William Stuntz died in 2011, before this book could be published. Since then, the high rates of incarceration & crime for which he indicts our criminal justice policy continued to decline (there was a brief uptick attributed to “the Ferguson effect”, but it’s gone back down since then). Many of the fundamental defects he points to remain though, and as with Mark Kleiman’s “When Brute Force Fails” it suggests that we could have much less of both.

While known to be a conservative evangelical Christian (an odd thing for a Harvard Law prof to be), Stuntz is mostly making a “liberal” argument: the system is stacked against criminal defendants, most egregiously for poor blacks. Where he differs from liberal legal scholars is that he regards the attempt by the Warren Court to help such groups as a big mistake. His ideal is the criminal justice system as it first resembled anything like what we’d recognize, in the Gilded Age (though specifically in northern cities). That was a system which dealt with a large influx of immigrants who committed crimes at a higher rate than natives, but it had a homicide rates much lower than those cities would later along with a very low rate of incarceration and a system much more favorable to criminal defendants. Someone like Steve Sailer might note that this was before the Great Migration brought a large influx of African Americans (then mostly living in the south, which Stuntz notes had a higher homicide rate) into those cities, but Stuntz cites Randolph Roth’s “American Homicide” to the effect that whites had higher homicide rates than blacks in both the north & south (which is contrary to what I’d read on the earliest data on homicide rates by race). That something happened other a geographic shift of criminals is plausible given how low the rate of homicide was in (then mostly rural) Ireland, compared to how murderous the Irish immigrants from there were after they arrived in those Gilded Age cities. Like Ed Glaeser, Stuntz is well aware of how those immigrants became politically powerful enough to elect notoriously corrupt representatives like James Curley, but he regards this as a good thing. A large & heavily Irish police force, functioning partly as a mechanism of patronage for urban political machines, managed to keep the peace while treating various “vice” industries & organized crime as a source of payoffs rather than something to be stamped out. Not only did the police resemble the criminals, so did the juries. But even when they didn’t, they resembled the broader class of defendants enough to acquit in circumstances we’d find unacceptable today, disregarding the simple letter of the law for their own notions of fairness. Stuntz writes that “White women and black men achieved surprisingly good results in that older justice system because defendants as a whole achieved good results”. To put it in numerical terms, 22% of turn-of-the-century Chicago homicides led to criminal convictions (this was even lower for wives who killed their husbands). A positive evaluation of this would seem to slight victims compared to ours (though elsewhere Stuntz notes that the inferior Jim Crow system of justice had a whiter incarcerated population than offenders because crimes with black victims tended to go unpunished, and that the failed prosecution of Emmet Till’s murderers was at least an improvement over the past), but the lower overall crime rate meant there were less of them. Stuntz’ claim that African Americans have never held the political clout of Gilded Age immigrants would seem to be falsified in the case of cities like Detroit where the Curley effect has driven most of the whites away, but perhaps he’d respond that white dominance of state & federal government prevents those cities from properly reflecting the political desires of their voters.

The wave of immigrants and the crime they brought with them resulted in an early “culture war”, with legal changes presaging what would happen later. However there were far fewer prosecutions for early laws against drugs, gambling, “transporting women across state lines for immoral purposes” and polygamy (whose enforcement was aided by the desire of Mormons to be law-abiding citizens) than vice laws today. Even Prohibition, which did result in large scale federal enforcement efforts, was carried out with much more transparency and a desire for fairness (Herbert Hoover’s commission noted that enforcement targeted working class pubs rather than upper class clubs, which Stuntz acknowledges may be more socially desirable if not for the reduced respect it led for the law). After noting that it was more successful in its aims than commonly thought of today, he declares “It seems closer to the truth to say that Prohibition was America’s good culture war, one fought by democratic means with reasonably fair and open legal rules.” This transparency (helped by the need for more tax revenue in the Great Depression) is part of why the political decision was made to adopt another amendment striking down the one which created Prohibition, in contrast to all the drug laws today which have no amendments justifying them. Still, these laws represented the displacement of common law crimes with statutory ones whose precision made them easier to prove, frequently used as a substitute for crimes the public actually cares about (like homicide) which happen to be difficult to prove. Many have noted that African Americans make up a far larger proportion of people convicted of drug crimes than among drug users, and Stuntz highlights this without chalking it up simply to racism. Instead he views the timing & demographics of drug enforcement as explainable only with reference to violent crimes. Homicides among whites in smaller towns & rural areas tend to be easy to prove (while drug deals are not) and less frequent, homicides among blacks in inner cities are difficult to prove given gang intimidation (and the hostility of many residents to police) and frequent, whereas it is easy to bring up arrest numbers via buy & busts at open-air drug markets or crack houses (which police hate than more discrete drug trafficking due to their connection with violent crime). Mark Kleiman & Stuntz have very different opinions of “Operation Ceasefire”, in which gangs were presented with evidence sufficient to convict them of trafficking drugs and/or guns and were threatened with arrests conditional on them being involved in shooting incidents with other gangs. Stuntz decries this as an inefficient & indirect means of targeting the crime police actually care about (gun violence), and by going after a collective rather than individuals leave said individuals with little of a marginal incentive not to fire off shots (he wants to expand the doctrine of racketeering law that the least culpable member of a group is excused from liability for the most serious offesnes). Kleiman is satisfied enough with the results to conclude that these gangs are able to discipline their members (which may not be the case with the increasingly fragmented & violent gangs of Chicago’s far west & south sides). On the other hand, Stuntz is fine with injunctions prohibiting gang members from meeting with each other, as they just result in short stays in local jails on contempt charges rather than long prison sentences. Stuntz also dislikes discretionary police stops, whose decline in Chicago following an ACLU consent decree has been blamed for a spike in gun crime, similar to the decline which sociologist and former cop Peter Moskos (also a supporter of drug law abolition & devil’s advocate for replacing prison with flogging) blames for a similar spike in Baltimore, although I suppose Stuntz isn’t around to respond.

It’s not surprising that Stuntz disagrees with Kleiman, Moskos and Paul Cassell, none of whom are particularly well known even now. What is surprising is how directly he targets one of the most sacred cows in American political history: James Madison’s Bill of Rights. Stuntz doesn’t have a high opinion of the thinking behind it, but at least early in American history it had little effect as few federal cases were overturned based on it. The real trouble came when the “incorporation doctrine” resulted in it being applied to state & local cases, with the Warren Court particularly blamed for the approach it took in response to what Stuntz also views as a real problem. He points to the French Declaration of the Rights of Man as being preferable to the Bill of Rights due to its emphasis on the substantive rather than procedural limits of the law. In his view, procedure is a matter of pragmatically responding to empirical reality, whereas the substance of the law is supposed to be based on political principles. The former should flexibly respond to changing conditions (as with the “living constitution”), while the latter is cemented into constitutional law. The Bill of Rights was written before a professional police force even existed, and the people who wrote it could not anticipate how it would be applied. Stuntz thinks it’s revered today by politicians in part because its procedural constraints are so easily maneuvered around by the actual institutions of criminal justice. I think of the enumerated powers in the constitution as also placing substantive limits on legislation (hence Prohibition requiring an amendment) which were ignored when it was sufficiently politically popular to do so, and since the French regime which enacted that Declaration didn’t persist it’s harder to point to as a more successful example. Stuntz views the Warren court’s errors as answering three questions in the least helpful way: whether to focus on the substantive legislation that could define crimes or the process by which they were enforced, whether to devise anti-discrimination rules to explicitly advance the goal of equal protection in the face of discriminatory justice system or to adopt facially neutral rules which would assist all defendants, and finally whether to expand on the loose “due process” jurisdiction which had struck overturned certain verdicts in the south influenced by mob violence or the federal Bill of Rights cases arising out of the turn of the century culture war (whose defendants had very little cause for procedural complaints but enough money to hire lawyers the courts granted a small amount of credence to before affirming their convictions).

The Supreme Court attempted to assist defendants, particularly indigent & black ones in the south (which Stuntz noted long had a more dysfunctional justice system), but the way that it did so was to place greater scrutiny on the behavior of police officers (who eventually learned how to get around such rules, particularly once the “law & order” backlash against the Warren court eased pressure on them) rather than defendants (whose behavior had led juries to acquit many defendants in Stuntz’ ideal system). The defendants who benefited most tended to be habitual criminals who knew not to talk to the police without lawyers (only 20-25% of suspects persistently decline to talk without a lawyer) especially those who could afford better lawyers, putting blockages in the way of law enforcement no more likely to help the innocent over the guilty than evidentiary rules which only made a difference if officers found something they would want to present in court. “Suppression of illegally seized evidence is a particularly poor sanction to apply to the police, because police officers lose nothing when evidence is excluded at defendants’ trials. Most criminal litigation is like a relay race: police identify and investigate the suspect, make the arrest, and then hand off the case to prosecutors. The exclusionary rule penalizes the anchor in this race; prosecutors and the public suffer for police officers’ mistakes.” These procedural rules which tended to free the “mistreated” guilty rather than railroaded innocents (like many of the blacks in the south the court was concerned with), particularly during the great crime wave of the 60s, set the stage for a political backlash. People like J. Edgar Hoover, Thomas Dewey, Estes Kefauver, William O. Douglas, Joe McCarthy and Robert F. Kennedy all attained national celebrity by positioning themselves as anti-crime crusaders (of various sorts). With the 60s crime wave & controversial court decisions which appeared to encourage it, street crime became a national issue and more local politicians like governors Ronald Reagan & George Wallace were able to posture as “tough on crime”. Without the power to overturn such decisions, this posturing was symbolic (and incarceration actually fell under both governors), which was ideal for politicians who want to avoid responsibility. At the same time, protections for defendants became increasingly symbolic, as the higher cost of trials resulted in prosecutors aiming for guilty pleas instead (which now account for 95% of all convictions), assisted by more statutory laws to charge defendants with. The state appointed defense attorneys guaranteed by Gideon visited crime scenes in 12% of homicide (and 4% of all) cases, interviewed witnesses in 21% (4%) and used experts 17% (2%), but filed motions in 26% (11%). This focus on motions rather than things relevant for guilt or innocence is not in anyone’s interest, but is the result of the Warren court’s revolution. The increase in crimes made it difficult for states to budget for the burden that had been placed on them, and per-case spending fell by half between 1979 and 1990 on indigent defense (the reliance on plea bargains over full trials is one way to reduce costs).

Just as there had been changes in crime & criminal law over that time, there had also been changes in local politics. Whereas blacks had once been swing voters that Republicans tried to court, after JFK & LBJ they were solidly in the Democratic camp and blue collar whites became the more competitive constituency. At the beginning of the century, most residents of a metropolitan area lived in the city itself, but with the growth of suburbs (accelerated by white flight in response to crime) and concentration of crime into black ghettos, politicians dealing with crime were frequently representing people with the least contact with it. Since middle-class whites were more likely to flee to the suburbs, taking their tax-dollars with them, local politicians catered more to them than their numbers might suggest. The shift had already begun in the low-crime midcentury, and this insulation from the realities of crime is one of the things which had caused the criminal justice system to prioritize the norms of a professionalized bureaucracy over that of constituents and lower the incarceration rate even while crime rose. The backlash against this supposedly apolitical and “expert-based” criminal justice also continued past the point when crime fell far below any level fitting the rate of incarceration, because it remained insulated and with the people in the most high-crime neighborhoods having the least influence. Thus they pay for the practical results of laws that politicians pass for symbolic reasons. The incentives of politicians are further skewed because local government pays the full cost of police*, but state governments pay the cost of incarceration, so the will overly rely on the latter rather than the former. Hence one of his suggested proposals.
*In Canada & Germany it’s states/provinces, whereas in the UK & France it’s the national government.

Stuntz recommends that local governments should pay half the cost of the inmates they send to state prisons, so as to internalize at least some of the externality. At the same time he would prefer if they only had to pay half the cost of their police payroll, with some combination of the states & federal government picking up the rest of the slack, which is somewhat similar to how education used to be (and would have been for police if one of the crime bills during the Clinton administration hadn’t gotten watered down). Housing a person in prison is cheap relative to hiring an additional police officer, but per Steve Levitt a 1% drop in crime nationally would require $700 to $840 million dollars, while the amount of additional incarceration to get the same effect would be $1.6 billion (Mark Kleiman thinks that at our current margins, additional incarceration actually causes more crime). But since the largest increase in discretion & power has been for prosecutors, he seeks to rein them in more directly rather than merely hoping their behavior changes once arrests go down (which has been the counter-intuitive result of expanding the police force from 1990 to 2003, as urban arrests fell by 22% and the rate for black suspects fell even more). He advocates the rule that for all sentences of incarceration beyond a certain minimum, prosecutors should be required to show that this is a normal sentence in that state for that crime given similar facts (accepting Anthony Kennedy’s logic on rarity showing a sentence to be “cruel & unusual”). Stuntz intends for this to be used against “pretextual” crimes charged against suspects of more severe crimes that can’t be used, which I expect would be a very unpopular change. Such changes wouldn’t constrain judges, who had so much less power in the past that ones on the Supreme Court used to resign in order to run for offices they found more significant. Here he actually likes the degree of discretion federal judges have after US v Booker struck down the federal sentencing guidelines, which practically made them a ceiling which judges could reduce, and wants state judges to follow a similar model. Multiple SCOTUS decisions should be overturned: the Miranda decision which prioritized the rules of interrogation over the facts involved should be replaced with the requirement that interrogations be videotaped and presented as evidenced. McClesky v Kemp and United States v Armstrong, which limited charges of discrimination regarding crime victims & defendants (respectively) should simply be overturned*. The confrontation clause (which Scalia of all people made a lot of pro-defendant decisions with) should be discarded to permit evidence other than live testimony, such as is common in continental europe (which has a less adversarial justice system which purports to focus on neutral fact-finding), preventing intimidation of witnesses and enabling a refocus on the violent crimes we actually care about rather than detailed statutory crimes that make for easier convictions. Procedural rules imposed on all law enforcement should be disfavored and replaced with institutional injunctions against those governmental bodies which violate the 14th amendment’s guarantee of “equal protection”, like the injunctions used to mandate school bussing and decrease prison overcrowding. Given the political backlash against bussing and withdrawal of white students from heavily minority school systems, I’d be skeptical of leaving that to the courts rather than the enforcement acts that amendment leaves to Congress (on the other hand, the persistent refusal of Congress to update the Voting Rights Act’s geographic designations doesn’t speak well of its responsibility either). He also thinks as a last resort judges may have to mandate certain forms of training to correct discriminatory policing, having never gotten to read Jim Manzi’s analysis of the dismal experimental record of training programs vs crude incentives. Sixth amendment jurisprudence should cease focusing on the quality of defense lawyering and look into quantity (ratio of lawyers to defendants), although the economist in me thinks that it would be difficult to maintain a standard of quality in the workforce while ramping up the supply. He also proposes that “in all jurisdictions that set up expert commissions to recommend appropriate funding for indigent criminal defense and then follow those recommendations, effective-assistance-of-counsel doctrine will not apply. Elsewhere, [assistance standards] will be ratcheted up sharply.” This granting of waivers actually does sound like something Manzi would approve of. Jury selection should be even more localized by basing it on neighborhoods (I don’t think this rule would work for less populous regions, as existing rules already enable the perfect crime in part of Yellowstone), and peremptory challenges reduced. His implicit hope is that the communities most affected by over-incarceration will engage in jury nullification (as was more common in the Gilded Era), and if such nullification was limited to drug laws I might be sympathetic, but I have read the account of a juror who caused the mistrial of a homicide case because he shared Stuntz’ outrage. Stuntz also favors the more traditional form of federalism by which federal & state law should not overlap so much, recommending both that local prosecution of federal crimes should have sentencing fixed by state rather than federal law**, and federal prosecution should be strictly limited to crimes not enforced by lower levels of government rather than taken up on a discretionary case-by-case basis.
*Stuntz acknowledges that African Americans commit certain crimes at higher rates and that arrest, conviction & incarceration rates should reflect this, so I’m not entirely sure how his embrace of “disparate impact” is supposed to work.
**Because there’s so much less prosecution at the federal level, there’s less cost to politicians “overproducing” it and it tends to be written even more out of symbolism than pragmatism.

There’s a lot more in the book I haven’t covered (like his view, contra Scalia’s textualism & Herbert Wechsler’s rationalized Model Penal Code, that vagueness in criminal law is a good thing), but I actually finished reading it weeks ago and have been starting to feel like I’d never finish writing this. You’ve just read me express skepticism at a number of things Stuntz wrote, but part of that is that this would otherwise too much resemble an infodump in which I just spewed out rephrased versions of what he said. Overall I appreciated the book as an unusual perspective and more than the common howl of outrage at the status quo. It should certainly be supplemented with some of the criminologist perspectives I’ve mentioned, but the history of how we got here is valuable if there’s to be hope of avoiding making more such whiggish mistakes in any future attempt to reform the system (which the “Ferguson effect” may have helped delay for some time).