Chris Blattman

Close this search box.

Does it matter if you’re black on the Internet?

Sadly, yes. A field experiment involving baseball card auctions on eBay:

Photographs showed the cards held by either a dark-skinned/African-American hand or a light-skinned/Caucasian hand.

Cards held by African-American sellers sold for approximately 20% ($0.90) less than cards held by Caucasian sellers, and the race effect was more pronounced in sales of minority player cards.

A new paper by Ayres, Banaji, and Jolls.

51 Responses

  1. I follow international soccer, and last World Cup saw the individually-talented, and majority black, French team dumped out.early. One of their players, Yohan Gourcuff, one of the most creative playmaking midfielders in the game today, was routinely excluded from possession by the other players. It was really clear from his gestures and demeanor toward the other players that he knew his teammates were deliberately not passing to him due to his race. there was too much rumor and evidence from game tapes to conclude anything else.

    Again,I’m not emotionally upset about this bias, but that is what it is. And I have experienced the same treatment, so, it’s not isolated.

  2. I just read a week ago that a Harvard study found white americans consider anti-white bias to be stronger than anti-black bias. many whites even rated anti-white bias at the maximum, 10 out of 10.

    So embarrassing to be a white male american. I feel like i can never apologize enough for the other idiots of my ethnicity.

    1. @Diego and EH

      When surveyed, about who they would trust more in a business deal, Americans of all races preferred whites to blacks. From what I remember, the findings held steady over several years.

      The Ebay example offers a test of the law of one price. If you have identical goods with identical availability and identical risk of faultiness, then over time you’re going to have them converge to one price. Otherwise, you have arbitrage opportunities. So, if you’re really sure that the average cards offered by blacks on the internet are equivalent of whites then buy at the black price and sell at the white one.

      I’m not going to pay for access, but I wonder if the researchers controlled for the seller ratings and number of sales by seller. A little secret of the NFL is that 8 of the top 10 jerseys purchased by white men is of a white player, and 9 of 10 by white women is the same case. On top of that, the one black player in the top 10 for women, one of two for men, is the very light-skinned Hines Ward.

      That white and non-white players tend to sell to different demographics combined with perceived differences in seller trustworthiness/quality account for the differences.

      Okay, so let me ask the question: is there some mechanism to distinguish between trustworthy and untrustworthy blacks?

      Why is it that such a clear and perfectly logical question elicits such hysterically hand-wringing responses. I get the distinct impression that if Diego and Eh watched the neighbor’s cat get hit by a car they would immediately pee their pants and start sucking their thumbs.

    2. whites from working class backgrounds are probably right. the main form of discrimination is against poorer whites in favor of all other demographics. Even when controlling for SES, income and education race was a very strong predictor of crime rates, far stronger than those three.

      I remember reading how the city of Knoxville (at least one of the big Tennessee cities) made a concerted effort to shift section 8 housing into the suburbs, on the theory that concentrating a bunch of poor people, disproportionately unemployed, created a snowball effect. They shifted the section 8 housing into middle and working class suburbs and the crime rate in those suburbs shot up. Of course,the well off neighborhoods were largely unaffected.

      Bias is a pretty subjective experience, so, we’d have to know a lot more of what they meant by that. If you concentrate bias toward a small section of a population then that section is likely to perceive that bias as generalized. Not sure what there is to apologize for.

      There are many areas of the country where it is completely infeasible to raise children, if you are white. I doubt there are many such areas you could say that same for black people.

      Here’s an interesting method of measuring bias. In the NBA there is a stat called win-share, IIRC, which measures the value-added a player brings to the floor. I believe that something like 9 of the top ten white players are foreign-raised. Despite the fact that basketball is clearly an American sport. The simplest explanation is that white players are systematically discriminated against during the formative playing years, so that when it comes to the big-time there are not suitable white prospects from the US.

      Who knows about the source of that bias, but I suspect it has two sources:

      A) Interpersonal social intimidation of whites by blacks.
      B) Differences in the rate of physical maturation/black kids being held back which affects relative physical maturity

      I can personally attest to the first. I grew up in a neighborhood that was majority working class, older white people with a bunch of low income housing that was pretty much black single mother families. By twelve, I was well familiar with the refrain “brotha and a white gettin’ in a fight, brotha’ don’t win, we all jump in”. This is not a complaint, just an observation that black guys are allowed to get away with stuff that white guys wouldn’t even consider trying. Spots on a basketball squad are scarce and confer social status, so young men compete for them, and if they’re allowed to compete in such extracurricular ways it is likely that over time they will.

      I don’t know if you’re aware of the recent increase in what are called “flash mobs” perpetrated by, almost exclusively, blacks in their teens and early 20s. I’m used to that behavior on a smaller scale, so, it’s pretty much just par for the course to me. A convenience stop opened kitty corner from the nearest high school and ended up closing down within a year because dozens of black kids would suddenly show up and nearly clean the place out on a regular basis.

      1. “This is not a complaint, just an observation that black guys are allowed to get away with stuff that white guys wouldn’t even consider trying.”

        You wouldn’t consider trying it out of fear of discomfort or lack of resources, the same factors that keep you from doing most things (everything you’re not doing right now, for instance). In this instance, your inability to gather a group of people to carry out an act of violence is due to lack of such a group and also fear of physical harm. Perhaps you’re not as well-motivated to carry out violence (small favors), so you’re not doing it because you simply don’t want to.

        I can’t think of a case in which blacks (especially men) can commit wrongdoing with fewer consequences than whites. Blacks are incarcerated at a much higher rate than whites, and disproportionately to their crimes. The same social mores that apply to you apply to blacks, they inhabit white society for most of the day. Amongst themselves, they adhere to some other, traditionally black cultural norms that you’re probably not even aware of, limiting their behavior further.

        Ever ride along with a black man in the same car? LEOs call it “salt-and-pepper” and target the car to be pulled over. Have you ever experienced a fearful reaction from a stranger after dark? It’s a trope in black life, blacks barely bother discussing it. Blacks are asked point-blank how they paid for their possessions by complete strangers, something that probably seems bizarre to you. I know black _lawyers_ driving $60K luxury cars that have had to leave a gas station because the woman at the next pump was frightened and could not calm down, threatened with arrest for getting gas at night.

        Even if the prejudice is justified statistically, taking responsibility for racial equality means you address the problems plaguing the black community, because the color of a person’s skin is impossible to change. The civil rights movement is the story of removing one after another barrier to success for minorities, until there are no legal ones left. That doesn’t mean we don’t have to deal with the consequences of 400 years of slavery and violent, systematic, authorized oppression. We can either deal with it on the backside, just locking black men up, keeping black families working at a subsistence level, not caring to grow the number of blacks completing higher education, ignoring ethnic health problems, failing to insist on clean, healthy food and water in black communities (yes fucking water, google it), figuring out a way to deny the results of studies like this. It’s just going to take a lot longer for blacks to achieve equality this way, which means more crime, but most victims of black criminals are other blacks, so who cares?

        You have to decide that for you, racial equality has moral merit. That it’s important to who we claim to be as Americans. You have to put your pragmatism aside to do good, just like you do when you take valuable time to vote, or send someone a card, or buy a round of drinks.

        Black people don’t sit around and bitch about the 20th century’s lineup of evil assholes, almost none of whom were black. I would be willing to bet a paycheck that the decisionmakers behind the recent depression were entirely white. You suffer at the hands of your white boss (practically all management is white) and every white asshole boss you’ve ever had. Lawyers: white. Cops: mostly white. Landlords, bankers, meth addicts, fraudsters, politicians, identity thieves, even the guys that write viruses and spyware are almost entirely white! I heard this one white lady even killed her own kid and got away with it!

        How come when you meet a white person you don’t think “asshole boss”, “serial murderer” or “violent separatist”? Why doesn’t your so-called reasoned prejudice against one race of people also apply to another, the one that you are coincidentally a part of?

  3. If you have two groups and one group produces more incidences of failed trust than the other then it is appropriate to label the former as likely to produce more instances of failed trust. In lay terms, they’d be described as “less trustworthy”, although there are more accurate ways of describing the differences. The evidence is pretty incontrovertible that any randomly selected interaction with a black person is statistically more likely to result in an outcome of failed trust than any randomly selected interaction with a white person.

    Are you serious? What do you mean the “evidence” is incontrovertible? What are you talking about. This is a joke, right?

  4. I just happened upon your blog by chance. I would like to say that this is one of the most sophisticated and hyperbole free discussions of the racial dilemmas we face in our country.

  5. Oh yes, it’s all so very horrible… Most naughtily unjust…! So…? What exactly is to be done about this obnoxious truth? This is a stupid complaint, larded with self-conscious adjectives and adverbs… If anything, this is a study in candor–a poll, if you will–indicating the genuine inner-psyche of the respondent without passing it through the subconscious prism of political correctness… which begs the question: What’s worse, an overt bigot or a faker…?

  6. What i find interesting is zeal to justify race differentiating but the fear to call it what it is…racism. Today in America, there is no racism…unless you’re Black.

    1. The term “racism” no longer means anything. Like the terms “evil” or “Nazi” it has been used to describe such a staggeringly diverse array of phenomena that it is functionally meaningless. A term that describes everything describes nothing. Consider a double-blind study where interactions between two people are put into dialogue format on paper. Now a separate group evaluates those interactions and decides whether the first person is “nice” to the second person. Now, let’s say that we divide the first group into blacks and whites but that distinction is unavailable to the group evaluating for niceness.

      Now, let’s say that the evaluating group, comprised of a random sample, describes white people as being “nicer” to black people, than black people are to black people. Having grown up in a very diverse environment I will submit that white people are, in fact, quite a bit nicer and more respectful of black people than black people are of black people. That would be evidence against the hypothesis of so-called racism. Of course, to a committed ideologist such evidence would be immaterial.

      Anyways, the point is that charges of “racism” are not about facts but about the ideological fervor of the one producing the charge.

      Words like “racism” and “evil” that are used to describe everything end up describing nothing.

    2. Also, your use of the term “zeal” is just flat out weird. The first synonym I found for zeal was “passion”. However, you might want to describe my interest in a mechanism to distinguish between untrustworthy and trustworthy black people, “passion” is clearly a poor fit.

      Frankly, your inappropriate use of the term zeal would rather indicate that you are the one operating from passion, not me.

  7. The defense of racism in this thread is astounding. “Maybe black people are just less trustworthy” is not an adequate answer because a) when speaking about a group as large as African Americans in the United States trying to judge something as subjective as “trustworthiness” is impossible. Whatever anecdotal experience you’ve had with black people is irrelevant. Blaming those that are discriminated against is a simple transfer of guilt and is easily seen in this thread.

    1. Um, dude if your position is so strong why the need to offer straw man arguments. Yes, black people are diverse. Some are untrustworthy, some trustworthy. Now, let’s say we randomly select a thousand men and a thousand women from the general population and measure their heights. Now, if you hear someone remark that “men are taller than women” we all know that we’re hearing a description of statistical averages. Let’s say that you are running a soup kitchen and want a volunteer who can reach things on the top shelf. A rational person would discriminate against women, if they needed to make a selection without knowledge of a particular individual’s height.

      If you have two groups and one group produces more incidences of failed trust than the other then it is appropriate to label the former as likely to produce more instances of failed trust. In lay terms, they’d be described as “less trustworthy”, although there are more accurate ways of describing the differences. The evidence is pretty incontrovertible that any randomly selected interaction with a black person is statistically more likely to result in an outcome of failed trust than any randomly selected interaction with a white person.

      Again, the evidence is not even debatable, really. Which brings me back to the question of how to find a mechanism of distinguishing the majority of trustworthy black individuals from the minority of untrustworthy ones.

      I didn’t realize that the very basic concept of statistics was so difficult to understand. I mean, if I said that “the average male is taller than the average female” people interpret that as “all men are taller than all women”? Are people really that ignorant of basic statistics?

      Also, I am not implying any specific causes of differences in trustworthiness. You can blame economic disparities, and that may be the primary explanation.

      People seem to confuse descriptive explanation with prescriptive justifications.

      1. It’s remarkable to me that you talk about “incontrovertible evidence” and accuse people of being “ignorant of basic statistics” without even ATTEMPTING to provide any of this evidence. Because it doesn’t exist. The best you can do is make a false comparison between height of the sexes, a quantitative thing that is easily measured. The idea that “trustworthiness” can be measured is flat-out laughable.

        So uhh… my condolences on your black niece and nephew? Good luck proving to people “Yeah, they’re black. But don’t worry, they’re not THAT black” and then trying to justify not being racist.

      2. We’ve decided that in America, we are not going to tolerate racism.

        That means that when studies like this come out, we don’t say “maybe it’s because blacks are, on average, worse people”, because the study shows that you already think that. A study to find that black people are, on average, worse people, by any objective measure has already been done, it’s called “The United States Census”.

        Discussing whether or not blacks are, on average, poorer, less educated, more likely to be perpetrators (or especially victims) of violent crime, more likely to spend time incarcerated, more likely to be raised by a non-parental guardian, more likely to talk during movies or be good at sports should be out of your system by now. Grow up. The “I’m just sayin” defense is as old as the hills, nobody’s fooled.

    2. No kidding.

      Every single person who has ever tried to rip me off on eBay (maybe 5-6 times) has been white.

      Mistrust of blacks in this country is part of this country’s continuing racism. It’s why black people get harsher sentences for similar crimes, and are less likely to receive consideration for jobs that have white applicants.

      Any white person who cannot see this is simply not paying attention.

      1. Awhile ago there was a stark example of how the justice system appeared to be discriminating against black drivers on the New Jersey turnpike. Studies clearly showed taht blacks were not any more likely to speed than whites, yet blacks were stopped at much higher rates. Discrimination? No, blacks who were speeding were speeding at an average of more than ten mph faster than the average white speeder. Once corrected for mph over the speed limit the difference disappeared.

        The main source of the disparities is that the average black arrestee has a far more extensive criminal background, more likely to have outstanding warrants, etc. Where is your evidence that blacks and whites commit identical crimes with identical contexts and results but receive different punishments. I know that charge gets thrown around but do you have any real evidence? Reports of race status by victim group is identical to arrest records. Roughly 60 percent of all robbery victims report being robber by a black person, while 58 percent of all arrestees for robbery are black. And the numbers hold across race and category of crime.

        From everything I’ve seen, blacks and whites simply do not commit the same crimes, as there are often aggravating outcomes that increase the severity of the crime. In fact, in crime categories where law enforcement has the greatest leeway, such as public intoxication, in determining punishment outcomes there is the least outcome disparity.

  8. This is a very real concern for me. My sister is married to a black guy. My niece and nephew are “black” (ugh, I hate that one drop rule). The guy is one of the most trustworthy dudes I’ve ever met (although he’s a little nebbish). I have little doubt their kids will be any less statistically trustworthy than my other nieces and nephews who have zero black ancestry.

    So, I’m looking for a mechanism to distinguish my trustworthy black niece and nephew from black people who are untrustworthy.

    1. Is this serious? You’re looking for a mechanism to distinguish between trustworthy and untrustworthy blacks? Geezus.

      1. Sure, it would just be a particular example of mechanisms we use to distinguish between trustworthy and untrustworthy people, in general. How is this even controversial or debatable?

    2. There is an established protocol in codifying trustworthiness that is used throughout the Western world. It gives strangers the opportunity to engage in social interaction, which is a necessity for modern society (think of it as being on the tech tree in Sid Meyer’s Civilization).

      I can’t believe I’m explaining this, but there are two totally different paradigms for establishing trust: in an anonymous transaction (like this one right now), and an identified transaction, like getting a job. Admittedly, trust without identity is a relevant problem, but mechanisms have sprung up to establish trust anonymously: wisdom of the crowd (ratings, reviews), an third-party authority that has the identity of the trustee, but does not share it with you (verisign, paypal), or even no-trust transactions (most illegal drug deals are carried out this way, every business deal was like this until about 100 years ago, The Silk Road (google it)).

      Trust with identity is a done deal. Credit ratings, background checks, security clearance, all very well-documented and commonly understood. But I’m thinking you don’t want a mechanism for establishing verifiable trust, I think you want a mechanism for having _feelings_ of trust, which is a job best left to a mental health professional.

  9. Occam’s razor sez … rational discrimination. Black folk are, ON AVERAGE, less trustworthy. Next question … can we provide some mechanism to distinguish between trustworthy and untrustworthy black people?

    Why does no one ever ask that question?

  10. This is the “statistical discrimination” vs. “animus” question — and a lot of studies do try very hard to sort this out, but it’s obviously difficult. (In the iPod study, they find that black sellers do worse in markets with high property crime rates, for instance, implying that buyers might be using race as a proxy for the probability of being mugged or the iPod being stolen. They also analyze the text of the emails buyers send and find that those buying from black sellers seem less trusting.)

    Also, the fact that different types of buyers might be attracted to different types of sellers doesn’t *really* matter, from the seller’s point of view. The black seller still winds up making less money, even if it’s not because the buyers are actively “racist” in the way we typically imagine. This has policy implications, but I can’t imagine the buyer cares.

    1. That is an interesting point about crime rates affecting purchasing habits. I want to see that they have the box and the full set of accessory items when they have something for sale because I do not want to buy anything that has been stolen. Same reason I will not go to pawn shops! Of course, a lot of people don’t have the same compunction and just care about price.
      Are you sure that people don’t choose to lowball sellers in high property crime markets because they assume the goods are stolen? If someone steals something then the entire take is profit. So even if I buy an ipod for 100 dollars, that is pure profit for the seller. If someone is selling their own used ipod, then they are less likely to accept a lowball offer because they still feel like they paid 400 for that ipod last year and they are more emotionally invested in getting a closer return on their investment. I am a very rare buyer on either ebay or craigs list- more of an etsy girl myself. But I never notice where the seller is until it comes time to check out shipping costs. I also don’t usually buy from someone who is holding anything. It looks amateurish. To me the focus should be on the product. Have a little pride in your production value!

  11. Very few, if any, studies of the “THAT’S RACIST” variety go the next step and verify whether or not the bias is justified. Is a purchase from a black the same as from a white? What sort of differences may there be between the two other than “trustworthiness”? Just off the top of my head I can think of a few. For the average (ie, not black) customer there is a greater likelihood of a communication barrier. Obviously there is also a cultural barrier. It is hard to deny that for any race on average one feels more comfortable interacting within their own cultural and racial borders than dealing with those outside.

    The glaring question of whether or not a black man is more trustworthy *on average* than a white one isn’t answered either. Can you justifiably be upset if what you’re complaining about is ultimately true?

  12. Plus, who was on the other side? If a certain kind of people are attracted toward black people, who tend to bid lower for whatever reason, that could also explain this.

  13. black people are poorer on average, if people trading incorporate this information, offering a lower price may simply be Bayesian updating.

  14. Two questions:

    1. Was the race of buyers known and, if so, did different races show different patterns or were black people as likely to shy away from black sellers as whites, Hispanics and Asians were?

    2. Has anyone studied whether black e-vendors do or do not cheat customers more often? I’m not sure it’s relevant because even if such a studied showed that black vendors did cheat customers so much more frequently that the bias was mathematically justified, the buyers would not know of the results of such studies. Still, it would provide some evidence on whether their biases were justified by reality or not. If there was proof that the bias was unjustified, it would make the study stronger.

  15. I find these studies fascinating and disturbing. Being white, it’s so hard to understand how black men and women live day to day with such racism.

    1. I’m not sure you should call it racism. I would expect that Chinese would rather buy from other Chinese. But Asians can’t be racist, right? Seriously, it’s just a matter of comfort and trust. Yes, the consequences are unfortunate, but it’s a deeply ingrained human trait to feel more comfortable with/around people that are similar to oneself.

    2. I’m not black, but participate in black culture, and have heard this many times from many different blacks: racism is very real and effects their lives in many ways every day. There are behaviors that are not appropriate for blacks due to social consequences brought about by racism, like walking alone at night, shouting or otherwise showing anger at work, hitting on white women at bars, wearing sports apparel, etc. Too many to list, that are all things whites do without a second thought, but blacks suffer increased inconvenience or discomfort in the exact same situation.

      Race no longer dictates where you can live or work, or who you can marry, but the effects of racial prejudice are real and pervasive, as this study shows. The number is not 20% lower because everyone thought “I’ll pay 20% less for this to counteract the untrustworthiness of the seller, which I have determined by research”, it was 20% lower because the pool of possible buyers decreased by some number. Some people refused to buy a baseball card from a black person at any price. You can’t tell me they passed over a potentially lucrative business transaction because they felt 20% less sure. This doesn’t stop at baseball cards or ipods, blacks basically pay a “black tax” of 20% in any situation where the price is flexible.

      The real lesson, which all white people should learn, is that engaging in business with minorities can be 20% more profitable!

  16. I sincerely believe every man or woman is an individual deserving the same rights as everybody else, but I observe that my emotional attitude towards other people is based on my experience of the group I catalogue them within. I honestly don’t understand how that is can be avoided.

      1. You just refuse to consider my question. We all the time make decisions based on our experience, and we can’t avoid it. If you are nearly run over by a green Mercedes thrice, you jump when you see a green one coming – and this without any intellectual deliberation at all.

        There was a story reported that Black taxi drivers in Washington refused to pick up Black customers at night, afraid of being mugged. Whether that risk had something to do with Americans treating Black people badly I don’t know, here in Europe I admit to checking if I still got my wallet when passing some Gipsy’s (or Roma people as the correct word now is, maybe even in the USA?). This is certainly pretty automatic, and based on reall and unhappy experiences. Which has nothing to do with the in itself lamentable fact that the Roma’s are given fewer legal opportunities than other people.

  17. Never underestimate the racism of people toward blacks, is a good axiom that smart black people follow.

    It is pretty taboo to talk about racism against blacks in this society, so I congratulate you on doing so.

    1. It’s not really racism per se, because it isn’t predicated on race itself but on perceptions of trustworthiness for which race is a clumsy proxy — I bet tattoos show an even more marked effect. Studies like this tend to show black buyers also prefer to buy from whites and asians.

      What this study really says is that buyers tend to find whites more trustworthy. The REALLY uncomfortable question is whether that is irrationally racist or simply rational.

      1. But what if it’s not an assumption but rather a learned behavior? Did the researchers attempt to measure whether people of different skin color are equally trustworthy?

        That’s why these tests don’t really make sense except as a club of igorance to blindly claim people are racist. You could run a similar test and say “Aha, people are more likely to think a white person is going to lose a race, therefore people have negative racist views of whites” when in fact people are just noticing the factual dominance of black athletes in racing.

      2. To demonstrate the effect you believe exists, you would need to conduct follow-on experiments in which people who’ve never met black people participate, or have the “racist-appearing” members of the survey explain their decision, or etc….

        I’ll put my money on the racism.

      3. Pretty much. There seems to have been no effort made to determine the influence of the race of the seller on the outcome of the general population of baseball card auctions. One would imagine that auctions where the seller was unprofessional enough to hold the baseball card in the pictures would tend to involve less satisfaction overall.. expecting the race of the unprofessional seller to not provide useful information of the likely outcome is pretty arrogant.

  18. There has been a similar study with similar results, involving iPods. (And also tattoos…) If you care for the details, go here.

Why We Fight - Book Cover
Subscribe to Blog