Colour and favours on the bus? It matters if you’re black or white!

Is there discrimination on colour in Queensland? In order to find out if black and Indian people are a discriminated ‘out group’ in Queensland, together with Redzo Mujcic I carried out a large-scale experiment involving bus drivers in Brisbane. We sent test subjects of various races to bus-stops all over Brisbane where these individuals would board a bus with a faulty bus card, subsequently asking the bus driver whether or not they were allowed to stay on the bus.

The 1,552 encounters recorded between bus drivers and test subjects revealed that in the main scenario, some 72% of white Caucasian test subjects were generously given a free ride, versus only 36% of black test subjects (Indigenous Australian, African, African American, and Pacific Islander). Indian test subjects were treated slightly more favourably than black test subjects and were let on 51% of the time, whilst white Asian test subjects (Chinese, Japanese and Malaysian) were let on 73%, almost as much as white Caucasian test subjects. Males were let on 8% more than females, and bus drivers were 6% more likely to favour someone of the same race.

In order to see the importance of dress, we then sent the same test subjects to bus stops in business suits and with briefcases. In such white-collar attire, some 75% of the Black and Indian test subjects were let on versus 93% of white test subjects, indicating that higher social status engendered more favouritism from bus drivers.

We conjectured that the reason bus drivers were more reluctant to give black and Asian help-seekers a free ride was that they did not personally relate to them. In order to test this, subjects were sent to bus stops dressed in army suits, making it appear that the test subjects were patriots and were defending the same community as the bus driver.

Some 97% of white test subjects were given a free ride if they dressed in an army suit, whilst 85% of black and Indian test subjects were let on with an army suit, showing that the reluctance to let on normally dressed blacks and Indians was indeed related to an ‘out group’ reaction by bus drivers that could be overcome by ‘in group’ clothing.

An important possibility to account for was whether black and Indian test subjects were perceived as less trustworthy and more aggressive than the other test subjects. To test this, pictures of the test subjects were presented to random passers-by in selected regions, asking those passers-by to rate how aggressive, trustworthy, and pretty the test subjects looked. The test subjects that were rated as less aggressive were let on slightly more often than more aggressive ones, while perceived trustworthy individuals were also favoured at a higher rate than untrustworthy ones, but this did not affect the results on race.

As a follow-up, we then conducted a survey of random bus drivers at selected resting stations around the city, presenting them with pictures of the same test subjects that had gone to bus stops and asking the bus drivers whether they would let them on or not with an empty travel card.

Some 80% of the bus drivers at resting stations indicated they would give free rides to Indian and black test subjects, even though in reality less than 50% were let on. Indeed, bus drivers said they would let on white subjects 5% less often than black subjects, whilst in reality white test subjects were favoured at least 40% more than black testers. When asked for reasons why a person would be let on or not, the main reason given for not letting someone on was that it was against the rules, whilst the main reason given to let someone on was that it was no burden to do so.

When the bus company learned of the survey, it sent out a general message to all bus drivers in Queensland not to participate any further.

What this study showed is that when it comes to doing favours, there is still significant racial discrimination in Brisbane. People with Indian or black complexion are still more likely to be treated as an ‘out group’ and thus less part of the ‘in group’ compared to white Caucasians and white Asians.

There were many important further elements to these experiments to talk about in future blogs. For those who like to see the preliminary paper for themselves, see here: mujcic_frijters racial bias March2013.

The study is an important stepping stone towards a general theory of human behaviour that is in a recently published book ‘An Economic Theory of Greed, Love, Groups, and Networks’. The first launch of that book is next week in Brisbane, 28th March 2013, 5:30pm, The QIC Building, Level 5, Central Plaza Two, 66 Eagle Street. CNR Elizabeth & Creek Streets. Brisbane, QLD 4001. RSVP [email protected]

This entry was posted in Uncategorized. Bookmark the permalink.

21 Responses to Colour and favours on the bus? It matters if you’re black or white!

  1. Jolly says:

    Asians, Blacks, and Browns are different!! Good Lord, their skin colour is sooooo obviously different from the rest of us. They need to be treated differently. We step out of Australia and we are confronted with Yellows, Browns and Blacks. We are afraid that given time, the Caucasian will disappear from this land because all of Asia is coloured. At least at one time, we had the “Yellow Peril” fear officially in place. Then we had Handson championing our innate fear of the coloureds. Today we have no one to champion the caucasian cause …just bus drivers. Viva bus drivers! The author has nothing better to analyse!

  2. RT Green in China says:

    Unable to use my normal log in (or twitter).

    It’d be very interesting to see a similar study in a generation. For me, the relevant distinction between East Asians, against whom you found no discrimination, and South Asians, and Africans, against whom there was plenty, is less their melanin content, but that they are a later migration cohort. If multiculturalism is working with a generations lag, maybe busdrivers will end up empathising with the latter the way they apparently do with the former.

    That puts a hopeful spin on things. It suggests that the discrimination, present, nasty and undeniable, might be washed out with time.

    That said, this would require disentangling Africans from the Murris. The discrimination against the latter seems more indelible.

    • conrad says:

      You’re generalizing that too much. There is a big contexutally specific aspect of racism. For example, as for my note below, when Andrew Leigh did a study using names and jobs, he found a decent effect of being Chinese (this is replicated a number of times around the world — in France they send off Arab sounding names). But this isn’t surprising, because if you want to try and make some sort of predictive model, you needs theories of attribution, group formation and so on, and these predict differences in different situations.

  3. conrad says:

    I’m surprised you’re getting into this area — it’s such a done to death area in psychology it’s hard to see what’s left to contribute, which is not to say that you couldn’t think of a novel take on it, although you’d have to do some really exceptional stuff such that people would care over other well entrenched ideas out there (good luck). That being said, I hope you don’t start reminding me of Andrew Leigh where he did really great stuff in some areas but his stuff on racism looked like rehashed stuff people had done in psychology decades before, including making the same mistakes (like using crappy “implicit” tasks). Seemed like a bit of waste to me.

  4. Pappinbarra Fox says:

    This is a very interesting item. I wonder if there is any breakdown of the stats by age? Were older people more likely to be let on than younger people?

  5. Michaelc58 says:

    Pseudoscience at best.
    Numbers everywhere ant not a statistical test in sight.
    How were confounding factors controlled, such as appearance of socioeconomic status, education, and language controlled?

    While no doubt biases of us vs them exists until the new ‘colour’ is assimilated, as the Chinese, Italians, Greeks etc have been, a bit more critical analysis of advocacy ‘studies’ would be good.

    Anyway, humans are necessarily categorising machines for the sale of efficiency. Some bikies are no doubt saints, but it would be stupid to pick an argument with someone in that category. Three black people in a dark alley could be the St Andrews choir, and the white woman, Jackie the Ripper… well you know. Discussions should recognise the utility as well as costs and damage of ‘discrimination’ and not lecture from high moral ground as if it was come incomphehensible moral defect in people.

    • FDB says:

      Pseudoscience at best.
      Numbers everywhere ant not a statistical test in sight.
      How were confounding factors controlled, such as appearance of socioeconomic status, education, and language controlled?

      RTFP, dude.

  6. observa says:

    Did you do anything remotely useful and find out why they’re letting any of these freeloaders on the bus without paying? Would sacking bus drivers that do be more racially sound? Are you just another middle class white guy living off the sweat of trammelled minorities and blue collar workers? Did you offer to pay for some poor unfortunate without a valid ticket? These are the great questions that remain unanswered.

  7. Richard says:

    Hi Paul,

    What a fascinating study.

    I have a thought: could the same effect be caused by the perceived authority of the person getting onto the bus? It might also explain why the suit and army uniform make such a difference to outcomes. Perhaps white people are seen as having more authority in Australia: just look at the Cabinet or police to see why someone might think that!

    Perhaps an interesting follow-up would be to perform the same experiment in a country where a racial minority group tend to make up the ruling class?

  8. Mel says:

    Absolutely fascinating study, Paul. I only wish more studies like this were done, altho I understand it must require a huge amount of work.

    I wonder if something akin to stereotype threat plays a role in the bus drivers’ response to lower status groups?

    Wiki defines stereotype threat as:

    “Stereotype threat is the experience of anxiety or concern in a situation where a person has the potential to confirm a negative stereotype about their social group.”

    Maybe lower status group members are more likely to exhibit behaviours, like less eye contact and nervousness, that are associated with untrustworthiness.

  9. Mel says:

    Another couple of points

    – how did you determine the race status of the bus drivers? Was this done on appearance, were they asked etc..

    – why lump Chinese, Japanese and Malay together? Ethnic Malays are usually markedly darker than their North Asian Chinese and Japanese neighbours. Did this impact the results?

    – persons who don’t look obviously indigenous often identify as such. Did this situation apply to any of your testers?

  10. observa says:

    You call THAT absolutely fascinating? THIS and THIS are absolutely fascinating!

    My folks drank instant coffee, walked if they didn’t have a bus ticket and called them toilets which you didn’t hang around in and the older I get the more common sense I realize they all had back then.

  11. Paul Frijters says:

    thanks all, appreciate the comments. Things are a bit too hectic here with the upcoming launches to respond in detail to these comments, so let me just say that the actual paper answers nearly all questions posed and that i hope to get back to this topic soon!
    Jolly, observa: wow! Conrad: see Jolly and observa. Still think there is no need for this research and that we all already know this?

    • conrad says:

      Paul, I think the data and experiment are fine (and good) — I think this sort of data is really useful in many contexts. My comments was really with regards to your comment that:”The study is an important stepping stone towards a general theory of human behaviour that is in a recently published book”

      If you really want to come up with a general theory of human behavior (which is a great goal!), then I don’t see how this particular piece of data will contribute anything meaningfully (perhaps excluding longitudinally over a long period).

      The reason for this is that we already know the typical results of this type of experiment. It’s not like people haven’t manipulated physical characteristics of the person (race/age/gender/fatness/baldness/hair length/spoken accent/handicap status…), things that the person wears (lab coat/suit/clothes of different nationalities/cleanliness of clothes/price of clothes..), task (job/housing/need for help [both for positive and negative outcomes]/acquisition of a public service/acquisition of a private service..), or relationship between the two people (knowledge of other/power status/amount of time known/previous interactions..) before (this latter one should be especially fun if you want to develop a theory, because it appears many of these effects disappear especially quickly — something that was not investigated until comparatively reasonably).

      So in the end, I don’t think I need to update my comparatively poor understanding of models of racism because of that data point (it basically lives in the distribution of all of the rest). Not that I want to invoke the spirit of Mike Pepperday on you :) but if you want to create such a theory (and I think that economists look at things differently to social psychologists, even though they often look at essentially the same problems, so you may well have something really unique to contribute), then you need you to think of something that your theory predicts that other theories do not and tell us why (and then get the data if it can’t be scavenged). This is especially the case because it appears you obviously have pretty decent resources at hand, as that study looks like a lot of effort to have done.

  12. Jason Soon says:

    why lump Chinese, Japanese and Malay together? Ethnic Malays are usually markedly darker than their North Asian Chinese and Japanese neighbours. Did this impact the results?

    Actually the study said ‘Malaysian’ not ‘Malay’. The former is a nationality, the latter an ethnicity. However Mel is right to pick up on this because he would’ve rightly assumed that you were referring to the ethnicity. However a lot (if not most) Malaysians in Australia are ethnically Chinese so that’s how I read it.

    More interesting puzzle is why the Indians were singled out with the Aborigines. I tend to associate ‘Indian’ with professions in IT i.e. a ‘model minority’ rather than a ‘historically discriminated underclass’ stereotype. Given the stronger racial prejudice towards Aborigines could it be many of them were simply mistaken to be Aborigine or is it really as simple as the color of their skin being darker?

  13. Magpie says:

    Dear Prof.

    I haven’t read your paper, yet (thanks for the link).

    Beyond the subject of racism, this paragraph caught my attention:

    “In order to see the importance of dress, we then sent the same test subjects to bus stops in business suits and with briefcases. In such white-collar attire, some 75% of the Black and Indian test subjects were let on versus 93% of white test subjects, indicating that higher social status engendered more favouritism from bus drivers.”

    Judging by what you say above, it would seem that not only racism is at play. Could this be interpreted as evidence of __class__ discrimination? Has this effect been found somewhere else?

  14. observa says:

    The problem is definitely one of class. Firstly my class who are past masters at grabbing economic rent out of taxpayers (largely off the sweat of those who can’t) and indulging themselves in ever more marginal pursuits.

    Firstly they largely wouldn’t be seen dead on public transport but they’ll internalise the horror of their pampered offspring should they ever be kicked off a bus for bad behaviour or lacking a ticket. Consequently bus drivers are rather circumspect about who they refuse entry to, lest they cop the ire of the chattering classes. Furthermore they’ve learned to shrug their shoulders at the problematic ticketing systems that have been foisted on them by a management class that couldn’t organise a booze up in a brewery. Think Adelaide’s Cruzet ticket system and Melbourne’s too by all accounts.

    When you’re on the Gummint teat and not responsible for the bottom line naturally it’s all about moi and internalising any possible unpleasantries for you and yours and woe betide the dumb slob that could possibly interrupt that class enjoyment, either actually or hypothetically. Consequently the time honoured, egalitarian, classless system of no pay no ride doesn’t enter the picture and the poor slob bus drivers have to dance on middle class eggshells all the time.

    Me? I take the car and try and ignore the worst of my class trying to find racism, misogyny or homophobia around every corner. Well not their particular corner because they don’t mind utilising the free market for houses and rents to sift out any unpleasantries and remain above it all.

    • Mel says:

      What is it about Obby that reminds me of Norman Bates? Is it just the skin crawl factor or is there more to it? Mmmmm.

  15. Nicholas Gruen says:

    Discrimination: what it feels like.

  16. Pingback: Ray Martin is shocked, shocked at the way media framing is fanning social division: Racism in Australia Part One | Club Troppo

Leave a Reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.