than-typical gender wage gap may lie in its larger-than-typical gap between the rich and poor. As noted in the article, ‘Famously, the state has more than its share of super-earners — and they tend to be men.’"

Stephanie goes on to give a great example explaining the differences between "averages" and "medians" and why and how the statistics can get manipulated.  It’s definitely worth a read.

What conclusions should we reach about this? Numbers themselves may not tell an entire story. Thus, the next time you hear about the gender wage gap use the power of the Internet to get a complete story.