The Visible Hand
STEVEN D. LEVITT
06/01/2010 | 11:00 am
Let’s say you were in the market for an iPod and wanted to find a bargain, so you searched in a local online market like Craigslist to find one.? Would it matter to you whether, in the photograph of the unopened iPod, the person holding the iPod (all you can see is their hand and wrist) was black or white?? What if the hand holding the iPod had a visible tattoo?
I suspect that most people would say that the skin color of the iPod holder wouldn’t matter to them.? More people likely would say the tattoo might keep them from responding to the ad.
Economists have never liked to rely on what people say, however.? We believe that actions speak louder than words.? And actions certainly do speak loudly in some new research carried out by economists Jennifer Doleac and Luke Stein.? Over the course of a year, they placed hundreds of ads in local online markets, randomly altering whether the hand holding an iPod for sale was black, white, or white with a big tattoo.? Here is what they found:
Black sellers do worse than white sellers on a variety of market outcome measures: they receive 13% fewer responses and 17% fewer offers. These effects are strongest in the Northeast, and are similar in magnitude to those associated with the display of a wrist tattoo. Conditional on receiving at least one offer, black sellers also receive 2-4% lower offers, despite the selfselected-and presumably less biased-pool of buyers. In addition, buyers corresponding with black sellers exhibit lower trust: they are 17% less likely to include their name in e-mails, 44% less likely to accept delivery by mail, and 56% more likely to express concern about making a long-distance payment. We find evidence that black sellers suffer particularly poor outcomes in thin markets; it appears that discrimination may not “survive” in the presence of significant competition among buyers. Furthermore, black sellers do worst in the most racially isolated markets and markets with high property crime rates, suggesting a role for statistical discrimination in explaining the disparity.
So what can you conclude from this study?? The clearest result is that if you want to sell something online, whether you are black or white, find someone white to put in the picture.? I suppose you could say that advertisers figured this out long ago, and actually go one step further, making sure the white person is also a good looking blond woman.
It is much harder, in this sort of setting, to figure out why buyers treat black and white sellers differently. As the authors note, there are two leading theories of discrimination: animus and statistical discrimination.? By animus, economists mean that buyers don’t want to buy from a black seller, even if the outcome of the transaction will be identical.? Buyers wouldn’t like black sellers, even if black sellers provided exactly the same quality as white sellers.? With statistical discrimination, on the other hand, the black hand is serving as a proxy for some sort of negative: a higher likelihood of being ripped off, a good more likely to have been stolen, or maybe a seller who lives very far away so that it will be too much trouble to meet in person to do the deal.
The most impressive part of this paper by Doleac and Stein is their attempt to distinguish between these two competing explanations: animus versus statistical discrimination.? How do they do it?? One thing they do is to vary the quality of the advertisement.? If the ad is really high quality, the authors conjecture, maybe that provides a signal that could trump the statistical discrimination motive for not buying from the black seller.? It turns out that ad quality does not matter much for the racial outcomes, but possibly this is because the quality difference across the ads isn’t great enough to matter. The authors also explore the impact of living in an area with more or less concentrated markets, and also across places with high and low property crime.? Black sellers do especially bad in high crime cities, which the authors interpret as evidence that it is statistical discrimination at work.
I really like this research a lot.? It is an example of what economists call a “natural field experiment,” which has the best of what lab experiments have to offer (true randomization) but with the realism that comes from observing people in actual markets, and with the research subjects unaware they are being analyzed.