Republican Schoolmaster and the Narcissism of the Minor Differences

Now that we all know who the new President of the United States will be, people are preparing for a new type of government, with a new and markedly different agenda than the previous one. Most people are very contend with this new agenda, but some will be disappointed. How does this influence the people’s opinion, one might ask? Will conflict be the result, or can one expect that in general the new agenda will be accepted and that those who voted McCain will change their opinions to generally accept the new policy?

In their classic article, Franklin and Kosaki studied something similar, although they focused on the impact of a Supreme Court ruling regarding induced abortion. Remember though, how Obama and MaCain differed on this issue, with Obama stating to protect the Roe v. Wade Court Ruling, and McCain trying to overthrow it. It is exactly the effect on public opinion of this Roe v. Wade Court Ruling that Franklin and Kosaki studied.

According to the authors, it has for long been held accepted that public opinion tends to converge with present legislation. So, in the case of abortion, the effect of Roe v. Wade (which, in a practical sense, legalized induced abortion in 1973) would be that public opinion on induced abortion would become more permissive. They refer to this as the positive response hypothesis. They contrast this hypothesis with a much more interesting one though, which they call the structural response hypothesis. This hypothesis states that it is possible that average levels of public opinion remain stable, but that different groups diverge. In other words: polarization does not necessarily mean a change in average levels of permissiveness towards induced abortion. According to the authors, prior studies failed to take that into account.

Using survey data from the General Social Survey (GSS) the authors found that the Court’s decision did indeed influence the public opinion, but not in the way the positive response hypothesis would expect. Roe v. Wade did raise approval of health-related induced abortions, but not for discretionary abortions. Regarding the latter, it was found that the average level of support remained the same, but that attitudes indeed became more polarized, as proposed by the structural response hypotheses. This can be understood based on a model of interpersonal communication: when new legislation is pressed, people talk to their peers about it and thereby form their own opinion. From this, it results that opinions become more homogeneous within groups, and that the relative differences between groups become larger. Moreover, it was found that this increased divergence between groups was stronger for groups that were diverged to a large extent intitially (religious v. non-religious), compared with groups that were not so much diverged (lower and higher educated) on the issue of discretionary abortions.

One methodological nag

I think, overall, that this is a very interesting article, the findings of which should necessarily be taken into account in other studies on public opinion as well, especially when studying polarisation. Sure, they only take into account one indicator of polarisation (diverging group-means). Since the authors only invesitage the direct response to the Roe v. Wade ruling, they neither investigated the emergence of the pro-choice and pro-life movement in the wake of the decision. Nevertheless, with this single measure of polarisation, they are able to convincingly show the empirical support for their structural response hypothesis, in addition to showing the shortcomings of the positive response hypothesis.

Nevertheless, I cannot resist to point out one methodological shortcoming, which may of may not have influenced the findings of this study. Using survey data from the General Social Survey (GSS), the authors create two scales by counting the number of conditions under which respondents would allow a women to have an abortion. In general, this is a doubtful strategy. Since the authors do not test for the feasibility of a so-called Mokken-scale, the ordinal-level scale ((according to the authors)) that they created assumes that acceptance of each condition is independent of the acceptance of other conditions, and that each condition is equally ‘difficult’ to accept. Despite the disntiction between so-called ‘health’ related conditions (abortion to save mothers’ health, after a rape, or in case of a birth defect), and ‘discretionary’ conditions (mother too poor, unwed mother, parents do not want anymore children), which helps to make this assumption less strong, I doubt that these two scales are good representations of respondents’ acceptance of or objections against induced abortion.

Theory of the minor Differences

Now for a completely different story, but also based on what has become a classic paper amongst scholars from a different background than Franklin and Kosaki. Put shortly, the theory of the `Narcissism of the minor differences’ has been described by Anton Blok (1998) as `the idea that identity lies in differences, and difference is asserted, reinforced, and defended against what is closest and represents the greatest threat’ (p. 39). Based on this general statement, it is expected that when differences are small, or are decreasing, people will accentuate these differences, enlarge emphasize them in order to protect their identity. According to the theory, this allows people to distinguish their own identity in a situation in which the similarities are actually much more evident than the differences.

From the perspective of the theory of the Narcissism of the minor differences we cannot decide which differences are large or small. This would require an a-priori judgement on, for instance, whether church members and non-members are much different from each other. The theory does not provide guidelines for this decision. However, it is possible to state expectations on relative differences.

In that case, from the theory of the Narcissism of the minor differences the hypothesis can be derived that polarisation takes place to a larger extent between groups that are relatively close to each other. When legislation comes into effect, which results in a clear, homogeneous, and identical situation for members of all groups, according to the theory people will need to emphasise their own position in the debate more strongly in order to defend their own identity. In other words, diverging attitudes towards abortion after the Roe v. Wade ruling would be expected, especially between groups that were relatively close initially.

However, as it was argued and empirically shown by Franklin and Kosaki: the opposite is true. In the case of the coming to effect of `Roe vs. Wade’ Franklin and Kosaki showed that it are the groups that initially are more different from each other that took increasingly divergent positions in the abortion debate after the legislation came into effect, whereas members of rather similar groups converged. So, in other words, conflict increased between already different groups, in clear contrast with the expectation based on the theory of the Narcissism of the minor differences.


Charles H. Franklin, Liance C. Kosaki (1989). Republican Schoolmaster: The U.S. Supreme Court, Public Opinion, and Abortion The American Political Science Review, 83 (3), 751-771

Anton Blok (1998). The Narcissism of Minor Differences European Journal of Social Theory, 1 (1), 33-56

6 comment on “Republican Schoolmaster and the Narcissism of the Minor Differences

  • Hey, that is a cool thing to think about… would you have a pdf of the Blok paper? Now… imagine that we use searches in google as some sort of measurement of public opinion… and test both papers in the moments before and after Verdonk and before and after Vogelaar… hum….

  • @ Inti: indeed, that would be a very interesting test. Although I don’t really see how Google could provide a proper measure of public opinion. But, I’ll think about it. I read this paper a while ago, but don’t have it. nor can I access it. I do have a highly similar version of the article as a book-chapter (or otherwise on paper). I’ll scan and send it to you soon.

  • The google thing is as follows: you can imagine that if you have a question about something that you do not know, you go and google it. So, if you can follow the dynamics of the hits of certain keywords, you might be actually detecting some proxy to the interest of a population in those keywords along time.

    Haven’t got the reprint yet, but somebody used this approach to evaluate AIDS progress…

  • No, not again, the narcissism of minor differences ;). Nice though to use it as a contrasting hypothesis, but to say it is rejected, I’m sure there are people who’d like to argue on that with you!
    Love from New Zealand

Leave a Reply