By Chris | November 12, 2007
I’m always a little suspicious of research studies looking for discrimination in a competitive market. So, when I read Tim Hartford’s piece over at Slate examining research out of Middlebury College on whether coffee shops discriminate against women, I wanted to take a look at the study.
Caitlin Myers from Middlebury College had five undergraduate students stake out coffee shops in the Boston area and take detailed notes on how long different customers had to wait for their order. Women have a tendency of not only ordering fancier drinks than men, but also including special requests with their orders.
“In this case [not accounting for special requests], the positive coeﬃcient for female would be the result not of discrimination but of unobserved order complexity.”
In order to avoid this problem, the authors examined whether male employees might be responsible for longer female wait times. Neither “proportion female employees”, nor “interaction of female customer and proportion female employees” were statistically significant at the 5 percent level (although the coefficients indicated greater discrimination). In the authors’ words:
“In a coﬀee shop with all male employees, a female customer waits
an average of 37 seconds longer for her order than a male customer. However, in
a coﬀee shop with all female employees a female customer’s wait is estimated to
be 7 seconds longer than that of a male, a diﬀerential which is not statistically
signiﬁcantly diﬀerent from zero. Although the coeﬃcient on the interaction
term is not signiﬁcant (with a p-value of 0.31), the result is suggestive that it is
not order type but rather some action on the part of employees that is driving
Fair enough. I don’t doubt that male employees take slightly longer than their female counterparts to serve women customers. Why might this be? Myers speculates:
“However, the source of such discrimination is not
clear. One possibility is animus-based discrimination on the part of male coﬀee
shop employees who wish to impose higher costs on female customers. On the
other hand, these longer waits presumably mean more time spent associating
with the waiting women, which may oﬀset any utility gains prejudiced employees
receive from imposing greater costs on female customers. Another possibility is
that rather than reﬂecting ill-will towards female customers, the diﬀerential is
indicative of male servers garnering utility from interacting or being near female
customers. In this case, the diﬀerential reﬂects not a desire to impose a cost on
women but rather to spend time ﬂirting, chatting with, or just being around
In other words, men are out to get women and can exercise their animosity but taking an extra 30 seconds to brew that cappuccino. Or, male employees hoping for a future date, waste women’s valuable time. The study goes so far as to use Massachusetts’ median-wage to estimate the lost earnings due to discrimination: approximately 10 cents per drink.
The study to a certain extent concedes, and I suspect, that extra requests are the root cause of wait differentials. However, going to a coffee shop is about more than just getting your drink as quickly as possible. Talking with the barista is part of the experience, and will increase wait time. Is it really surprising that women might engage in this aspect of the experience more than men do? I have never worked at a traditional coffee shop but I did spend a month working at a snack bar that sold coffee. I found women significantly more chatty than men. It isn’t hard to image that they might be more talkative with male baristas than females. Surely, male baristas aren’t the only ones looking for dates.
The study also acknowledges past research suggesting that women tip less than men. Although, I doubt the difference is large enough to have much impact on service ex ante. I know at least one reader of this blog has significant work experience at a coffee shop. Any thoughts?
Notes: The study found no statistically significant discrimination of blacks. Also, almost exactly 50 percent of the sample was female. Tim Hartford’s anticdotle observation that many of his female colleagues don’t drink coffee doesn’t seem to be representative of Boston’s coffee market.