Get the F*** off OUR roads’ — Motorists, Cyclists and Intergroup Bias

In Australia it is that time of year again… sum­mer. Where the weath­er gets nicer, and in Adelaide the Tour Down Under arrives in town. Now unsur­pris­ing­ly this annu­al event sees the  sea­son­al rise of vis­i­ble cyclists, and of course accom­pa­ny­ing it the usu­al dia­tribes and vit­ri­ol flash­ing about in all direc­tions over the top­ic. There are many direc­tions that these ‘con­ver­sa­tions’ inevitably go in, be it down the path of licens­ing, or psy­cho­path­ic motorists, or appar­ent fla­grant dis­re­gard for the law… from both sides. However, none of these are what I want to address in this post. Rather, I think it is help­ful to look at some of the under­ly­ing fac­tors with­in the cyclist/motorist inter­ac­tion, specif­i­cal­ly that of group bias­es and Social Identity Theory (SIT). It is espe­cial­ly help­ful in this case because the inter­ac­tion is rel­a­tive­ly arbi­trary and cross­es many oth­er more com­plex social bounds in a rel­a­tive­ly equal fash­ion. This helps as it acts as a type of micro­cosm or case study that can inform much more com­plex interactions.

ingroup-outgroupFirstly, an exceed­ing­ly brief overview of SIT and some of the bias­es at play. SIT was for­mu­lat­ed by Henri Tajfel and John Turner in the late 70s and ear­ly 80s as a means of explor­ing inter­group rela­tions. [Ref]Tajfel, Henri, and John C. Turner. “The Social Identity Theory of Intergroup Behaviour.” In Psychology of Intergroup Relations, edit­ed by William G. Austin and Stephen Worchel. Chicago, Ill: Nelson-Hall, 1986. 1 There they note some of the fac­tors that impact upon inter­group aggres­sion; quot­ing from their abstract:

Perceived inter­group con­flict of inter­ests, the pos­tu­lat­ed moti­va­tor of aggres­sion, pre­dict­ed it strong­ly. The effects of con­flict on aggres­sion were par­tial­ly medi­at­ed by 2 index­es of dehu­man­iz­ing the out-group (per­ceived val­ue dis­sim­i­lar­i­ty and trait inhu­man­i­ty) and by 1 index of prob­a­ble empa­thy with it (per­ceived in-group–out-group bound­ary permeability).

In effect they name ‘inter­group con­flict of inter­est’ as the pri­ma­ry moti­va­tor, and impact­ed by the dehu­man­i­sa­tion of the out-group and the per­me­abil­i­ty of the bound­aries. Finally in anoth­er study by Mackie et. al. they found sig­nif­i­cant appli­ca­tion of the fun­da­men­tal attri­bu­tion error with­in groups, nov­el­ly nam­ing this ‘group attri­bu­tion error.’ 2 In lieu of the longer post on FAE to come the sim­pli­fied under­stand­ing is such that in-group mem­bers char­ac­terise out-group mem­bers by indi­vid­ual actions (and usu­al­ly those that serve the in-group con­fir­ma­tion bias).

Aggressive Motorist CartoonSo how does this impact upon our lit­tle case study? Well if the motorist/cyclist dynam­ic is dichotomised between cyclists and motorists, as the debates ensue, then SIT can be utilised in look­ing at the inter­group inter­ac­tions. Addressing the first of the sub fac­tors from Struch and Schwartz, even though the per­me­abil­i­ty between groups is incred­i­bly high, with many bicy­cle rid­ers own­ing cars, and obvi­ous­ly vice-ver­sa, the per­ceived per­me­abil­i­ty is exceed­ing­ly low. I would sug­gest that this is due to the mutu­al exclu­siv­i­ty of the means of trans­port, its impos­si­ble to oper­ate both at the same time, and only a mar­gin­al per­cent­age of cars are seen with bike racks. Furthermore the pro­lif­er­a­tion of the ‘own­er­ship’ of the roads, as high­light­ed by the large num­ber of aggres­sive claims to ‘our’ roads from both sides serves to fur­ther delin­eate the groups.

The sec­ond of the Struch and Schwartz char­ac­ter­is­tics is that of dehu­man­i­sa­tion of the out-group, and this is extreme­ly easy to see in the lan­guage used in the debates. Via that eru­dite medi­um of Facebook I have seen a pletho­ra of invec­tives such as ‘death cage oper­a­tors’, ‘lycra scum’, etc with many more that aren’t worth repeat­ing. All of these serve to remove the per­son from the out-group, and replace them with a dehu­man­ised label. For an even more preva­lent exam­ple of this, see the American pro­pa­gan­da dur­ing the Vietnam war dehu­man­is­ing the Vietnamese as mon­keys (c.f. the work of Albert Bandura on the same). The last of Struch and Schwartz’ char­ac­ter­is­tics is that of con­flict of inter­ests, which in this case is the usu­al and pre­dictable con­flict over space on the roads.

Mackie’s appli­ca­tions of group attri­bu­tion error can be rel­a­tive­ly eas­i­ly seen as well with the anec­do­tal evi­dence base sig­nif­i­cant­ly out­weigh­ing any sta­tis­ti­cal or Bayesian mea­sures. The usu­al argu­ment appears: ‘I saw a cyclist break­ing the law, there­fore all cyclists break the law’ or ‘I was harassed by a road Cyclist Denigrated Cartoonrag­ing dri­ver once, there­fore all motorists are out to kill me.’ As with most, if not all, attri­bu­tion bias­es there is an ele­ment of truth there, but lit­tle to no sta­tis­ti­cal sig­nif­i­cance or repeata­bil­i­ty. So these anec­do­tal ‘evi­dences’ serve only to strength­en the out-group dis­crim­i­na­tion bias, and rein­force the in-group bias. Furthermore the inverse is true, motorists don’t self-char­ac­terise by those ‘hoons’ or crim­i­nals who kill peo­ple in acci­dents, and nei­ther do cyclists self char­ac­terise by those who run red lights and knock down pedes­tri­ans. The con­fir­ma­tion and attri­bu­tion bias flows in both directions.

Finally it is worth acknowl­edg­ing that there are a pletho­ra of oth­er fac­tors at work, from con­fir­ma­tion bias­es to clus­ter­ing illu­sions, empa­thy gaps and many more. However, the major­i­ty of these serve to rein­force exist­ing group bound­aries, rather than dis­solve them, so while they con­tribute to the big­ger pic­ture it is in terms of detail rather than applicability.

So what can be done with this sit­u­a­tion? It is all well and good to use SIT to describe an inter­group inter­ac­tion, but as with many aspects of acad­e­mia it is hol­low if left there. One of the advan­tages of describ­ing the inter­ac­tion in this way is that par­tic­i­pants in the groups get to see how their bias­es shape the inter­ac­tion as a whole. This is where edu­ca­tion comes into play. While edu­cat­ing cyclists that not all motorists are homi­ci­dal psy­chopaths, and edu­cat­ing motorists that not all cyclists are fla­grant­ly law-flaunt­ing dilet­tantes will not remove those who are gen­uine­ly homi­ci­dal psy­chopaths and fla­grant law-flaun­ters, it does break down the bound­aries somewhat.

This break­ing down of the bound­aries is impor­tant on two lev­els, first­ly as it dis­man­tles some of the con­flict, and sec­ond­ly as it removes places for those who gen­uine­ly are psy­cho­path­ic or law flaun­ters to hide with­in their respec­tive in-groups. I note that the Motorcycling Victoria is doing sig­nif­i­cant­ly more on the edu­ca­tion front than I have seen the cycling and motor­ing groups do in recent times. See this video here: There are many more appli­ca­tions of SIT in iden­ti­fy­ing bias­es and break­ing down the stereo­types, such as serv­ing to re-human­ise the par­tic­i­pants in each in-group and many more that I don’t have the time to explore here now. Suffice to say that prop­er analy­sis of the inter­group bias and inter­ac­tions helps to inform efforts to resolve issues. But I would also sug­gest that with­out a good under­stand­ing of the group dynam­ics at hand there will be lit­tle trac­tion in the pletho­ra of dis­cus­sions to be had.

Lastly, its worth not­ing that while the cyclist/motorist exam­ple is a salient one for many, myself includ­ed as I span both groups (dis­claimer: motor­ing AND cycling enthu­si­ast), it can read­i­ly be extrap­o­lat­ed to oth­er inter­group con­flict. The oth­er swirling debates over ‘Islam vs the West’, var­i­ous racial dis­putes, Republican v Democrat, Liberal vs Labor, lib­er­al vs con­ser­v­a­tive, reli­gious vs athe­ist, and many more all find appli­ca­tion with­in the realm of SIT. Furthermore they all can be assist­ed in bet­ter con­ver­sa­tion and pos­si­ble res­o­lu­tions 3 to var­i­ous degrees by iden­ti­fy­ing the inter­group con­flicts and see­ing the ori­gins and rein­force­ment of the bias­es present.

What do you think? Weigh in on the com­ments below.

About Chris


  1. Primarily SIT seeks to define groups and their rela­tions such that there is a form of pre­dic­tive capa­bil­i­ty of the inter­ac­tions between the groups. At a sec­ondary lev­el it allows for a struc­tured method­ol­o­gy for analy­sis of inter­group rela­tions and con­flict, the pri­ma­ry use for it in this sit­u­a­tion. Since SIT’s pro­pos­al has been aug­ment­ed by a series of papers that have inves­ti­gat­ed how SIT may be used to elu­ci­date fur­ther aspects of inter­group inter­ac­tion. Of par­tic­u­lar rel­e­vance here is the work by Struch and Schwartz. 4Struch, N., and S. H. Schwartz. “Intergroup Aggression: Its Predictors and Distinctness from in-Group Bias.” Journal of Personality and Social Psychology 56, no. 3 (March 1989): 364–73.
  2. Mackie, Diane M., Scott T. Allison, and David M. Messick. “Outcome Biases in Social Perception: Implications for Dispositional Inference, Attitude Change, Stereotyping, and Social Behavior.” Advances in Experimental Social Psychology — ADVAN EXP SOC PSYCHOL 28 (1996): 53–93. doi:10.1016/S0065-2601(08)60236–1.
  3. Many res­o­lu­tions are like­ly impos­si­ble, but at least not debat­ing over use­less top­ics