Weisberg v. Kahan

In anticipation of next week’s visit by Robert Weisberg, I reread one of his articles that made a big impression on me when it first appeared a few years ago: “Norms and Criminal Law, and the Norms of Criminal Law Scholarship,” 93 J. Crim. L. & Criminology 467 (2003).  This is basically an extended — perhaps even a little too extended, at 125 pages — critique of the scholarship of Dan Kahan.  Kahan has had a huge influence on many other criminal law scholars (I’ll include myself among those influenced), so Weisberg is taking on a worthy adversary.  And, on the whole, I think his criticisms hit the mark.

(Kahan himself, I should note, delivered a fascinating lecture at Marquette Law School not so long ago.)

In order to appreciate the significance of Kahan’s agenda, it is helpful to think about it (as Weisberg does) as a response to the law and economics movement.

In the classic L&E model, prospective criminals are rational cost-benefit balancers.  Because increasing penaltes also enhances the expected costs of committing a crime, crime rates should more-or-less predictably go down as sentences go up.  This straightforward predictive claim has important implications for criminal-justice policy.  Applied crudely — and how else will any punishment theory be applied by the political system? — the L&E model seems to justify limitless increases in penal severity, at least during periods of rising crime.

The L&E approach faces two major sets of criticisms.  The first is ethical.  The L&E model seems to have no place for rights, at least to the extent that rights are not cost-benefit justified.  Although we might view various kinds of acts, such as criticizing the government or worshiping the deity of our choice or having sex with another consenting adult, as absolutely protected from government interference, the L&E model offers no a priori assurance that these acts will not be criminalized.  Nor does the L&E model ensure that the severity of the punishment will be made proportionate to the gravity of the crime.  In principle, minor crimes might be punished much more severely than major crimes, depending on how the cost-benefit calculus plays out.

The second set of objections — and the ones more pertinent to the Kahan-Weisberg debate — are practical in nature.  The claim, in essence, is that the L&E model rests on an overly simplistic view of human psychology.  We are not the rational cost-benefit balancers that L&E assumes.  We process new information poorly and badly misestimate probabilities.  We are also subject to all sorts of social influences that cause us to act in ways that might seem irrational or self-destructive from the L&E perspective.  Once these considerations are taken into account, we can better appreciate why crime rates do not respond in simple, predictable ways to changes in punishment.  Increased severity in the 1970s and 1980s did not prevent crime rates from climbing in those decades.  Meanwhile, crime rates have dropped dramatically over the past twenty years, despite a leveling off and perhaps even a slight retreat from the severity peaks reached in the 1990s.  These patterns at least pose a challenge to the predictive claims of the conventional L&E model.

One way in which I think Kahan has made a valuable contribution is in drawing attention to a wide range of psychological and sociological factors that bear on crime rates.  He has helped us (or me, at least) better understand the limitations of the L&E model, and given us a new vocabulary, borrowed from the social sciences, to describe and evaluate the effects of new criminal-justice policies.

I don’t think Weisberg would disagree that the L&E model is overly simplistic and that many of the psychological and sociological factors highlighted by Kahan play some sort of role in the complex relationship between crime rates and legal responses to crime.  He does, however, question the novelty of these insights, finding key antecedents in the work of the Legal Realists and their intellectual heirs.  That may be so — and, indeed, I think much of the power of Kahan’s claims lies in their intuitive quality — but perhaps we can still credit Kahan with drawing attention to some important ideas that are often neglected in the political and policymaking realms, and articulating those ideas in new ways that may help to bridge the disciplinary divide between legal scholars and social scientists.

Where I think Weisberg’s critique is more compelling is in his treatment of Kahan’s prescriptive program.  It would be one thing if Kahan’s project were merely defensive, so to speak — undermining or adding necessary nuance to the L&E model.  But Kahan plays offense, too, advancing an ambitious program of legal reform.  He has been a high-profile supporter of community policing, broken-windows policing, antiloitering laws, and shaming sanctions, as well as an outspoken critic of constitutional interpretations that impair the ability of the police to address the needs of high-crime urban neighborhoods.  (See my post here.)

What links together many of the planks of Kahan’s reform program is a line of reasoning that goes something like this: (a) the key to crime control at the neighborhood level is the sense of trust that neighbors have in one another and in the police; (b) trust is undermined, and crime flourishes, when residents lose the sense that there are shared norms of good conduct in the neighborhood or that norm-violators will be effectively sanctioned; and (c) police, and the legal system more generally, can and should work to reinforce community norms and build confidence that violators will be sanctioned, even in relatively small matters, so as to build trust and cooperation.

The concept of “social meaning” does much of the work here.  Individual acts of criminality and individual responses to crime convey broader messages about the content and importance of community norms.  The task for police and for the legal system is ensure that responses to crime convey the right sorts of social meanings.

The problem with all of this, as Weisberg demonstrates, is that the underlying social science does not actually provide much support for the sort of ambitious reform program that Kahan has in mind.  To be sure, there are studies that suggest that fear of legal sanctions is not the only, or even the most important, reason that people obey the law.  Likewise, neighborhoods do vary in their crime rates, and relatively low-crime neighborhoods do seem to display a more generalized sense of trust, mutual support, and shared norms.  But, as the saying goes, correlation does not imply causation.  It is not clear to what extent low crime causes or is caused by other positive neighborhood characteristics, nor the extent to which both types of variables result from still other factors.

And even if the causal link were established — that is, low crime results from a sense of strong shared norms — we would still be left with little basis for knowing whether and how law and policing can generate the right sort of social meaning so as to produce the desired neighborhood characteristics.

Indeed, in claiming to discern the “meaning” that is conveyed by innovative police practices and legal reforms, Kahan seems to fall victim to the same sort of reductive psychology that has undermined L&E.  It seems more likely that police practices lack any fixed, determinate, widely shared meaning.

What does it “mean,” for instance, when police dramatically increase their number of car stops in minority neighborhoods, as the Milwaukee police have done in recent years?  Probably quite different things to whites and blacks, to drivers and non-drivers, and to those who fear criminals more than the police and those whose fears run in the other direction.  And those meanings are likely to shift over time, as crime rates go up or down, as people grow accustomed to the new practice, and as the inevitable police errors and overreactions are publicized in the media.  Positive meanings can slide into negative, and then back to positive — or maybe a police practice simply becomes part of the background noise in a community, and in some sense becomes “meaningless.”

I suspect that Kahan himself would today agree that he spoke overconfidently about social meaning earlier in his career.  Indeed, since Weisberg’s critique, Kahan has explicitly disavowed his earlier support of shaming sanctions, citing the diverse and contested meanings of shaming.  I think Kahan’s even more recent work on cultural cognition reflects an increasingly rich appreciation of the multiplicity meanings that may be generated by a single legal or policy decision.

Back to early Kahan, Weisberg’s critique makes at least one additional point that bears note: Kahan’s emphasis on supporting community norms through policing and the legal system may, in practice, be hard to distinguish from simply giving the criminal-justice system over to populist anticrime sentiments, which are apt to be excessively harsh and inconsistent with the robust protection of individual rights.  Weisberg does not really push the point, but one implication of his critique is that Kahan’s approach may be no less vulnerable than the L&E model to the sort of ethical objections that I identified earlier in this post.