In my last post I mentioned the concept of correlation vs. causation. Of course you know that causation refers to two things where one has the specific effect of causing the other. A good example of causation is what happens in my house when I open the refrigerator…it causes an immediate stampede of cat and Chihuahua to see who can be the first to intercept any cheese bits that fall to the ground.
In contrast, correlation is what happens when two things occur together, but one is not the cause of the other—in other words, things may relate to each other but could well be anything from connected to near coincidence. For instance, when said cheese appears, the Chihuahua might growl at the cat, but frankly, the dog could growl at the cat anytime anyplace, not just in the presence of falling cheese.
It is not uncommon for people to confuse correlation with causation. Hey, Fox News has built an entire network out of this. But I recently (August 4th) read a great venture capital story by Joanna Glasner on PE Hub that highlighted this concept in a particularly provocative way.
Entitled, “High Obesity Correlates with Low Venture Activity,” Glasner reported on a recent Gallup Poll on U.S obesity in order to determine how obesity rates correlate with venture capital investment. (According to Glasner, the report showed a wide disparity in obesity rates state-to-state, ranging from 34% in West Virginia to 19.5% in Colorado, with an average of 26.7% nationwide.)
Her conclusion: the “correlation between obesity and low venture funding was shockingly high. The eight states with the lowest obesity rates saw 100 times more venture funding per capita than the eight states with the highest rates.” The following is an excerpt from the article with Glaser’s supporting data:
The eight states with the highest obesity rates are: West Virginia (34%), Kentucky (32.8), Mississippi (32.7), Arkansas (32), South Carolina (31.1), South Dakota (30.8), Oklahoma (30.7), and Louisiana (30.6%). Here’s how much VC funding companies in those states have raised so far this year, according to Thomson Reuters:
West Virginia: 0
Kentucky: 3.8m
Mississippi: 0
Arkansas: 0
South Carolina: 24.7m
South Dakota: 0
Oklahoma: 3 m
Louisiana: 14m
Total: $45.5 million
Population: 25.5m
VC per capita: $1.78
The eight states with the lowest obesity rates were: Colorado (19.5%), Massachusetts (20.9), Utah (21.3), Hawaii (21.8), Connecticut (22.2), Rhode Island (22.6), California (23.2), New Mexico (23.5). Here’s how much venture capital companies in those states have raised this year:
Colorado: 159m
Massachusetts: 1.84b
Utah: 100.3m
Hawaii: 6.2m
Connecticut: 109m
Rhode Island: 51.6m
California: 8.394b
New Mexico: 20.8m
Total: $10.68 billion
Population: 59.2m
VC per capita: $180m
Clearly there are a lot of data points that would show similar correlations. For example, the states with high obesity rates also tend to be, in general, poorer, more rural, more southern, less coastal, and further from a high concentration of VCs than those with lower rates. And it could be argued that California –which drew close to half of VC funding this year – skews the overall stats.
Glasner states that while she recognizes that the data she cites are largely examples of correlation, not causation, she is yet, “somewhat disturbed by the divergence in funding levels between the two groups.” She goes on to say, ”I’m surprised at the number of states with no companies raising reported venture rounds this year. Is there truly no fast-growth, innovative company worth backing in all of Mississippi, Arkansas and West Virginia? Are these regions simply ignored? Or, once a start-up begins to see traction, does it move headquarters to be closer to a big money center? Moreover, what does it mean that so much of the money spent fostering innovation is clustered in few areas? Will we see further bifurcation in opportunities – with those in the ‘have’ regions getting access to capital and jobs in innovative sectors, and those in the ‘have not’ areas unable to get a foothold in growth industries?”
This article got me to thinking about what other attributes one might find in concentration around Silicon Valley and Route 128 and whether those attributes could be categorized as correlation or causation. In particular, it got me thinking about one of my pet subjects, women in venture capital (or the lack thereof).
As previously reported, women make up fewer than 14% of all venture capitalists and the number appears to be shrinking. Does this suggest that the more mature the venture industry becomes, the less likely women are to seek careers in the field?
In the average venture firm that has any women at all, it is an incredible rarity for such a firm to have more than one woman who holds the rank of partner. Does the presence of one female partner cause a partnership to cease considering additional women for partnership, maybe because the first one fulfilled an invisible, maybe even unintentional, quota? Or is this just correlation? Perhaps there aren’t a large enough number of women seeking jobs as VC partners to generate enough critical mass to get more than one woman into the average firm?
I have a hard time with this one because I teach a class in healthcare venture capital at UC Berkeley’s Haas School of Business and the class is decidedly 50-50 on the male-female front. I realize the sample size is small (about 22-24 students per year for 3 years), but there seem to be just as many female as male students in my class who wish to break into the venture field (even after being forced to listen to me for a semester). What are the factors that cause women to be under-represented in venture capital if they are pouring into the entry funnel at an equal rate? Is there a woman-shaped hole in the side of the funnel where they tumble out unnoticed?
As many venture firms have contracted in size during the recent challenging economic times, there seems to be a disproportionate number of women who are the first ones out the door. Is this causal in that women are the last in so the first out due to having fewer years of tenure? Or is this correlation, with women raising their hands first to leave because they have foreseen the poor market conditions and been the first to identify other satisfactory alternatives?
In any event, if low body mass index is correlated highly with venture capital, mascara appears to be poorly correlated. This extends beyond the venture capital team members and into the pool of entrepreneurs as well, as women-led firms also receive a disproportionately small amount of venture capital investment relative to the whole.
While the current VC financing environment doesn’t lend itself to growth and diversification, many believe we are at the earliest stages of an upswing in the cycle (boy, I hope they’re right—this downswing part is giving me a rash). If that is the case, I sincerely hope that this causes a more diverse crop of new hires to be considered. It would be a shame if the future of our industry is correlated with the status quo.
Looking at returns for venture over the past decade, maybe women are just smart for staying away:) But seriously, this is something we’ve puzzled about, particularly in the healthcare field where, unlike engineering, women are heavily represented.