Why are you hitting yourself? Part 8: extra credit questions

This is part 8 of an ongoing discussion of a paper by Jost, Banaji and Nosek discussing System Justification Theory. Read Part 1Read Part 2. Read Part 3. Read Part 4. Read Part 5. Read Part 6. Read Part 7.

Having summed up my lengthy exploration of System Justification Theory, I teased you this morning with the question that you’ve likely been asking youself from the beginning: now that we know about system justification, what can we do to correct for it? Are we doomed to keep making the same mistakes, or can we overcome our terrible mammal brains and become better critical thinkers?

In order to answer this question, I must first re-iterate a point that I’ve been making for almost as long as this blog has been in existence: we can overcome cognitive biases by becoming more aware of them. Just like we, as skeptics, have learned to recognize faulty arguments like straw men and fallacies like appeal to authority, we can also learn to recognize when we (or others) base their arguments on streotypes instead of evidence. System justification lives on stereotype – confronting those will go a long way on its own to reduce the amount of system justifying we do.

There is also something important to be learned from Part 6, which is that system justification is directly connected to the level of inequality present in a society. As we reduce gaps between groups – be they through legislative policies like pay equity or through changing the social stigma associated with being in the minority – we reduce our tendency to ‘explain away’ disparities as being part of the natural order of things. By engineering societies that are more fundamentally equal, we simultaneously rob fuel from the system justifying machine.

Finally, and perhaps most crucially (or just my preferred method), we can reduce system justification by talking about it. The more people are aware that they have a tendency to do this kind of backfilling, the more likely they are to notice themselves doing it in the future. Successes in my own ongoing struggle to become less misogynistic suggests to me that awareness (and acceptance) of the fact that we all have cognitive demons operating below the level of conscious awareness will help us police our own attitudes better. We may never become perfect at it, but we can certainly become better.

Now, I would be a really crummy scientist if I didn’t use this opportunity to raise some research questions of my own that this paper did not address.

Does making group status (high- vs. low-) more salient lead to a change in system justifying behaviours?

The issue with implicit processing is that we incorporate attitudes towards certain groups, often without realizing that we are processing group status. One of the examples the authors use is that of white and non-white students alike demonstrating a preference for partnering with a white person for a co-operative task. I highly doubt that students consciously processed race when making their decisions. My question is, are we less likely to engage in system justifying behaviours when group status is made more explicit? Will reminding members of minority groups of the fact of their low-status relationship make them more likely to advocate for in-group causes rather than demonstrating out-group favourability?

Related to this question is the conservative talking point: that talking about race issues makes people resentful of each other. I think this is nonsense, designed at preserving the status quo, but how does this attitude relate to system justification? Will people who are prone to system justifying behaviour demonstrate increased out-group hostility when reminded of their status? Will members of the majority demonstrate more explicit antipathy toward minority groups if reminded of the power divide? Is that effect moderated by level of self-confessed politicial conservatism?

Is system justification moderated by the size of the status gap between groups?

One could make the argument that there are some groups that are not as relative low-status as others. Atheists are a low-status group compared to theists, for example, but they (we) do not face nearly the same kinds of discrimination that Arab Americans or gay people do. Does the effect to which people demonstrate system justifying behaviours change in proportion to the magnitude of the status gap? For example, do black women demonstrate more system justification than black men when compared to white men – this is assuming we can control for other factors like socioeconomic status.

Related to this question is whether or not the gap between explicit and implicit attitudes toward group favouritism (see graphs in part 3) is moderated by the size of the power gap. The graphs for age discrimination and race discrimination look quite different from the graph for sexual orientation (see paper page 898-900) – is that related to the magnitude of age discrimination vs. anti-gay discrimination (two types of discrimination that look very different), or something else?

Are some people predisposed to system justifying behaviour? Is it related to a ‘capacity’ for cognitive dissonance?

Anecdotally, some people are better at ‘suspending disbelief’ than others. People who are otherwise intelligent but manage to compartmentalize the various contradictions within religion, for example (I am not referring to those that pretend they don’t exist – denial is something else entirely) – are they just “better” at simultaneously holding conflicting ideas in their heads? If that’s the case, and some people are better at handling cognitive dissonance than others, are those kinds of people also more prone to system justifying behaviours? Will they more readily adopt out-group favourable attitudes if they are in the minority? Basically – can SJT explain why women are more religious than men, poor people are more religious than wealthy, and why communities of colour are more religious than whites? Is it because the high tolerance of cognitive dissonance needed to be a believer is related to the high tolerance of cognitive dissonance needed for out-group preference?

I think these are all interesting and relevant questions for reducing the impact of system justification, or at least moderating its effects. I am sure you can think of more questions that you’d like to see answered. Please consider leaving them in the comments. Bonus points if you can propose a method of measuring them.

Like this article? Follow me on Twitter!


  1. Allienne Goddard says

    Don’t think so. It ends:

    Basically – can SJT explain why women are more religious than men, poor people are more religious than wealthy, and why

  2. Crommunist says

    God damn. I have ‘fixed’ this thing 5 or 6 times and it keeps reverting to an old save.

  3. quantheory says

    If I had to make a hypothesis related to your second question, I would say that one thing that really sets the LGBT community apart is that guilt, shame, and low self-esteem (what we call “internalized homophobia”) were identified very early on as important factors that harmed us and prevented us from speaking out on our own behalf. Coming out is an event where one explicitly, consciously switches from the default straight identity, to an LGBT one, and a statement that one values that choice. As a result, virtually the entire history of gay rights since Stonewall has involved putting ourselves, our community, and our lives in a positive light (the major exception, perhaps, being the worst years of the AIDS epidemic, when the struggle was seen as being one for survival rather than for tolerance/acceptance).

    We have “Gay is Good”, “It Gets Better”, celebrity icons, and annual events dedicated to coming out and to “pride” (pride, of course, being the antithesis of shame). Encouraging gay individuals to value themselves above oppressive systems is an explicit goal of these movements, in some cases even more so than raising the status of the LGBT community in general (“It Gets Better”, for example, is not targeted to winning over straight people at all). It would be interesting to see if out-group favoritism in gay people tracks with exposure to these messages, and also how it correlates with self-esteem at different stages of life. Do gay people with an in-group bias have higher self-esteem than before they came out, or lower? What about gay people showing out-group favoritism? Are these changes correlated to “coming out to oneself”, to finding LGBT resources, or coming out more generally (to family or the public)?

    If these messages work, they are effectively a way of reducing out-group favoritism without necessarily decreasing the status gap, except as a side-effect.

    On top of all that, for practical reasons of dating and acceptance, LGBT people are usually compelled to seek out LGBT peers. When the explicit anti-gay attitudes of straight people in an area are very strong, the contrast between the prejudice or outright meanness of many straight people, versus the more easygoing, accepting atmosphere of a gay community, probably increases in-group favoritism. An interesting way to test this would be to compare bias among LGBT people who live in more accepting areas, to those who live in less accepting ones. The difficulty would be in controlling for internalized homophobia, which would presumably be stronger in less accepting areas (perhaps this could be done by picking people who had moved at around the same time as coming out, or had come out some time ago and had the chance to cast off those sources of out-group bias).

    Or a more general question would be this: All other things being equal, does increasing the explicit in-group favoritism of the high-status group increase, decrease, or have no effect on, the in-group favoritism of a low-status group. System justification would seem to push for a negative correlation (low-status group adopting the values and justifications of the high-status group), while social identity factors would seem to push for a positive correlation (the low-status group pushing back against prejudice). Perhaps both effects occur in different contexts, with system justification being more powerful among groups with a poor sense of community and group identity (older people, gay people before Stonewall, surely many others) and social identity taking over when a group is able to reframe prejudicial attitudes as a moral deficiency rather than mere common sense.

    I think that there’s an unfortunate side-effect of gay in-group favoritism, which is that the gay community and allies have tended to promote “complementary” stereotypes more than they should. If being gay is all about sex and art and wild novelty that straight people are too boring for, that fits in quite well with the stereotype that gay people are shallow hedonists and incapable of having the sort of stable relationships that form the basis of family life. It would be interesting to see if there is a correlation between complementary stereotypes and sexual-orientation bias (already established for income and gender related biases).

    As a matter of purely personal experience, I felt a sort of disconnected discomfort with my bisexual identity that was never quite resolved until I started hearing other people’s coming out stories (which tended to dispel stereotypes and uncertainties that I’d been unconsciously carrying around up until then). Bearing that feeling (for about two years), and then feeling that burden lift, without my conscious attitudes really changing at all, was quite an extraordinary experience.

    (As has happened to me a lot during this series, my comment became far longer than I intended. Hopefully it provides something interesting to consider.)

  4. Enkidum says

    Does making group status (high- vs. low-) more salient lead to a change in system justifying behaviours?

    This is an interesting one. I’m sure you know about stereotype threat, but as a brief example, if you include the question “Gender: M/F” at the start of a math test, women will, in general, perform worse on that test than if it was not there. Which is an insane finding, but a well-replicated one (same thing for a question about race for black people).

    But it could be argued that such a question does not make gender (or race) super salient – just enough that it kicks off those damn implicit processes. I believe there is actually research showing that if you explicitly discuss stereotype threat and explain the purpose of the study, the effect goes away. So at least in that example, your intuition is right.

    Of course I think making it salient can sometimes lead to its own set of problems. The whole treating-you-as-a-BLACK-guy-instead-of-as-a-human thing that you’re certainly well acquainted with. But I suspect that is a price worth paying, especially since the problem generally lessens with time (I assume).

    Props to you for this series – I have to say that reading PZ and feminist blogs over the past year has led me to reconsider many of my actions, and to a lesser extent that may be happening with race too. So thanks!

  5. bemused says

    I recently stumbled onto this blog series about SJT and it is very timely since I recently have had discussions with family and friends about this sort of phenomena. The historical example I use occurred several years ago when George W. Bush was running for his second term as the US President. A local paper interviewed several “NASCAR Dads” who all stated they would be voting for Bush (again). What really shocked me was that most, if not all, stated that they were well aware that economic policies promoted by Bush and the Republicans would hurt them but they were voting for him anyway because of “the (religious) values thing.”

    In thinking of SJT in relation to the above example:

    1) It seems that SJT has very good explanatory power to make understandable why people in a lower socioeconomic status would support those in power who promulgate policy that hurts those in the lower strata.

    2) I am not sure that SJT explains why this group would choose religious belief over (potential) economic security (although this might be explained by various features of particular brands of Christianity such as Christ suffered so you should suffer too, etc…). Perhaps SJT would explain this problem in terms of the dominant social structure having a particular religious belief combined with a certain amount of economic disparity. The lower strata would then identify with the dominant group through the commonality of religious belief, a belief that justifies (and thus encourages) greater economic disparity.

    3) When I read part 1 of this blog series I immediately thought of the Stockholm Syndrome.

    4) I wonder if the phenomena described by SJT are a result of the fact that humans are inherently social animals and it is thus better to be part of a group, even if that group treats one badly.

    Thank you for your summary and analysis of SJT. I have already forwarded this to others for mental mastication purposes.

Leave a Reply

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