Implementing decision A may also lead people to see B as less extreme and thus more acceptable. When we’re at position 0 (no handgun ban), the leading policy options may be 0, A (a ban on small, cheap handguns), and B (a total handgun ban), and B may seem like a large step. But after A is adopted, the leading options may become A (the narrow handgun ban), B (the total handgun ban), and C (a ban on all firearms, whether handguns, rifles, or shotguns), and B may thus seem more moderate; position 0 might no longer be considered, because it’s been tried and rejected.
In principle, such framing effects—whether B is seen as the extreme option among 0, A, and B or as the middle option among A, B, and C—should be irrelevant. When the choice is between A and B, people shouldn’t be influenced by the presence of options 0 or C.
But social psychologists have shown that people do tend to view proposals more favorably if they are presented as compromises between two more extreme positions. In one experiment, for instance, one group of subjects was asked to decide which of two cameras, a low-end model and a mid-level model, was the better deal; 50% chose the mid-level as the better deal. Another group was asked to choose among the same two cameras plus a high-end model; in this group, the mid-level was favored over the low-end by over two-and-a-half to one.
The result may seem irrational; the addition of the new option might reasonably decrease the fraction of people choosing either of the other two options, but it shouldn’t increase the relative fraction preferring the mid-level option. At the very least it reflects bounded rationality. But in any event, that’s the result, which has been replicated for legal decisions by mock juries. And it fits our experience: people are often (though not always) more sympathetic to options framed as “moderate” than to those framed as “extreme.” To the extent this phenomenon occurs among voters, it can produce slippery slope effects, as the enactment of even modest steps makes a formerly extreme proposal seem more moderate.
Let me close out this section with a final question: Are Attitude-Altering Slippery Slopes Good or Bad?
When a decision A alters people’s attitudes about B, this alteration may be part of a good learning process. People might, for instance, initially oppose a broad market in human organs (B), but once they see that a limited market (A) works well, they may change their views. And if we decide our initial aversion to B might be mistaken, we might want to try A and see if we learn something from it. “The spirit of liberty is the spirit which is not too sure that it is right,” wrote Learned Hand, and surely this is even more true of the spirit of sound policy analysis.
The danger, though, is that our experience with A might leave our aversion to B unchanged, but might lead others—in our view, erroneously—to support B. After all, what some people call good “learning” is precisely what others might call bad “desensitization.” Maybe being confronted with happy beneficiaries of an organ market will lead our fellow voters to underestimate the moral harms of such markets. If that’s so, then we might regret having supported A in the first place, because it would have indeed brought about a B that we continue to oppose.
This approach might at first seem improperly paternalistic or anti-majoritarian, but it simply reflects political reality. We want the political process to reach results we like. Sometimes it doesn’t reach those results, because others disagree with us. In our view, those people are mistaken; but their votes can force their mistake on us. So if we do think that implementing A would lead others to support B while we ourselves would continue to oppose B, that’s a reason for us to oppose A.
Recall the question at the heart of this article: “Does it make sense for me to support A, given that it might lead others to support B?” The possibility of good attitude-altering slippery slopes shows that even if we oppose B, we might still endorse A if (1) we have reason to doubt our judgment that B is bad, and (2) we are fairly confident that if A persuades our fellow decisionmakers that B is good, it will also persuade us. But if these conditions aren’t met, then we’re back to the slippery slope that we’d like to avoid rather than embrace.