Is there a Gender Salary Gap in private ventures?
My previous post showed the percentages of men versus women in IT versus Life Sciences executive teams, across various VP-level and C-level positions. In this post, I focus on salary differences by gender, first describing the overall results using our full 2008 survey results, then delving into various subsets of the data.
The regressions included data on 2,200 executives from almost 480 private IT and Life Sciences ventures. The regressions controlled for differences in the following factors:
- Positions — separate variables for each of 15 VP-level and C-level positions
- Company maturity — rounds of financing raised, number of employees
- Industry — IT vs. Life Sciences
- Geography — whether the venture is located in each of the two “hubs” of start-ups (California and Massachusetts) or not
- Executive backgrounds — founder versus non-founder status, years of prior work experience, years working in the venture, educational background (i.e., advanced degrees)
- Executive’s equity holdings — percentage of the venture’s equity held by the executive
(Note: Where appropriate, some variables were log-transformed or square-root transformed.)
The table below summarizes the gender-salary gaps that were statistically significant (p<.05) after controlling for the above factors. The sections that follow describe the results, followed by auxiliary results analyzing whether the gap is different if the CEO is a woman, and analyzing gender differences in equity holdings.
In short, there is indeed a gender salary gap within private ventures, but that gap is greater in IT ventures than in Life Sciences ones, outside the “hubs” of California and Massachusetts, and in later-stage ventures.
It is important to note that, compared to the 20-25% gender wage gap consistently found in studies of large companies, these gender gaps in private ventures are much smaller (less than 10%). Small, private ventures may indeed be a good place to escape the factors that cause wide gender gaps in large companies.
Overall Salary Differences
I first focused on whether there are overall differences in salaries between men and women. This initial regression showed that overall, men made $10,200 more than women (with very high statistical significance of p=.005). The other very significant factors affecting the salary gap were number of employees, founder status, years of work experience, position differences, geography (CA and MA higher than elsewhere), and industry (IT lower than Life Sciences).
IT vs. Life Sciences
I then ran the analyses separately for each of the IT and Life Sciences industries. In IT, there were 1,464 executives from 313 ventures, and in Life Sciences there were 709 executives from 163 ventures.
After controlling for all of the factors listed above, in the IT model men made $13,300 more than women (very high statistical significance of p=.001). In the parallel model for Life Sciences, the difference between the genders was not significant (p=.328).
Geography
The next models separate California and Massachusetts from the rest of the country. For California, there were 630 executives from 146 ventures, for Massachusetts there were 344 executives from 74 ventures, and the rest of the country included 1,199 executives from 249 ventures.
After controlling for all of the factors listed above, in the California model there was no statistically significant difference between the genders (p=.914). In the parallel Massachusetts model, the difference between genders was of borderline significance (p=.06). In the rest of the country — outside California and Massachusetts — men made $16,300 more than women (very highly significant at p=.002).
Stage of Venture
The next models separate early-stage ventures (2 or fewer rounds of financing) from later-stage ventures (3 or more rounds). The early-stage model included 892 executives from 220 ventures, and the later-stage model included 1,281 executives from 246 ventures.
After controlling for all of the factors listed above, in the early-stage model there was no statistically significant difference between the genders (p=.254). In the parallel later-stage model, men made $12,300 more than women (very significant at p=.009).
Any thoughts on these gender differences (and non-differences)?
For instance, why does the gap widen as ventures grow? Why is the gap higher in IT than in Life Sciences?
Auxiliary Analysis #1: Female CEOs vs. Male CEOs
Does it make a difference if the CEO of a venture is a woman? Given the low percentage of female CEOs in these ventures, it’s a little hard to draw definitive conclusions, but the data do suggest that it does make an important difference. I ran separate models for ventures with female CEOs (this model included 92 executives) versus male CEOs (the other 2,019 executives in the dataset).
In ventures with female CEOs, the difference between the genders is not statistically significant (p=.223), though directionally, the model suggests that women make more than men in ventures with female CEOs. In ventures with male CEOs, men make $13,900 more than women (extremely high significance at p=.000).
Auxiliary Analysis #2: Differences in Equity Holdings
Are there also gender differences regarding equity holdings, the other major element of compensation in private ventures? Regression models using the percentage of equity held by each executive (square-root logged) indicate that overall, men do hold more equity than women (high significance at p=.007).
Looking at the different industries, the gender equity difference in IT ventures is not significant (though it’s not too far from significance, at p=.106), and in Life Sciences it is of borderline significance (p=.051). Geographically, the differences mirror the salary differences: no difference in California, some difference in Massachusetts, strong difference outside those hubs. Regarding stage of the venture, there are borderline-significant differences in both early-stage and later-stage ventures, each in favor of men.
Are there differences across the genders that are not captured by the control variables used in the models above? (For instance, are there more part-time executives in later-stage ventures than in early-stage ventures, in IT than in Life Sciences, or in non-hubs than in “hubs”?) Do you have any other explanations for these differences?
I suppose the real headline here is less about the “5% compensation gap” and rather more about the fact that there are so few women in startups!Regarding the growing gender gap as companies mature, the obvious answer would be that early stage companies are meritocracies whereas mature companies are bureaucracies that reward factors other than performance which are biased against women.If you really want to get into compensation bias, the question you ought to ask is, “how tall are you?” I’d bet this year’s salary that taller executives make more money. ;-)
That was indeed the headline for my “Part 1″ post about this, Furqan. :-> Guess we'll have to add a height question to next year's surveys!The answer you suggest above may be the most “obvious” one about the growing gap as ventures mature (though it's not the only possible answer), but it begs the also-obvious follow-up of what causes young meritocracies to so quickly adopt policies that lead to gender gaps.
So why do gender gaps appear as companies mature? I bet it has to do with the fact that most founder/CEOs get replaced by “seasoned executives” many of whom come from larger organizations. And those executives bring not only their “own team” but they also bring their existing preferences and biases. Do you have data to see if the bias exists in mature companies if one or more of the founders remain?
I like the thought, Furqan! It’s consistent with the view of founders as being the culture bearers as the venture grows.I do have data to test that. I just created a flag for whether any founders remained among the C-level executives (likewise for the broader VP-level and C-level top management team, which had similar but weaker results), and added it to the model for the later-stage ventures. Directionally, it showed that the presence of founders did counterbalance the gender gap a bit (reduced it by about one-third), but it was not statistically significant, so I’d be hesitant to draw any definitive conclusions. But quite interesting nonetheless!
I think the issue here is company growth. Many women today, particularly those in their 40′s who are experiencing independence that coincides with older children, are starting companies. It’s large companies that have stood the test of time and have taken on bureauocratic tendencies in which gender gaps surface. Noam’s article in HBR “Planning a Start-Up? Seize the Day” (Jan. 2009) supports the claim that women may be better suited for start-ups because they have not fallen prey to large corporations or the outsourcing that being part of that environment ensues. Women are accustomed to multitasking and what Edward P. Lazear calls bieng “the jack of all trades”.As far as compensation, well, you’re usually “the jack of all trades” because you can’t afford to be anything but.
Hmmm, I guess I can’t rip on the hired gun executives so much any more! Or maybe not.What if you just look at mature companies where the founder remains as CEO and whether that group has a gender gap? I could imagine it being the case where one or more founders remain but they are not in a position to control culture.
That seems fair and wherein lies the gender gap. The question is why? And is there really compensation bias or is it that male executives have been at it longer; or is it that “they’re worth more” ie: seasoned executives who may have moved into a start-up to take it to the next level.
I am a women entrepreneur on my 3rd startup (not including those that only got to the prototype stage). There is a lot of bias against women as entrepreneurs and inventors…I can sense it but I cannot prove it. I suspect if you interviewed the few of us here in SV and really listened, you might be able to figure out exactly what is going on. For instance, I’ve had three ideas (that were at the working prototype stage) stolen by others (one VC firm tossed it to their EIR, another person who claimed to be an investor took the idea for himself, and a colleague absconded with the third idea). The first two ideas/prototypes were turned into actual funded companies by men, one supposedly now has a $1B valuation. While these horror stories happen to ALL entrepreneurs, I suspect they happen MORE to women entrepreneurs, partly because it is harder for them to get the funding to go past the seed/prototype stage. Suggestion - contact Adeo at TheFunded.com and ask for help in contacting women CEO entrepreneurs. He probably knows all (5?) of us.
Great analyses, Noam. An interesting counter trend to disparity in compensation/power/status is here in wicked old Hollywood. Dreamworks, Universal and Sony/Columbia all have installed women in top executive positions now. Both Variety and the Hollywood Reporter have run pieces in the last year noting that trend. It’s unclear whether the unusual, parochial and sometimes byzantine relationship-dependent dynamics of Hollywood are any kind of paradigm compared to tech/IT/life sciences. For example, in general, the entertainment industry’s social milieu is sometimes pegged as “…junior high school with money.”An ironic, waggish litmus strip for the difference between H’wood and Sand Hill Road is that one end arose from vaudeville and schlockmeisters and the other from government-financed mad bombers. Which end of California will ultimately prove to be the more equitable — and trend setting — in paying the ladies has yet to be determined.
How do you differentiate the source of this gap between the start up effect and nation-wide bias? I think it wuld help to <>compare this gap to that of national average<> of Fortune 500 potentially. Also, it is difficult to put significance to a number (10 000$) and the <>percentage gap<> would put perspective. Thanks for the interesting results as usual, and looking forward to more!
Thank you for all of the comments regarding these results!I fully agree with Kerry that the key question is <>why<> the gap exists. I think the results showing where the gap exists and doesn’t exist (more in IT than in Life Sciences, more in non-hubs than in hubs, more in later-stage ventures than in early-stage ones) are key to our being able to come up with some answers. Furqan has taken a good shot at tackling the later-stage/early-stage question; <>does anyone have thoughts on the non-hub and IT results? For instance, are the “biases” (mentioned by several people) different in those different areas/industries/stages?<>Jim H: As I mentioned towards the top of the post, this gap is far smaller than the national/large-company one (“less than 10%” in contrast to “20-25%” in large companies). In case it helps you translate into percentages the specific dollar gaps I list, the average salary across these 2200 executives was $183,700. (Thus, the all-ventures gap of $10,200 is 5.6% of this average, the IT gap is 7.2%, the non-hub gap is 8.9%, and the later-stage gap is 6.7%.) The more detailed averages in those specific buckets vary a little, but those percentages give a pretty good idea of what the gap-percentages are.
I wonder if the growing gender gap as companies mature is correlated with companies taking on venture capital. Venture capitalists tend to be male, especially in IT investing, as opposed to health-care related sectors, and view male CEOs as more likely to be risk-takers, and aggressive deal-doers and less likely to need to skip work to cover a family need. Venture capitalists have a lot of influence on CEO choices since they typically take board seats as a condition of investment and sometimes insist on a CEO change at the time of their initial investment or a subsequent round of funding.
We saw similar gender gaps in our analysis, specifically in the product management roles in technology vendors. Back in 2000 when I did my first survey on the topic, I neglected to ask the question and in 2001 I was frankly surprised to find a delta between genders. The gender gap is reduced somewhat when you incorporate experience and technical savvy. All of our surveys are available free at http://www.pragmaticmarketing.com/survey
When the Harvard Business blog picked up on this post (see < HREF="http://blogs.bnet.com/harvard/?p=772&tag=nl.e713" REL="nofollow">here<>), Jeanine Boderic contributed the following comment:<>Women need to read the book “Women Don’t Ask”. Much of the difference is because women do not negotiate as well as they could upon hiring and when other opportunities present themselves.Learning to negotiate better would show that many times the pay difference is not discriminatory but because the men tend to counter employment offers whereas women (even ones who know they should negotiate) seem happy to be “wanted” and are fearful negotiating might result in the offer being taken off the table.<>
This post has been removed by the author.
This is a really interesting blog. It's a shame you haven't posted much lately.
@ BV: I totally agree.
English lessons online
Really interesting information. It's not really a secret, but it's still frustrating to see. Thanks for posting, hope to see more from you.
-Kate