In December, Pragmatic Institute released the 2006 product manager survey results. At first glance, there appears to be a huge disparity in compensation between male and female product managers. When we look in more detail, the evidence does not support that conclusion.
History of data
We took a cursory look at the 2000-2005 product management survey data back in October, and The Cranky Product Manager has done a similar analysis, including the 2006 results. Unfortunately, we and she have both jumped to a couple conclusions about salary inequality.
Here’s the conclusion we jumped to…
Notice also the unreasonably large gap between blue (female) and maroon (male) overall compensation data.
And today’s interpretation…
On the brink of the year 2007, female product managers in the US and Canada only make 87 cents for every dollar their male counterparts make. For the SAME JOB.
The Cranky PM asked some good questions about the analysis:
So why is this happening? The Cranky Product Manager wants to find out…. Her top-of-the-head hypotheses include the following:
- The women respondents skewed younger or less experienced than the male respondents?
- The women respondents tended to be less technical than the male respondents?
- The women respondents are in worse-paying industries than the male respondents?
- Sexism? (which might be increasing over time?)
The first three hypotheses could potentially be ruled out by slicing and dicing the raw survey data.
And we respond…
Responses by gender and experience
We started by looking at the experience level of the respondents to the survey (435 of 439 respondents specified either maleness or femaleness). We start by filtering to ignore the four respondents that did not provide gender data.
In the survey, two questions were asked about experience:
- How many years have you worked in a technology role?
- How many years have you worked in your current role?
We found that some respondents considered themselves to be non-technical (for example, 15+ years in their current role, 0 years in a technology role). We decided that using the larger of the two responses would give us the best proxy for years of experience.
The following chart shows the number of respondents by answerable experience range in years.
What immediately jumps out at us is that there are far more male respondents at the higher levels of experience. We may be inclined to conclude that there are far fewer female product managers at these experience levels, but we can’t. All we can conclude from the survey data is that more males responded to the survey, not that more males are in the role.
Salary by experience
We would expect salary to rise with experience, although it surprisingly plateaus at the 3-5 year mark.
With just this data – higher salaries at higher experience levels, combined with a higher number of male respondents with more experience – we would expect to see higher salaries for males. And that’s exactly what our cursory examinations showed. What about looking at the “gender gap” when normalized for experience?
Product manager salary by gender and experience
When we split the salary data by experience across gender lines, we see something interesting.
For product managers with more than five years of experience, the women who responded to the survey are out-earning the men. The only large disparity we see in either direction is for people with less than a year of experience (about 1% of our respondents).
Almost 85% of the respondents have 6 or more years of experience, and for those groups, the women have higher compensation than the men.
The soundbite analysis appears to show women having markedly reduced earnings relative to their male peers in 2006.
Looking at the next level of detail, we see that within ranges of comparably experienced product managers, women earn more than men For the experience ranges that represent the bulk of the respondents, women reported about 5% higher earnings than men as product managers. We also saw that product managers with more experience tended to earn more than their neophyte peers, although experience beyond the five-year mark (for our experience proxy) did not seem to have much effect on earnings. These data were obscured by the larger number of male responses within those higher-earner categories.
Although an earnings disparity may exist, the data from this survey does not support the argument.