Quote:Quote:
Someone on the forum did a fairly detailed statistical analysis of the performance of female run companies. It was pretty negative.
I think that was the regression analysis i did on the effect of the % of women on revenue per a person. Let me go dig it out.
Edit:Found it
Source:
http://archive.fortune.com/magazines/for...012/women/
These are the best 100 companies to work for and have a range of women from 95% to 10%. We can assume that this is as good as it gets for revenue per an employee and should probably dip after this list. A few companies didn't have any info for revenue and/or % of women so i removed them. I would have preferred profitability per an employee since it would give an idea of costs; but you work with what you got and I don't want to go through nearly a 100 financial reports. So the measure i used was revenue per an employee. Revenue per an employee was used since if there are staff that aren't doing anything, it should reduce as you have more non-revenue generating employees.
I would say industry does explain a lot of the revenue per an employee. I do recognize that different industries have different cost structures hence why i wanted profit and because women flock to certain industries. Health was a big one in the sample. If I had more time, I would do an industry by industry comparison but the numbers are good enough to get the idea there is a trend going on even after removing industry from the model. An R^2 of 20% which is about what your SAT is at for assessing your intelligence. It's not a perfect correlation but it's good enough given the data. Also due note, the clearest relationship for drop in revenue is at the highest percentages which might mean there is a sort of "cattiness" threshold.
Anyway the main takeaway is for every absolute percentage point increase of women in the workforce, the revenue per an employee drops by 2.45% which means if you have a 100% female workforce you could be potentially making 92% less revenue per an employee than a full male workforce.
General Regression Analysis: log10_rev versus % women
Regression Equation
log10_rev = 6.16215 - 1.09753 % women
96 cases used
Coefficients
Term Coef SE Coef T P
Constant 6.16215 0.110912 55.5589 0.000002
% women -1.09753 0.217357 -5.0495 0.000002
Summary of Model
S = 0.398729 R-Sq = 21.34% R-Sq(adj) = 20.50%
PRESS = 15.5224 R-Sq(pred) = 18.30%