Compromising with Racism

Over at Slate, Tanner Colby has a critique of liberal US school busing policies that’s well worth reading.  Some historical context: in the wake of Brown v. Board’s 1954 mandate to integrate school districts, a pattern of ‘white flight’ emerged – white parents moving from city centers to the suburbs to avoid having to send their children to racially integrated schools.  School busing was a court-enforced reaction to this movement, designed to force the children of those who had fled to the suburbs to integrate by busing students in the whiter suburbs to more minority-dominated schools and vice-versa.  Busing has more recently been rolled back by various courts and local governments, much to the chagrin of liberals – but Colby argues the policy was actually a massive failure to begin with.  He makes some important points concerning a central goal of integration (to get students of different races to truly socialize and interact, not merely sit in the same classrooms and cafeterias) that busing did not achieve, and towards the end offers a glimpse of an alternative Colby thinks is superior.  This alternative essentially involves compromising with racism by having blacks be bused to predominantly white schools, but (acceding to the racially-motivated demands of white parents) not vice-versa.  Yet despite the allegedly good consequences of the compromise, there are inherent problems with it.  These problems, I submit, give us strong reason to reject compromising with racism in this instance.

 

The compromise

            One of the main failures of school busing was that it led to even more white flight.  Just as parents previously moved to the suburbs to avoid having to send their children to integrated schools, busing led them to move even further or withdraw entirely from the public school system (opting for private education or homeschooling).  This significantly weakened the financial and social support for public schools and thereby exacerbated existing racial and socioeconomic disparities in educational attainment that persist today.  What’s more, those who didn’t exit the public system were nevertheless bitter over being forced onto buses and resisted becoming socially integrated in the new schools.  School cafeterias would have more representative proportions of blacks and whites, but they wouldn’t eat together.

But Colby recalls an alternative outcome in Birmingham, Alabama.  Like elsewhere, white parents in Birmingham strongly resisted busing policies and pressed the courts against it.  The ruling was a compromise:  black students would be bused to predominantly white schools, but white students would not be bused to predominantly black schools.  This significantly dampened the motivation of whites to flee the school district, as they could keep their children in their old predominantly white schools and not have to bother with the hassle of long cross-district bus rides each day.  The result, according to Colby, was at least somewhat more positive than elsewhere in the country:  many whites stayed in the school district and there was greater social integration between blacks and whites.

 

Two objections

It is difficult for me to evaluate Colby’s claims that Birmingham’s compromise was relatively successful compared with other areas of the country.  Colby’s argument is anecdotal, based on his personal experiences in Birmingham.  These may well not be representative of all students’ experiences in his school, and his school may not be representative of most Birmingham schools.  But granting Colby the benefit of the doubt for the moment, it remains questionable whether the effects justify an evidently unfair and iniquitous compromise meant to accommodate parents’ racism.

The first flaw of the compromise is that it is unfair to blacks.  Being reassigned via busing from a local school to one further away another is in itself burdensome.  Students spend more of the day in transit.  The relative distance hinders their ability to fully participate in after-school activities.  They must leave an old institution, well-known teachers and established social networks behind.  With the compromise, this burden is borne disproportionately by blacks – an already-disadvantaged group that had already been subjected to years of oppression under Jim Crowe laws.  This seems to get things backwards; if anyone is to bear a disproportionate burden in order to right the wrongs of segregation, privileged whites (members of a group that bears responsibility for the harms to blacks under Jim Crowe) should be the ones to do so.  But because of whites’ disproportionate political power (a further injustice), it was blacks who had to bear those burdens.

One could respond that it is perfectly fair for blacks to bear a disproportionate burden, because they disproportionately benefit from desegregation.  Indeed, while integration may provide benefits to all sectors of society, the policy of busing does seem predominantly aimed at aiding blacks.  And generally, it could be argued that those who stand to benefit more from a policy can fairly be asked to bear a greater proportion of the costs of that policy.  However, the present case is exceptional because the policy of desegregation is one of righting the past wrongs of Jim Crowe perpetrated predominantly by whites upon blacks.  Consider an analogy: a vandal smashes your mailbox, and it will cost some money to be repaired.  You would benefit greatly from the repair, and the vandal would not.  Would it be fair, on those grounds, to say that you should bear the brunt of the costs of the repair?  Surely not – it is the vandal who should pay the costs.  This is even more the case if the vandal is relatively wealthy, powerful and overall privileged than you.  The same sort of thing is going on with desegregation – a relatively wealthy, powerful and overall privileged group caused significant harm by perpetuating segregation, and it is unfair if the group harmed is asked to bear a disproportionate burden of an attempt to remedy that harm.

The second flaw of the compromise is that it accommodates to the unreasonable racist proclivities of white parents.  The compromise is supposed to be attractive because without it, white parents will pull their children out of the public school system in even greater numbers because they do not want their children to be bused to predominantly-black schools.  To be sure, such inclinations are not necessarily racist – parents may well simply want to avoid the burdens of long bus rides to far-away schools mentioned above.  But the cultural and historical context of white flight in Alabama suggests otherwise – the fact that the schools would be predominantly black is a significant, perhaps decisive basis for their opposition.  Put another way: it is unlikely the negative reaction would be so strong if white students were merely bused between predominantly white districts (perhaps for logistical reasons).  The compromise, then, accepts and accommodates such racist inclinations, rather than seeking a policy to overcome them.  Accommodation of morally odious dispositions like racism is plausibly unacceptable (or at least should be avoided) in public policy – some positions are too repugnant to take seriously in negotiations.

 

Adjudication

What we have is a case of a classic tension in liberal governance: an unjust procedure (a compromise that unfairly places more burden on blacks and accommodates racists) begets good consequences (better integration, less white flight).  Resolving such tensions is very difficult, and a general approach is far outside the scope of this post.  But I would suggest that, in this case, we should err on the side of fair procedures and against the compromise.  Recall that Colby’s account of the benefits of the compromise has not presently been verified by robust evidence.  He has his experiences, but that’s just an n of 1 and not clearly representative.  On the other hand, the unjust burdens imposed on blacks and the accommodation of racist attitudes intrinsic to the compromise are evident, more or less inherent to the nature of the compromise. 

The actual situation, then, is an uncertain good consequence conflicting with a certainly unjust procedure.  In such a case of a significant differential in the certainty of competing, similarly-strong considerations, we should err on the side of certainty and so reject such a compromise with racists.  This is not to say that equally-applied school busing is necessarily an acceptable policy either – the bad consequences outlined by Colby may well make it untenable.  But the compromise Colby floats is also one we should reject.

  • Facebook
  • Twitter
  • Reddit

16 Responses to Compromising with Racism

  • Rogue says:

    Your premise here is that predominantly white schools are better for students than predominantly black schools. So, if parents want to put their children to white schools, on what basis does it imply racism and not wanting what’s best for their children? Either your premise is racist or your argument is flawed.

    • Owen Schaefer says:

      Rogue,

      I didn’t claim that the predominantly white schools are necessarily higher-quality, nor is that a crucial part of the push for integration. The supposedly good consequences of the Birmingham integration presented by Colby are social: children of different races interact more with each other.

      Still, you point out that white parents might themselves believe that predominantly white schools are better than predominantly black schools; their opposition to busing is not based on race, but on school quality. Fair enough – but then we’re faced with the question of why parents think there is such a differential. If they think the differential in quality is itself due to race, that’s going to be a racist view. But if they think it is other socioeconomic or institutional factors – differential funding, infrastructure, teaching quality, etc. – then they should at least support a two-pronged integration plan: A) reform giving equal funding, comparable infrastructure, equal teacher quality, etc. to predominantly white and predominantly black schools and then B) bus both black and white students between the schools to achieve social integration. I’m sure that support for A is high, but B significantly less so. Instead, in Birmingham, they supported C: bus only black students. Given that A is on the table and being implemented, what (other than race) would motivate personal preference for C over B? You could point to the hassle of busing noted above, but then the policy is still unfairly burdening one group over another – and, moreover, the strength of opposition to busing seemed to go deeper than the burdens of a long commute.

      • Rogue says:

        “I didn’t claim that the predominantly white schools are necessarily higher-quality”

        Well, I didn´t mean to claim exactly that either. The implicit (though quite obvious I think) point of my comment was that it´s quite likely that parents are using race as a proxy for economic and social class. And I would guess that most people would admit that they think students of better economic and social class –> better schools (ceteris paribus). So if the proxy is correct (as I think it is at least in some degree), race might not be playing the role you are implying it´s playing here. I´m not saying people are not racist, just that in this case race might be playing a much smaller part than socioeconomic background.

        If there were two schools, A and B and

        A) predominantly higher-middle class blacks

        B) predominantly trailer-park whites

        I would bet most of my money that most white parents would prefer to put their children to school A .

        Btw I´m an economist and a philosophy student from Finland. Here we are now experiencing “white flight” for the first time in our history. There are some areas in our capital Helsinki, in which around 20-30% of popuiation is somali, iraqi, etc war refugees. Many Finnish parents dont want their children to schools where 20-30% of students don´t even know the teaching language well. So they use immigrant background as a proxy when deciding which schools not to put their children in. I think it would be an enormous mistake to call this behaviour plain racist.

  • Rogue says:

    “The supposedly good consequences of the Birmingham integration presented by Colby are social: children of different races interact more with each other.”

    I don´t buy that. If people thought that whiteness or blackness of a school is completely irrelevant from the view of future economic and social outcomes, nobody would try to force the children on those long and frustrating bus rides. In that case forcing children on those buses would also be horrible policy from people calling themselves liberals.

  • Owen Schaefer says:

    Thanks for the further comments clarifying your view. Yes, it’s of course possible that race is a pure proxy for socioeconomic status of one’s peers, which is in turn a proxy for educational quality. But there is evidence that it was not a pure proxy; opposition to busing has been significantly associated with prejudiced racial views, support for segregation and wariness of the civil rights movement. See: Sears, Hensler and Speer 1979: http://www.isr.ucla.edu/sears/pubs/A055.pdf.

    In addition, the pure proxy view predicts that people would be relatively accepting towards school integration between poor black and poor white students. However, one of the more ugly incidences in the busing movement surrounded just such a case in Boston – the attempted busing between a poor, predominantly black school and a poor, predominantly white school was met with fierce, violent resistance. (http://www.princeton.edu/~bsimpson/Modern%20and%20Postmodern%20America/week%207%20materials/Busing%20protests.htm) Boston isn’t Birmingham, so that’s not definitive proof that pure proxy issues weren’t dominant in the latter, but it at least suggestive that opposition to busing was not based on pure proxies.

    In the Helsinki case, I don’t know enough about the cultural and political background as well as the terms of the debate to evaluate whether parents’ opposition is based on race/religion/nationality or educational quality – a rigorous study of attitudes would be useful here. Also, the case is not quite analogous to Birmingham – I take it there is not a compromise proposal where children otherwise going to predominantly immigrant schools would be bused to predominantly Finnish schools, while children from predominantly Finnish schools would not be bused the to predominantly immigrant schools for fear of white flight. Do you think such a compromise would alleviate Finns’ concerns and avoid white flight?

    And, in response to your other comment, you might be right that busing in general is justified in part as a corrective relative differences in school quality. However, I was evaluating the specific argument put forward by Colby concerning the Birmingham compromise. He did not claim that better educational quality was achieved by the compromise, but rather that the gains were in terms of better sociocultural interaction and relationship-forming between diverse groups, one of which had previously faced systematic and widespread oppression by the other. This perspective is backed up by the rejection of ‘separate but equal’ in the Brown v. Board ruling: even if segregated schools are of equal educational quality, segregation in itself is harmful (due to social and psychological effects).

    • Rogue says:

      Thanks for the thoughtful answer.

      Well yeah, my point was mostly that the info you gave didn´t in itself justify the strong racism argument without knowing the context and the details well. My Helsinki example was meant just to emphasize that argument of the same form you used could also be used here, but it would probably be false as in Helsinki´s case it´s clearly obvious that socioeconomic factors are playing a large role. The public policy in Finland is very anti-racist as in most modern democracies, but not many people are even trying to argue that our “white flight from schools” -phenomena is mostly due racism.

      ” Do you think such a compromise would alleviate Finns’ concerns and avoid white flight?”

      I think most people here would oppose such one-sided busing as a discriminating policy.

      “He did not claim that better educational quality was achieved by the compromise, but rather that the gains were in terms of better sociocultural interaction and relationship-forming between diverse groups”

      That might be the goal in someone´s mind but I´m quite sceptical against such claims. It´s rare to see politicians or anyone else trying strongly to push mixing people in relation to attributes which are considered harmless and irrelevant to any social hierarchy. It seem´s to me that most mixing projects are pursued in the name of some kind of social justice. Ofcourse I might be wrong in this…

  • Ben Salemi says:

    I think you are valuing technical correctness over effectiveness in providing remedy.

    Brown v. Board ended De Jure segregation, but not De Facto segregation. The thrust of Colby’s argument is that bussing, as pursued in larger metro areas especially in the north, was ineffectual in ending De Facto segregation. His arguments on this point seem fairly solid. Whether his proposed remedy would have been good to adopt needs to be evaluated based on whether their is adequate evidence it would have been efficacious, not whether it would have technically correct.

    To use your mailbox example, if the USPS was the sole installer of boxes, and required the vandal to pay for a new one before they would install it, this does not leave the owner in a better position if the vandal flees the state. Far better to allow the the homeowner to pay for a new one and attempt to collect later.

    Ben

    • Owen Schaefer says:

      Ben,

      I take it what you call ‘technically correct’ is what I was referring to as procedural justice. I didn’t mean to necessarily value that over effectiveness (what I called good consequences); I take them both very seriously. My point was just that, in the face of weak evidence for good consequences but stronger evidence for procedural injustice, we should err on the side of justice. If the evidence for efficacy was stronger, the Birmingham compromise might start to look acceptable (even if non-ideal). But maybe you’re a consequentialist, and think that procedural justice is in itself irrelevant independent of its consequences. In that case, I’d just reframe the debate as over the (definite) bad consequences involved in the procedure (unequal burdens and accommodation of racism) vs. the (uncertain) good consequences that result (greater integration, less white flight). Again, we should err on the side of certainty.

      I think your revised mailbox example still leaves room for procedural justice. To make the analogy even stronger, the Birmingham compromise is like if the government said they would only ask vandals to pay for a quarter of the mailbox’s cost, because if they asked for more the vandals would flee and the owners get nothing. It’s entirely possible that strategy maximizes returns for the owners, but it seems procedurally unjust. Maybe the policy would still be justified (I think it’s an open question) – but the justification of the policy becomes *much* less defensible if the government has only weak evidence that the policy would maximize compensation for owners, as in Colby’s argument for the Birmingham compromise.

      • Ben Salemi says:

        Owen,

        Yes, I believe my reference to technical correctness is essentially what you are referring to as procedural justice. You object to this proposed remedy as it presents a case of unjust procedure begetting good consequences. However, I think you are discounting the opposite effect, the just procedure leading to bad consequences, worse than no remedy or previous remedies. This is, I believe the main thrust of his argument.

        It matters not whether the government requires the vandal to pay 100% or 25% of the cost of my mailbox if I am denied the use of the mailbox until the vandal pays.

        The courts used a test that used measured de facto segregation as a proxy for determining whether de jure segregation still existed. Given the range of ways to continue a governmental policy of segregation, that was required, certainly early in the process. However, when the measures imposed to combat de jure segregation increase, rather than decrease, de facto segregation, then the remedy could be seen as to be illegal itself. It certainly is pernicious to argue against holding measures which increased de facto segregation as wrongly conceived or executed.

        • Owen Schaefer says:

          Ben,

          I didn’t mean to discount the relevance of just procedures –> bad consequences, nor insist that just procedures should always win out. I don’t think just procedures always trump consequences, nor do consequences always trump procedures. There’s often a legitimate conflict between the two. My point in this case’s conflict where there is much weaker evidence for one (consequences) than the other (just procedure), we should go with the stronger evidence.

          I guess you just deny there’s any problem with asking vandals to pay less than the ought in order to get them to pay at all, even if there’s only weak evidence that would increase the likelihood of payment/reinstallation of mailboxes. So maybe, on your view, good consequences simply trump just procedures?

          Your analysis of de facto segregation also implies a view where consequences trump just procedures, from what I can tell. It sounds like it doesn’t matter, on your view, whether or not the remedy to de facto segregation is procedurally unjust – all that matters is whether it in fact increases or reduces segregation. In my view, by contrast, it is certainly possible (though not necessary) that a policy reducing segregation may be ill-advised because of the injustice of its procedures.

          But if the consequentialist framework is what you want to go with, I’d again suggest just reframing unjust procedures as itself a bad consequence, to be weighed up against the good consequence of reducing segregation. It’d be absurd to think reducing segregation will trump any other consequences – and in the present case, the weakness of the evidence for the good consequences of reducing segregation via the compromise will substantially undermine the extent to which it can outweigh the certain bad consequences of the unjust procedure.

          • Ben Salemi says:

            Owen,

            You are emphasizing his proposed remedy. But his proposed remedy does not seem, to me, to be thrust of his argument. Let us ignore his proposed remedy for a moment, and concentrate on the actual remedy and it’s documented effect.

            The actually remedy, a remedy aimed at reducing de facto segregation, increased de facto segregation. If a remedy has the opposite of its intended effect, with strong evidence for that result, isn’t that a case where consequences do trump procedure?

            In our mailbox example, you continue to ignore my main point, which is that a proposed remedy in which the mailbox continued to be unavailable for use ad infinitum is no remedy at all. The vandal has successfully continued to deny the homeowner the use of their mailbox, for the homeowner has now been prohibited by the court from replacing it themselves. It does not matter how correct it is for the vandal to pay for the mailbox if in enforcing that correctness you bring additional harm to the homeowner.

            It would certainly be more procedurally just to force the vandal to come replace the mailbox using their own labor, time and money. But that should not trump the need for the homeowner to have a replacement mailbox in a timely fashion.

            It strikes me that perhaps a more apt apology would be to stealing a classic car, wrecking said car, hence it is impounded. Again, it would be more just if the owner did not need to lift a finger to fix the car. If in enforcing the desire to have the vandal do the work and be responsible for all costs, the court requires the vandal to pay the impound fee before releasing the car, the car is now exposed to the elements and may sustain even more damage. This is a case where the proposed remedy is most just in the abstract, but perversely may result in more harm to the injured party.

            So, it is procedurally unjust if the owner has to go the impound lot and pay to have car released, and then pay to have the car towed. We can call that a negative consequence in a consequentialism framework, if you would like. In that case I’d argue this negative consequence is less than the negative consequence of having the car sit in the lot and rust.

            But I’d also argue that it can’t be actually procedurally just to empower the thief to do additional harm to the injured party.

            • Owen Schaefer says:

              I should clarify: my initial post was a critique of the Birmingham compromise, not a defense of busing in general (which seems to be the thrust of your further examples). As I note in the end, I don’t dispute that equal/reciprocal busing may well have been an ill-advised policy (and Colby provides much stronger evidence for this claim than his alternative). Merely having an equitable procedure does not generally suffice to give one positive reason to adopt that procedure; usually, it’s that having an inequitable procedure gives one reason to avoid that procedure. What’s more, it’s arguable that *both* busing and failing to bus are just procedures in themselves (so long as there’s no de jure segregation), insofar as they bus or fail to bus blacks and whites equally. In which case, we should definitely focus on the consequences, which Colby argues strongly militates against busing.

              • Ben Salemi says:

                In re-reading your article in light of your recent response, I can see I missed some of your points of emphasis and caveats on first reading. Thank for the additional clarification.

                However, you reject his proposed compromise, the Birmingham compromise as being of “uncertainly good consequence with a certainly unjust procedure.” I would submit that given 300 years of racial inequity starting in forced enslavement every procedure will be certainly unjust. Justice could only fully be done by undoing a past long gone. And every proposed remedy will have a large measure of uncertain good in it.

                You categorically reject his proposed compromise, but in that categorical rejection it seems you would likely include every single proposed remedy. Given this, it seems the proper treatment is to evaluate this proposed remedy against other remedies, choosing the “least bad” option.

  • Mike says:

    “The compromise, then, accepts and accommodates such racist inclinations, rather than seeking a policy to overcome them. Accommodation of morally odious dispositions like racism is plausibly unacceptable (or at least should be avoided) in public policy – some positions are too repugnant to take seriously in negotiations.”

    Owen,

    I believe many people would take issue with your framing this as a ‘racism’ issue– i.e.,
    that white flight is essentially racist in motivation;
    that any action that disproportionally harms a disadvantaged group, and indeed even the existence of groups that are disadvantaged, should always be viewed through the lens of racism;
    that the resolution of any racist situation takes moral primacy over all other concerns.

    I’m not saying these are correct or incorrect, but to me it doesn’t seem like you’ve supported these assumptions, and I’m certain you could frame this issue in a less incendiary manner. Perhaps, instead of leading with the R-bomb, you could have led with what you said in reply to Rogue: “segregation in itself is harmful (due to social and psychological effects)” and it benefits society to reduce segregation. I could get behind that, and it doesn’t cast aspersions on anyone.

    I think you’re writing about an interesting topic, and saying interesting things, but doing so in a way that will turn off many in your target audience. But perhaps I’m misinterpreting who your target audience is- who would you like to convince here?

    Mike

    • Owen Schaefer says:

      Mike –

      It’s true that I could’ve framed the issue entirely in terms of the harms of segregation and the first undue burden argument. Maybe that would be more appealing. But I included discussion of racism because I think it is important to recognize as an important aspect of modern society. Racist dispositions are unfortunately pervasive (I cannot claim to be immune from them myself), and I believe do affect a wide variety policies. That’s not to say that focusing on racism is always appropriate – I haven’t discussed it much at all in my previous blog posts on this site. But the issue of busing is indeed defined by race, and people’s reactions to it are themselves informed by race. It’s true my initial post did not provide much evidentiary force for this claim; I tried to back that up a little in my reply to Rogue.

      No one likes to think of themselves as being racist, or even having racist dispositions. It makes us uncomfortable. But it’s important to discuss these influences nonetheless; pretending that race doesn’t affect us just leads us to ignore the potentially harmful effects of racist dispositions. And recognizing when racism affects the construction of and reaction to public policy, as in the case of busing, will be important insofar as it potentially allows us to try and mitigate those dispositions. And I think a policy of accommodation of such dispositions – as in Birmingham – does the opposite, either ignoring or accepting racist motivations as a given and merely trying to work around them. I think we can do better, and can begin to do so by being honest about ourselves and our dispositions.

      • Mike says:

        Owen, thanks for the reply. I was very impressed with your post on Savulescu and Sparrow’s debate on designer babies.

        The issue of racism is a tough topic to discuss. It’s tough from one angle, as you allude, because we must all confront our ‘inner racists’. Racism IS bad, and we ARE all guilty to some degree (I too am not immune). It’s tough from another angle, which you do not mention, because often the word obscures more than it illuminates: mention the R-word, and suddenly a modest discussion of busing blows up into Who Are The Good Guys, And Who Are The Evil Racists? That is, these discussions turn into a vehicle for identity politics and the use of the term ‘racist’ is often simply as a cudgel in such status games. I don’t think this is what you intend, but it’s a big reason why I call out your use of the word.

        Speaking very honestly: being a white male myself, this word is most often used as a cudgel against people like me, which has left me with a deep suspicion whenever it’s used (I tend to cringe a bit and think to myself, “I would never use what-might-sort-of-come-across-as-a-slur against him, why’s he using the same against me?*”).

        This has led me to become a big fan of skipping the big, undifferentiated Who Has Racist Motivations And Is Therefore Wrong question, and diving into the details– e.g., *how* are peoples’ reactions to busing informed by race, and *what exactly* are the motivations involved? This is actually the interesting stuff anyway. (Who in this situation are being rational actors? How are these two groups’ interpretation of rational actors different? What’s the expected economic cost-benefit distribution of busing, adjusted for everything we can? What are the rights of the majority and the rights of a minority in this situation, and how do we determine their relative weights? Etc- certainly interesting stuff!)

        *Is ‘racist’ a slur against white males? Depends on the context. I would just note it’s a powerful word, as indeed you say: “No one likes to think of themselves as being racist, or even having racist dispositions. It makes us uncomfortable.” (Uncomfortable might be putting it lightly… here’s an exposition of the way I feel about the word. It’s a bit long but I recommend it unreservedly: http://slatestarcodex.com/2014/01/12/a-response-to-apophemi-on-triggers/ ).

Recent Comments

Authors

Affiliations