Tim Harford at Slate discusses a new study addressing the question whether coffee shops discriminate against women:
I'm a real cappuccino lover myself, but many of my female colleagues don't seem to go for the stuff. I'd never thought too much about it until recently. I suppose I carelessly assumed that men and women have different tastes, probably as a result of different social influences. Now I know better: My female colleagues don't go to coffee shops because they're shabbily treated when they get there.
That's the conclusion of American economist Caitlin Knowles Myers. She, with her students as research assistants, staked out eight coffee shops (PDF) in the Boston area and watched how long it took men and women to be served. Her conclusion: Men get their coffee 20 seconds earlier than do women. (There is also evidence that blacks wait longer than whites, the young wait longer than the old, and the ugly wait longer than the beautiful. But these effects are statistically not as persuasive.)
Perhaps, says the skeptic, this is because women order froufrou drinks? Up to a point. The researchers found that men are more likely to order simpler drinks. Yet comparing fancy-drink-ordering men with fancy-drink-ordering women, the longer wait for women remained.
It is also hard to attribute the following finding to a female preference for wet-skinny-soy-macchiato with low-carb marshmallows: The delays facing women were larger when the coffee shop staff was all-male and almost vanished when the servers were all-female.
I haven't looked at the study itself yet, because I wanted to address something else in Harford's piece, this:
This is an intriguing piece of research because coffee shops appear to be a competitive business, and one thing we economists think we know about discrimination is that competition should tend to erode it.
The idea comes from an article published 50 years ago by economist and Nobel laureate Gary Becker. The reasoning is simple enough: A business that deliberately offers shoddy service or uncompetitive prices to some customers, or that turns down smart minority applicants in favor of less-qualified white male applicants, is throwing money away. If it is a government bureaucracy or a powerful monopolist, that's a loathsome but sustainable choice. But racist or sexist businesses with many competitors are likely to be shut down by the bankruptcy courts long before the human rights lawyers get to them.
Becker's theory is powerful, and there is evidence to back it up. Economists Sandra Black and Elizabeth Brainerd found that the surge in international trade, which has increased competitive pressures in many markets, has reduced the ability of firms to discriminate against women.
It's worth pointing out that Becker's seminal (ovular?) work looked at three types of discrimination, all aimed at the workers in the firm, not at its consumers: discrimination by the owners of firms, discrimination by other worker groups (e.g. whites against black colleagues or men against women) and discrimination by the firm's customers against one type of workers (who might be directly serving them, say).
It's in the context of the owners discriminating against workers that Becker draws the conclusions Harford mentions. The conclusions in the other submodels are less clear-cut. Note, also, that Becker's model has no uncertainty and lack of information. In reality, women at coffee shops may not know that they are being served more slowly, and you can't react to something you are not aware of.
The current study is therefore not an obvious application of Becker's reasoning. It looks at a different type of discrimination, one possibly by workers towards customers, and it's not clear whether the female customers think (or know) that they are served more slowly or not. If they are unaware of this the firm has no real incentive to fix any problems.