Criminal Law

Index to Sections on Criminal Law

In Chapter 9 we consider the economic case for imposing a civil duty to rescue. For example, the law might impose liability for failure to effect an easy rescue: the person who failed to help might be held liable for the losses that the person who might have been rescued suffered.

    There is a criminal-law equivalent of the civil law duty to rescue. Suppose that someone observes his neighbor committing a crime. The law might enlist his aid in deterring crime and apprehending criminals by imposing a duty on him to intervene in the crime or to report his knowledge of the crime or criminal to the authorities. Several states have enacted duty-to-rescue and duty-to-report laws. For example, Mass. Gen. Law Ann. Ch. 268, § 40 (West 1990): "Whoever knows that another person is a victim of aggravated rape, rape, murder, manslaughter or armed robbery and is at the scene of said crime shall, to the extent that said person can do so without danger or peril to himself or others, report said crime to an appropriate law enforcement official as soon as reasonably practicable." Or Ohio Rev. Code Ann. § 2921.22 (West 1997): "No person, knowing that a felony has been or is being committed, shall knowingly fail to report such information to law enforcement authorities."

    At first blush one might think that these duties would be efficient: witnesses would be induced to come forward, making the successful prosecution of criminals easier, and, possibly, criminals, knowing that witnesses had a duty to come forward, might be marginally deterred from committing crimes. But in a recent article Professor Eugene Volokh of the UCLA Law School ("Duties to Rescue and the Anticooperative Effects of Law," 88 Georgetown L.J. 105 (1999)) argues that a duty-to-report/rescue law may make things worse rather than better. To see his intuition, imagine that Alex sees his neighbor criminally abusing a child. Alex neither intervenes nor reports the incident. He has, thereby, violated a duty-to-report/rescue statute. "By the time the remorse sets in, [Alex is] legally guilty, and either volunteering or honestly answering police questions will incriminate [him]. … [Alex therefore has] a choice: continue to keep quiet and likely never be conclusively found out to have been a witness, or speak up and risk a criminal prosecution or perhaps a civil lawsuit for [his] initial failure to act." (107-108). The result is that "in trying to achieve a procooperative effect, duty-to-rescue/report laws may inadvertently cause an anticooperative effect." (108)

    Professor Volokh does not insist that every duty-to-report/rescue law will necessarily have an anticooperative effect. For example, "[t]here will [] be no anticooperative effect when people do not know about the duty to rescue or to report." (109) And he also recognizes that law enforcement authorities can induce testimony from someone who has knowledge of crime but has not yet come forward by giving them a grant of immunity in exchange for their testimony. But that bargained-for testimony may not have the same credibility as freely given testimony.

    Professor Volokh reports that his colleague, Professor Peter Arenella, suggests adding a "second chance" defense to the duty-to-report/rescue statutes: the witness would be absolved from liability if they (eventually) volunteered information or answered questions.

    In situations other than crime, a duty to report or rescue may work well. For example, many states require doctors to report to the police any gunshot or stab wounds that they treat. (See Cal. Penal Code §§ 11160-11161 (Deering 1991).) Professor Volokh contrasts this duty to the duty to report knowledge of a crime: "Most witnesses who fail to rescue or report act out of callousness, fear, or deep-seated loyalty to family, friends, or confederates; it’s unlikely that knowing about a legal duty to rescue or report will have much normative effect on their behavior. Doctors who fail to report gunshot or stab wounds, on the other hand, are likely to be acting that way because they feel a less entrenched professional norm of loyalty to their patients. A duty-to-report law is especially likely to undermine this norm." (112)

See also Daniel B. Yeager, "A Radical Community of Aid: A Rejoinder to Opponents of Affirmative Duties to Help Strangers," 71 Wash. U. L. Q. 1 (1993).

Questions:

1. Referring to our discussion in Chapter 9, do you think that there would be a greater procooperative or anticooperative effect of creating a duty to effect an easy rescue in tort law?

2. Could a witness with knowledge of a crime avoid prosecution by submitting an anonymous tip to law enforcement officials?

3. Are illegal aliens and prostitutes who are the victims of crime or have observed crimes likely to obey duty-to-report statutes?  

Back to Index of Sections on Criminal Law

A Question about the Duty to Report

In May 1997, high school senior Jeremy Strohmeyer and his best friend, David Cash, Jr., went to Nevada to gamble.  While in the Primadonna Resort and Casino in Primm, Nevada, Strohmeyer accosted seven-year-old Sherrice Iverson in a restroom at the casino and raped and strangled her to death.  Cash observed Strohmeyer’s initial assault of the child but did not attempt to intervene.  Instead, he exited the restroom after two minutes and waited for his friend in the casino.  Cash failed to notify either the police or any casino employee of Strohmeyer’s actions.  Twenty-four minutes later, Strohmeyer emerged from the restroom.  He told Cash that he had sexually assaulted and killed Sherrice.  The two men gambled for a while longer and then drove home. 

    Some time later, Strohmeyer was arrested for his vicious attack.  To avoid a potential death sentence, Strohmeyer pled guilty to kidnapping, sexual assault, and murder.  A Las Vegas judge then sentenced him to life in prison without parole.  (67 Fordham L. Rev. 3169 (May 1999).) 

    David Cash was interviewed on the news program 60 Minutes soon after Strohmeyer’s sentencing.  He told the interviewer, Ed Bradley, “I don’t feel there is much I could have done differently.”  Later in the interview he stated,  “So you know, in my opinion, it was like time for me to get out of there.  When an 18-year-old male grabs a 7-year old child ... that’s not a position I want to be in.  It wasn’t something I wanted to stick around and, you know, see what would materialize.” 

    David Cash was never brought up on charges for his lack of action.  Indeed, it is probably the case that he did not violate any law.  In Nevada, as well as in 46 other states, including California, failure to inform authorities of a crime is not an offense.  Naturally, many were outraged that Cash should have walked away from this heinous situation without intervening directly or reporting it to the authorities.  A bill to correct this possible omission in the criminal law -- the Sherrice Iverson Act -- was introduced in the 105^th Congress but expired without action.  That bill was reintroduced on April 14, 1999. 

    The question that this incident raises is this, “Should there be a duty to report knowledge of a crime?”  Consider that question in light of the argument in the Article by Professor Volokh (discussed in the previous section of this page) and the facts in the case of David Cash. 

Back to Index of Sections on Criminal Law

Empirical Studies of the Economics of Crime and Punishment

One of the themes that distinguishes the economics of crime and punishment from the other subjects that we examine in Law and Economics is that there is a wealth of interesting empirical work on criminal law issues.  We survey some of that work in Chapter 12 -- e.g., the literature on the deterrent effect of capital punishment and the Wilson & Abrahamse work on boundedly rational decision-making among criminals.  

    In this section of the Web site, we shall be adding summaries of and references to new material on the empirical study of crime and punishment.  

Juvenile Crime and Punishment

An issue that arises in criminal justice policy discussions is whether juveniles respond to criminal justice system variables in the same way in which adults do.  (We summarize some of this literature in Chapter 12).  In an important recent article Professor Steven D. Levitt of the Department of Economics at the University of Chicago re-examines that question—“Juvenile Crime and Punishment,” 106 J. Political Economy 1156 (1998).  He begins with some casual evidence on the possible deterrent (or incapacitation) effect of imprisonment on crime by juveniles: 

The rate at which juveniles were arrested for violent crime rose 79 percent between 1978 and 1993, almost three times the increase over that time period for adults.  The divergence in murder is even more striking: over that same period, juvenile murder arrests rose 177 percent, whereas the murder arrest rate for adults actually fell 7 percent.  For property crime, both juvenile and adult arrest rates have been relatively steady.  (1156)  

    Naturally, the question arises as to why there should be such a remarkable rise in the violent crime rate among youth.  One theory is that the restraints imposed by the bonds of family and society have grown weaker, so that young men (the principal committers of crimes in any society) are less deterred than they have been.  Indeed, some commentators—such as William J. Bennett, John J. DiIulio, Jr., and John P Walters, Body Count: Moral Poverty—and How to Win America’s War Against Crime and Drugs (1996)—have suggested that some young males have become “super-predators, who are

radically impulsive, brutally remorseless, including even more preteenage boys, who murder, assault, rape, rob, burglarize, deal deadly drugs, join gun-toting guns, and create serious communal disorders.  They do not fear the stigma of arrest, the pains of imprisonment, or the pangs of conscience.  They perceive hardly any relationship between doing right (or wrong) now and being rewarded (punished) for it later.  (27) 

    Professor Levitt begins with a far simpler and less dire explanation for the pattern of juvenile crime: rational responses to changing expected punishment.  Recall that the economic theory of the decision to commit a crime imagines that criminals compare the expected costs and expected benefits of crime and commit a crime if the expected benefits exceed the expected costs and refrain from crime if the reverse is true.  Professor Levitt points out that the expected criminal punishments for adults and juveniles may have changed so as to be a complete explanation for the reported patterns: 

 [T]he divergence of juvenile and adult crime rates may not be the result of teenagers who differ categorically from earlier generations, but rather a rational response to a change in the relative incentives for juveniles and adults to engage in criminal activities.  …  Between 1978 and 1993, the ratio of adult state and federal prisoners per violent crime committed by adults in that year (a rough proxy for the punitiveness of the criminal justice system) grew from .34 to .55, an increase of over 60 percent.  The corresponding ratio for juveniles fell from .36 to .29 a decline of 20 percent.  Juvenile punishments, at least by this crude measure, were comparable to adult punishments in 1978, but were only about half as severe by 1993.  (1157) 

    But Professor Levitt is far too careful to leave the matter at this otherwise convincing evidence.  He notes that there are at least three alternatives to the expected-punishment explanation, and he proposes looking at these alternatives before fully accepting the expected-punishment explanation for the pattern of juvenile crime. 

    First, it might be the case that there is a subtle relationship between the juvenile crime rate and the expected-punishment rate for adults.  For instance, if adults and juveniles are substitute perpetrators for some crimes (and there are limited opportunities for crime), then an increase in the expected punishment of crime for adults could lead to a decline in the adult crime rate and a substitution of juvenile crimes for those now-fewer adult crimes.  Or it could be the case that juvenile and adult crime are complements.  If, for example, juveniles model their behavior on adults and if the expected punishment of adults increases so that the adult crime rate falls, the juvenile crime rate might fall, too.  There does not seem to be any clear theoretical resolution to this issue of whether adult and juvenile crime are substitutes or complements.  The issue can only be resolved empirically, as we shall see. 

    Second, there could important intertemporal choice issues.  In deciding whether to commit crimes, juveniles may consider not only the immediate expected punishment but also the future returns to a continued commitment to crime.  This consideration would be more important to the extent that commission of a crime has a learning-by-doing aspect or involves an investment in crime-specific human capital.  So, although there might not be a direct relationship between the adult expected punishment for crime and the juvenile crime rate, there could be this indirect effect: a juvenile today might not invest in a “life of crime” (even if the expected punishment of juveniles was light) if he perceives the returns to this life to be low for adults.  As with the first alternative to the expected-punishment explanation, this one, too, demands empirical investigation.  Theory alone cannot tell us whether this factor is significant.  

    Third, the imposition of punitive sanctions on juveniles may have independent and contradictory short- and long-run effects on the decisions of juveniles as between criminal and legitimate activity.  Suppose that harsh or more certain punishment deters juvenile crime in the short run.  But also suppose that a criminal record stigmatizes the juvenile offender so that legitimate employment in the future is less likely.  These suppositions could mean that even though harsh punishment reduces juvenile crime in the short-run, it might increase adult crime as the convicted juveniles grow older and find that their reduced legitimate opportunities in the future increase the relative returns of adult crime.  As before, whether this effect is present and significant is something that we can resolve only by careful empirical work. 

    Professor Levitt tests each of these alternatives.  Here are his conclusions: 

First, when state-level panel data for the period 1978-93 are used, harsher punishments for juveniles are strongly associated with lower rates of juvenile offending.  …  There does not, however, seem to be a high degree of substitutability between juvenile and adult crime; there is no systematic response of juvenile crime to the expected punishment facing adults, or vice versa.  Nor does there appear to be a strong relationship between the punitiveness of the juvenile justice system that a cohort faces when young and the extent of criminal involvement for that cohort later in life.  (1159)

    Levitt also sheds light on another persistent issue in the economics of crime and punishment, whether punishment reduces crime through deterrence or incapacitation: 

Evidence that a substantial fraction of the crime reduction results from deterrence (and not simply incapacitation) comes from analysis of changes in crime rates around the age of majority.  States in which juvenile punishments are lenient relative to adult punishments see much greater declines (or smaller increases) in crime as a cohort passes to the adult court.  For example, in states in which the juvenile courts are most lenient vis-à-vis the adult courts, violent crimes committed by a cohort fall by 3.8 percent on average when the age of majority is reached.  In contrast, violent crimes rise 23.1 percent with passage to the adult criminal justice system in those states in which the juvenile courts are relatively harsh compared to the adult court.  Similar but less extreme patterns are observed for property crimes.  (1159) 

    What are the implications of this very important work for criminal justice policy toward juveniles?  Professor Levitt recognizes that those implications are nuanced.  The results would seem to suggest that an effective method of deterring juvenile crime (with relatively few unintended consequences) would be to make expected punishment more certain and harsher.  But there is the important caveat that incarceration resources are limited so that there may be a decision to be made as between incarcerating one more adult or one more juvenile.  If so, then there must be a cost-benefit calculation of these alternatives.  Professor Levitt suggests that the annual cost of incarcerating a juvenile ($33,000, according to the Office of Juvenile Justice and Delinquency Prevention, 1995) is substantially greater than that of incarcerating an adult ($23,000, according to the U.S. Department of Justice, 1995).  But he also notes that the social benefit of saving a high-risk juvenile offender is substantial—$2 million, according to Mark Cohen.  (Cohen, “The Monetary Value of Saving a High Risk Youth,” Vanderbilt University School of Law, 1995.)   

Back to Index of Sections on Criminal Law

Legalized Abortion and Crime

Rarely does academic work rate attention by the popular press and commentary on the national television networks.  And rarer still is it that academic work rates this attention before it has even been published.  Both of those feats distinguish a fascinating piece of work by John J. Donohue III of the Stanford Law School and Steven D. Levitt of the Department of Economics, University of Chicago, and the American Bar Foundation.  In the late Summer, 1999, they circulated a draft paper entitled "Legalized Abortion and Crime."  The paper's central argument, which we summarize below, was so novel that The New York Times and other prominent newspapers wrote articles about it.  Even legal scholars who pay little attention to law and economics are aware of Donohue and Levitt's article.  It is a remarkable piece of scholarship.  (John J. Donohue III and Steven D. Levitt, “Legalized Abortion and Crime,” Stanford Law School, Public Law & Legal Theory Paper No. 1; and Stanford Law School, Olin Law & Economics Paper No. 17 (September, 1999).  Posted at www.ssrn.com on August 17, 1999.) 

    The focus of the Donohue and Levitt article is an explanation for the dramatic decline in crime rates in the United States since 1991.  "Between 1991 and 1997, crime fell substantially.  Murder rates dropped 30 percent; property and violent crime each fell between 15 and 20 percent."  (37)  There have been many theories to explain those declines -- many of which were published in the 1998 Summer issue (Volume 88) of the Journal of Criminal Law and Criminology.  Among those theories are these: (1) that the decline is a response to the increased use of imprisonment during the 1980s; (2) that it is attributable to the increase in the number of police and more novel police strategies (such as "community policing," said to be particularly relevant to explaining why crime declined in New York City); (3) that the market for illegal drugs has experienced a decline in demand; (4) that the allures of legitimate employment have been so strong as to reduce the attractions of illegitimate activity; and (5) that private security measures (such as the wider availability of handguns for defensive purposes, security guards, and alarms) are more widely used.  

    Donohue and Levitt suggest a very different explanation for the decline in crime after 1991: the legalization of some abortions in the landmark case of Roe v. Wade, handed down by the U.S. Supreme Court in January, 1973.  What is the connection?  

First, we see a broad consistency with the timing of legalization of abortion and the subsequent drop in crime.  For example, the peak ages for crime are roughly 18-24, and crime starts turning down around 1992, roughly the time at which the first cohort born following Roe v. Wade would hit their criminal prime. ...  Second, [] the five states [New York, Washington, Alaska, Hawaii, and California] that legalized abortion in 1970 saw drops in crime before the other 45 states and the District of Columbia, which did not allow abortions until the Supreme Court decision in 1973.  Third, our [regression] analysis shows that states where abortions were more frequent in 1973-1976 experienced substantially greater crime drops between 1985 and 1997.  This finding is true after controlling for a variety of factors that influence crimes, such as the level of incarceration, income, and racial composition. (3)   

     And they contend that the effect of abortion on crime rates is substantial.  A 10 percent increase in abortions yields about a 1 percent drop in crime.  The overall result is that "legalized abortion can account for about half the observed decline in crime in the United States between 1991 and 1997."  (3)  And it is worth noting that there was a substantial increase in abortions in the years after Roe v. Wade.  In 1973 there were slightly fewer than 800,000 abortions performed in the U.S. Between then and 1982 the number of abortions per year rose to about 1.6 million, where it remained through at least the early 1990s.  Decomposed into an abortion rate (whether per woman of child-bearing years or as a percentage of live births), there was also a significant increase after Roe.  

    The most obvious causal connection between legalized abortion and a subsequent drop in crime is that there are far fewer 18-24 year-old males beginning in 1991 (and earlier in those five states that had legalized abortion prior to 1973).  Their empirical work establishes that this cohort-size effect is present and statistically significant in explaining the decline in crime after 1991.  But somewhat surprisingly there is an additional effect that their empirical work uncovered.  Crime should fall if the young-male cohort size falls, but there may be no obvious reason that the young-male cohort crime rate should fall after Roe.  "While overall crime will fall (in the short run) as the post-legalization cohorts mature, the cohort crime rates will be unaffected if abortions randomly decrease cohort size."  (4)  But, in fact, what Donohue and Levitt found was that the cohort crime rate fell substantially after Roe.  Indeed, they found that of the two effects -- the cohort-size effect and the cohort-rate effect -- the cohort-rate effect explains the greater share of the decline in post-1991 crime.  

    Why should there be a cohort-rate effect because of legalized abortion?  And why should it be so strong?  

Cohort crime rates will fall if the women most at risk to have children who would engage in criminal activity have higher abortion rates.  In fact, women who choose to have abortions are not a random subset of all pregnant women, since teenagers, unmarried women, and African-Americans are all substantially more likely to seek abortions.  Children born to these mothers tend to be at higher risk for committing crime 17 years or so down the road, so abortion may reduce subsequent criminality through this selection effect.  Furthermore, an array of environmental factors can harm (or improve) the life prospects of children, and abortion may reduce the births of those whose chances are most at risk or improve the chances of other children who are born.  ...  If those at high risk for crime are disproportionately aborted, one can observe a substantial reduction in crime even in the absence of a large effect on overall fertility.  

      The evidence is that abortions were not randomly distributed among the population.  Donohue and Levitt cite the work of others -- notably, Levine, Staiger, et al., "Roe v. Wade and American Fertility," NBER Working Paper No. 5615 (June 1996) -- who found that abortion led to a decline in births that was twice as great for teenage and non-white mothers as it was for non-teenage and white mothers.  They also note that legalized abortion may improve infant outcomes by reducing low birthweight babies and neonatal mortality.   

Back to Index of Sections on Criminal Law

The Efficiency of Miranda Warnings

The Fifth Amendment to the United States Constitution gives each citizen a privilege against self-incrimination.  

[Oops!  We haven't quite finished this yet.  We'll have this done by March 24.]

References: Paul G. Cassell U Richard Fowles, "Handcuffing the Cops?" A Thirty-Year Perspective on Miranda' s Harmful Effects on Law Enforcement," 50 Stan. L. Rev. 1055 (1998); John J. Donahue III, "Did Miranda Diminish Police Effectiveness?," 50 Stan. L. Rev. 1147 (1998); and Paul G. Cassell & Richard Fowles, "Falling Clearance Rates After Miranda: Coincidence or Consequence?," 50 Stan. L. Rev. 1181 (1998).  

Back to Index of Sections on Criminal Law

Crime and Work

In Chapter 12 we consider whether there is a relationship between the business cycle and crime.  One simple theory of that relationship would begin from the presumption that rational actors compare legal and illegal alternatives and chose whichever of those alternatives offers the greater expected returns over some decision horizon.  (A more sophisticated theory might allow the decision-maker to choose a combination of illegal and legal work.)   As the relative returns to legal and illegal work vary over the business cycle -- probably rising in favor of legitimate work as the economy prospers and rising in favor of illegitimate work as the economy goes into a recession, rational actors would respond predictably -- by, for example, engaging in more crime as the economy worsens and less crime as the economy prospers.  But the relationship between crime and work may be more subtle than this.  For example, among non-violent, property crimes, there may some crimes that rise and fall with the business cycle precisely because the desired objects of the crime -- say, automobiles for stealing -- rise and fall in value with the cycle. As with many questions of criminal law, there may be no obvious resolution to the relationship without appealing to empirical work.  

    Fortunately, there is some new and interesting literature on this matter.  In a recent article, Jeffrey Fagan (Professor of Public Health and Director of the Center for Violence and Prevention, Columbia University School of Public Health, and Visiting Professor, Columbia University School of Law) and Richard B. Freeman (Professor, Department of Economics, Harvard University, and Program Director for Labor Studies at the National Bureau of Economic Research) have surveyed the literature on the relationship between crime rates and legitimate work.  "Crime and Work," 25 Crime & Justice 225 (1999).  The article bears a close reading because it covers such a broad range of topics.  Here is the abstract of the article:  

"Crime and legal work are not mutually exclusive choices but represent a continuum of legal and illegal income-generating activities.  The links between crime and legal work involve trade-offs among crime returns, punishment costs, legal work opportunity costs, and tastes and preferences regarding both types of work.  Rising crime rates in the 1980s in the face of rising incarceration rates suggest that the threat of punishment is not the dominant cost of crime.  Crime rates are inversely related to expected legal wages, particularly among young males with limited job skills or prospects.   Recent ethnographic research shows that involvement in illegal work often is motivated by low wages and harsh conditions in legal work.  Many criminal offenders 'double up' in both legal work and crime, either concurrently or sequentially.  This overlap suggests a fluid and dynamic interaction between legal and illegal work.  Market wages and job opportunities interact with social and legal pressures to influence decisions to abandon crime for legal work.  Explanations of the patterns of legal and illegal work should be informed by econometric, social structural, and labeling theories.  The continuity of legal and illegal work suggests the importance of illegal wages in research and theory on criminal decision-making."  

Questions:  

1.  Suppose that, as Fagan & Freeman suggest, a typical individual must allocate her time along a spectrum.  At one end of the spectrum she is spending all her time in legitimate work.  At the other end of the spectrum, she is spending all her time in crime.  Intermediate positions represent a combination of legitimate work and crime.  How would you explain where a typical individual will decide to allocate her time along that spectrum?  What factors will cause her to change her position -- opting for more legal work and less crime or vice versa?   

2.  Suppose that you are eager to recommend policies that reduce the amount of crime in society.  If you suspect that people are making decisions about crime in the manner suggested in Fagan & Freeman's abstract and in your answer to Question 1, what policies would you recommend?  Harsher punishment for crime?  More police or different policing methods?  More legitimate work opportunities?  A combination of those policies?  If so, in what proportion?   

Back to Index of Sections on Criminal Law

The Economics of a Criminal Street Gang

Another famous and as-yet-unpublished article on an empirical criminal law is Steven D. Levitt and Sudhir Alladi Venkatesh, "An Economic Analysis of a Drug-Selling Gang's Finances" (National Bureau of Economic Research Working Paper 6592, June, 1998 -- available at www.nber.org/papers/w6592.).  The article is so intriguing that there have reports in the popular press about its findings -- e.g., the Forbes piece shown below.  

    While this article gets the highlights right, you'll want to look at the original Levitt and Venkatesh piece for some additional insights.  Just to give you an idea of what some of those are, Levitt and Venkatesh report that the gang whose finances they studied initially controlled a 12 square block area that was bounded on each side by major thoroughfares, along which most of the gang's selling of crack cocaine took place.  Areas of the city to the North, South, and West were controlled by other street gangs, and these rival gangs frequently tried to encroach onto one another's territory.  A gang war arose when some gang sought to move its selling activities into that of a neighbor.  (Gang wars occupied approximately one-quarter of the four-year period that Levitt and Venkatesh studied.)  One method of doing that was to engage in drive-by shootings along the borders of the selling areas.  Sometimes these shootings were not meant to injure sellers or customers -- although that did happen at what Levitt and Venkatesh call an "astonishing" rate -- so much as simply to disrupt normal selling venues.  The hope was that customers who would normally have dealt with the gang that sold in that particular locale would come to perceive that area as dangerous and shift their purchases into the safe areas controlled by a rival gang.  But the gang members recognized the game theory downside of these battles: shooting up the selling areas of a rival gang's area could lead to a worse state of affairs for all of them if that rival retaliated by shooting up the aggressor's selling areas.  

Greedy bosses

By Scott Woolley

Forbes, Aug. 24, 1998

To understand the business model of a drug gang, think pyramid-selling organizations.  A highly centralized organization controls distribution to leaders of individual gangs, who function much like franchisees.  In return for rights to an exclusive territory, each franchisee pays a flat franchise fee, known as a tribute.  The central gang also controls the wholesale drug distribution network. 

That’s among the insights provided in a new study published by University of Chicago economist Steven Levitt and Harvard sociologist Sudhir Venkatesh.  They teamed to analyze rare data Venkatesh had found: a drug gang’s books, which give an inside peek into the numbers behind the crack business. 

Media folklore has it that drug pushing beats hamburger-flipping for ghetto kids.  Not necessarily so.  Foot soldiers who sell drugs on the street make just over $3 an hour, the academics discovered.  “There's the mythology that the media has built about the lucrativeness of pushing drugs, but the lessons of introductory economics suggest it can't be true,” says Levitt.  “These are kids with very low skills and there's a reserve army of them waiting to get into the gang.  Competition would suggest that the employer has all the power in this relationship.” 

Analyzing the data, the two economists discovered that the rules of economics apply to narcotics.  So why do kids risk a bullet in the head for $3 an hour?  Part of the lure is glamour:  There's a certain prestige that goes with drug pushing.  And there's the hope of promotion.  Levitt and Venkatesh found that some neighborhood gang leaders can make $65 an hour, and the hope of reaching the top keeps the other pushers going. 

What crime pays

A drug gang’s monthly income statement

^*Membership dues are paid by peripheral gang members who often don’t sell drugs.  Note: These figures are based on a 42-month period beginning in July, 1991.  Source: Levitt and Venkatesh. 

The numbers Venkatesh and Levitt use come from a since-disbanded drug gang in a Midwestern city.  Venkatesh lived with some of the gang members during this period and got the records. 

Yes, this is a high-margin business.  The economists discovered that a gang's gross margins for crack are 80 percent (see table).  So why are rewards so low at the end of the distribution chain?  It’s a classic case of exploitation.  The only ones making good money are the higher-ups in the pyramid.  The gangs keep it that way by keeping rival gangs out of their territory, a situation that enables them to exploit both crack users and sellers. These monopolies are secured and defended with guns.  This is about as nasty a kind of economics as anyone could imagine. 

“People researching street gangs typically focus on attitudinal surveys—‘Why did you join this gang?’” says Venkatesh.  “We were interested in treating them as an organized group, an economic unit.”  Levitt believes that educating gang members on the way the business works could push many of them back into legal activities.  “Right now,” Venkatesh says, “the kids are completely getting the calculation wrong about what the likelihood is they’re going to rise up in the gang.”  Drug pushers of the world, arise! 

Questions:

1.  Do you think that the young men who are participating in these gangs are behaving rationally?  Why or why not?  

2.  Suppose that society's desire was to reduce the amount of crime that these gangs engage in.  What sort of public policies would you recommend as correctives?  

Back to Index of Sections on Criminal Law

The Social Costs of Handguns and of Handgun Control

There is a great deal of anxiety and controversy about the role of handguns in U.S. society.  As we note in Chapter 12, the U.S. has far and away the highest rates of homicide and other violent crimes among developed nations -- even though our rates for other crimes are equal to or less than those of our peers.  Many people associate these high rates of violent crimes to the easy availability of guns in the United States.  Others point to socio-economic and historical factors that, they claim, explain the disparity.  

    One of the most fascinating attempts to come to grips with the role of handguns in the U.S. is the work of John R. Lott, Jr.  In a series of articles -- the most famous being John R. Lott, Jr., & David B. Mustard, "Crime, Deterrence, and Right-to-Carry Concealed Handguns," 26 J. Legal Stud. 1 (1997) -- and now in an important new book -- More Guns, Less Crime: Understanding Crime and Gun Control Laws (1998) -- Professor Lott has studied the impact of gun control laws on crime rates and concludes that the relationship is far more complex than most people believe.  The central insight of Professor Lott's work is that handguns confer social benefits as well as social costs and that we do a grave disservice to ourselves if we ignore those benefits and focus exclusively on the costs.  The potentially greatest social benefit of handguns is their use for defensive purposes.  Specifically, if law-abiding citizens may have relatively easy access to handguns, then criminals may be thereby deterred from committing certain crimes because of their quite rational fear of being injured (or worse) by a gun-carrying victim.  As in the case of the Lojack system described in the next entry on the page, a criminal cannot easily distinguish between those potential victims who are carrying a concealed weapon and those who are not.  This may induce criminals to try different kinds of crime or to abandon crime altogether.  In this way, handgun availability may confer an external benefit on society, even to those who do not carry a handgun.  

    Professor Lott goes well beyond this theoretical possibility of the deterrent effect on crime of handguns to attempt to measure whether this effect exists.  He seeks to do so by testing whether, when a state passes a law making it relatively easy for residents to carry concealed weapons, there is, thereafter, a discernible impact on various kinds of crime.  Let us first describe the laws that Professor Lott examines, then describe his data, and then summarize his results.   

    The focus of Professor Lott's study are "shall issue" laws.  These are state statutes that permit any resident of the state who fulfills certain specified criteria to receive a permit to carry a concealed weapon.  These laws are nondiscretionary in that they require the issuing agency to issue the permit to all qualified applicants.  The agency may not use subjective discretion to decide who may and may not receive a permit.  Prior to 1977 only eight states had "shall issue" laws -- Alabama, Connecticut, Indiana, New Hampshire, North Dakota, South Dakota, Vermont, and Washington.  Between 1977 and 1992 ten more states adopted "shall issue" statutes -- Florida (1987), Georgia (1989), Idaho (1990), Maine (1985), Mississippi (1990), Montana (1991), Oregon (1990), Pennsylvania -- excepting Philadelphia (1989), Virginia (1988), and West Virginia (1989).  Since 1993 an additional 13 states have passed nondiscretionary concealed-carry laws.  There are, therefore, a total of 31 states that have "shall issue" laws.  

    Professor Lott has compiled a data set of county-level crime rates for the entire United States over the period 1977 through 1992.  He uses this data set (and others) to test whether there is a discernible effect on various categories of crime of the passage of a "shall issue" law.  That is, Professor Lott takes various crime rates in the counties as his dependent variable and then performs a regression to see if a set of independent variables (such as income, incarceration rates, and population density) can explain variations in the county-level crime rates.  For his purposes, the most important of the independent variables is a "shall issue" dummy variable -- that is, a variable that takes only two values, in this case 1 if the state in which the county lies has passed a "shall issue" law, 0 if it has not.  

    Professor Lott's most notable regression results are that passage of a "shall issue" law (1) reduces violent crime by 5 percent and (2) increases property crime by about 3 percent.  Both of these results are significant at the 1 percent level.  Here is the first row of Table 4.1 from More Guns, Less Crime.  

Table 4.1  John R. Lott, Jr., More Guns, Less Crime (1998).  

    This is only the barest summary of Professor Lott's conclusions.  There are other important empirical tests in the book -- as well as a wealth of fascinating information on handgun ownership, crime rates, and related subjects -- that deserve your careful consideration.  

    Professor Lott did an interview with David Inge of WILL-AM radio on March 1, 2000.  WILL, the broadcast station of the University of Illinois, has archived that hour-long interview on its Web site.  You can find the interview at www.will.uiuc.edu/WILL_Contents/AM_Contents/AM_Focus_580_Webcasts.htm.  In order to listen to the interview you will need to have RealPlayer or its equivalent installed and have headphones or external speakers available.  Because WILL will archive this Webcast for a limited period, we urge you to listen soon.  We'll keep this link here only so long as the interview with Professor Lott is still available.  

A Critique of More Guns, Less Crime

    In the inaugural issue of the American Law and Economics Review, Professors Ian Ayres and John Donohue have reviewed this important work on handguns and crime -- "Nondiscretionary Concealed Weapons Laws: A Case Study of Statistics, Standards of Proof, and Public Policy," 1 Am. L. & Econ. Rev. 436 (1999).  While they greatly admire Professor Lott's attempt to shed light on this troubling issue through careful empirical work and, indeed, admire his empirical work, Ayres and Donahue take issue with some of the conclusions.  

    Ayres and Donohue focus on what they consider to be a puzzling difference between what Lott's theory seems to predict about the deterrent effects of shall-issue laws and the actual effects.  Consider, first, Lott's finding about the effect of shall-issue laws on the murder rate.  Note in the second entry in Table 4.1 above that passage of a shall-issue law caused murder to fall by almost 8 percent in that state.  In other regressions Lott reports that a 1 percentage point increase in the proportion of the population carrying concealed weapons in Pennsylvania and Oregon caused a 27 percent and 37 percent reduction in the murder rates in those states, respectively.  Ayres and Donohue believe those numbers are so far-fetched as to call into question Lott's theory, data, methodology or all three.  The simple reason is that a very small proportion of any state's population has permits for concealed weapons.  As they ask, "Is is plausible that issuing concealed gun permits to 2% of the population (as in Florida) could generate and 8% reduction in murders or a 5% reduction in rapes?"  (p. 458.)  Lott believes that it is plausible and explains why.  But Ayres and Donohue disagree.  They rely on the observation of Franklin Zimring and Gordon Hawkins -- "Concealed Handgun Permits: The Case of the Counterfeit Deterrent," 7 Responsive Community 46 (1997) -- that if concealed handguns deter crime, they are most likely to deter recidivists and the crimes they tend to commit, such as robbery, and least likely to deter crimes that a one-time criminal commits, such as murder.  (The reason is revealed in a simply exercise: "If 2% of the population carries concealed weapons, then a criminal who robs 100 people a year faces an 86.7% chance of encountering a concealed weapon [at least once] over the course of a year [; a] 2% chance of encountering an armed victim may not be sufficient to deter a one-time criminal, but it may be sufficient to deter someone from making a  profession out of robbery."  (pp. 458-59)  Of course, one wonders whether the typical criminal can make this relatively sophisticated probabilistic calculation.)   

    But Ayres and Donohue report that Lott's results do not confirm this prediction that the strongest deterrent effect of shall-issue laws will be on crimes that recidivists tend to commit, such as robbery, and the weakest will be on non-repeat crimes, such as murder.  To see this, they re-do his regressions using an alternative specification.  Ayres and Donohue say that one of the puzzling omissions in Professor Lott's study is his failure to use the incarceration rate as an explanatory variable.  Many crime studies include and lay great emphasis on the rate at which those convicted of crime are incarcerated as a possible explanatory variable of the crime rate.  The standard hypothesis is that a greater willingness to incarcerate will cause a later reduction in crime.  Ayres and Donohue re-ran Professor Lott's regression (reported in the table above) with the incarceration rate rather than the arrest rate (which Lott had used) as an independent variable.  (Clearly, the arrest rate and the incarceration rate are so highly correlated that it would not reveal much information to use both of those rates as independent variables; Ayres and Donohue prefer to use the incarceration rate for plausible technical reasons.)  They report the results of this alternative specification in the following table (from p. 450 of their article):  

Table 1.  From Ayres & Donohue.  Replicating Lott's regressions in Table 4.1 above but dropping the arrest rate and instead including the incarceration rate as an explanatory variable.  The starred entries are our addition.  

    A comparison of this table with the one above indicates that there are changes in many of the coefficients.  However, in most cases the changes confirm the effects reported in Lott's study.  But two changes stand out.  (We have put asterisks beside those two results.)  First, for murder the very high reduction in murder rates reported by Lott in Table 4.1 above falls dramatically and becomes statistically insignificant.  Second, Lott's reported effect on robbery almost disappears.  This result, combined with the theoretical implication mentioned above about the relationship between shall-issue laws and robbery, causes Ayres and Donohue to be skeptical about Lott's reported results.        

    One last set of criticisms is worth noting.  If shall-issue laws have strong deterrent effects, they should have particularly strong effects on particular kinds of crimes and criminals -- as we saw above with respect to robbery and murder.  But in addition, shall-issue laws are likely to have these three effects:  "(1)  the ratio of stranger to acquaintance killings [should] decrease; (2) the amount of the confrontational and economically motivated crime of robbery [should] to go down; and (3) the ratio of  young to old victims [should] increase (because youths cannot legally carry concealed weapons)."  (p. 460)   Ayres and Donohue report that none of these effects seems to have occurred.  We have already seen that there is little, if any, effect on robbery.  Moreover, Lott and Mustard (in their 1997 article) reported  that the proportion of stranger killings increased after the passage of shall-issue laws.  And finally, "Lott fins no statistically significant difference in the way that concealed weapon laws deter crime against adults and youth victims."  (p. 461)  

Back to Index of Sections on Criminal Law

The Economics of Lojack

Recall that the externalities can be negative or positive.  Pollution is a negative externality (or external cost), and inoculation against a contagious disease is a positive externality (or external benefit).  

    When you install a burglar alarm on your automobile, you do so presumably because the cost of the alarm is less than the expected benefits (the value of your car times the reduction in the probability of loss through theft).  Cars armed with burglar alarms usually have some external indication of that fact -- such as a blinking small red light on the dashboard, observable through the windshield -- so as to alert and discourage potential thieves.  When buying that burglar alarm, what you probably do not take into account is the fact that because your car is now less likely to be stolen, your neighbor's car may now be more vulnerable.  The implication is that some traditional security devices may not reduce the overall level of crime but simply shift its incidence.  

    Other devices and actions designed to deter theft may confer a positive externality on others.  For example, a motion-sensitive light on the front of one house may illuminate a neighbor's house, too.  A video camera that records comings and goings in the front and back of one business establishment may do the same for a neighboring business.  

     One recent anti-theft device that conveys substantial positive externalities is Lojack.  Lojack, introduced in the late 1980s, is a radio transmitter hidden in a car.  Only representatives of the company can install the radio, so that not even the car owner knows where the device is.  The radio is placed randomly within the car, so that a thief may not easily find it and disable it.  There is no external indication on the car that it is equipped with Lojack.  If a thief steals a car equipped with a Lojack device, the owner calls the company's local office.  The company then sets its receiver to pick up the transmission from the radio installed in that car.  When they locate the car, they inform the police, who retrieve the car and, if possible, arrest the thief.  

    The device is so effective that the Lojack Company warrants to recover a customer's car within 24 hours of his reporting its loss. Approximately 90 percent of cars equipped with Lojack are recovered, and those so equipped typically have about 20 percent of the damage done to other stolen cars.  The cost of the service is relatively modest (about $500), and there are no monthly fees.  Initially, Lojack was available only in major urban areas in the United States, but it is now available in nearly every state and in many foreign countries.  (See www.lojack.com.)  

     Because there is no indication that a car may contain Lojack, auto thieves have had to innovate against the possibility that any given car they steal might contain the device.  For instance, a thief may drive the stolen car to a new location, park it, and surreptitiously watch it for a day or so.  If no one comes to claim the car, then the chances are good that it is not equipped with Lojack.  However, the police now know that thieves are parking and watching stolen cars to learn if they have Lojack.  And so, the police, when they receive notification that a Lojack-equipped car has been found, go to watch the car to see if, after 24 hours or so, someone approaches it.  

    From an economic point of view, there are several remarkable points to be made about Lojack.  First, it significantly reduces automobile theft rates.  As Ian Ayres and Steven Levitt found -- "Measuring Positive Externalities from Unobservable Victim Precaution: An Empirical Analysis of Lojack," 113 Quarterly J. of Economics 43 (1998) -- a 1 percent increase in Lojack usage in a given area can reduce auto-theft rates there by 20 percent.  Second, note that Lojack confers a significant positive externality on other automobile owners.  This is because a thief cannot tell whether any given car is protected by the device.  This deters auto theft generally, not merely the theft of any particular car.  An interesting issue is whether the large reduction in auto theft in Lojack areas results in increases in other crimes, which may have become more attractive because of the reduction in the attraction of auto theft.  

Questions:  

1.  The usual economic method of dealing with an external benefit is to subsidize the externality-generating activity.  For example, the government frequently subsidizes inoculation programs against communicable diseases, such as the flu.  Should the government subsidize the cost of Lojack?   (See Steven Landsburg, "Property Is Theft: When Protecting Your Own Property Is Stealing From Others," Slate, www.slate.com/Economics/97-08-02/Economics.asp.)    

2.  The other side of this coin is that economics calls for the taxation (or other regulation) of those activities that generate external costs.  As Professor Landsburg asks in his column cited in the previous question, does it follow that the government should tax the owner who installs a traditional burglar alarm on his home or car?  

Back to Index of Sections on Criminal Law

Deterrent Effects of Errors

Suppose that the criminal justice system is, as we know it to be, imperfect.  From time to time the system finds truly innocent people guilty of a crime, and, just as surely, the system also exonerates some who are truly guilty people.  What effects are these errors likely to have on the deterrence properties of the criminal justice system?  

    In a recent article -- "Shocking the Conscience: Pragmatism, Moral Reasoning, and the Judiciary, A Review of The Problematics of Moral and Legal Theory by Richard A. Posner (Harvard University Press, 1999)," 16 Constitutional Commentary 675 (1999) -- Professor Dan Farber argues that convicting the innocent is likely to have little deterrence effect:  

Suppose that of 100,000 people 95 percent are innocent and 5 percent are guilty.  Suppose also that in a given year 10,000 people are convicted, of whom only a quarter are truly guilty.  Under the hypothetical, the odds of being punished if your are innocent are only about 8 percent (7,500 out of 95,000), while the odds of being punished if you are guilty are 50 percent (2,500 out of 5,000).  So even in in this extreme case in which most of those punished are innocent, the effect on deterrence is not substantial, because the guilty still face much higher risks of conviction. (n. 24)  

Questions:

1.  Comment on the example (beyond saying, "Makes sense to me").  To what extent does this example work simply because the proportion of the population that is criminally inclined is small?  Would matters change if the property so inclined were larger?  Why?  

2.  Professor Farber recognizes the point of the first question in n. 25 of his article.  He says, "[The effect on deterrence] is significant when guilt is widespread because, unless an even higher percentage of the convicted are actually guilty, the innocent have no incentive to refrain from crime."  Rework the example above when 75 percent of every 100,000 people are criminals and 25 percent of the 10,000 convictions are erroneous convictions of the innocent.  What are the odds of being punished if you are guilty?  What are the odds of being punished if you are innocent?  Does the criminal justice system have a deterrent effect?     

Back to Index of Sections on Criminal Law

The Six Percent Rule

The economic theory of crime and punishment might be read to suggest that anyone might commit crime.  All that is required is that the expected benefit of a crime sufficiently exceed the expected punishment.  This need not be read to suggest that criminal acts are randomly scattered among the population, but it might suggest that a relatively wide, rather than a narrow, slice of any society's population would commit crime.  

    The remarkable fact is, however, that a very narrow slice of any population commits the majority of that population's crime.  Criminologists believe that six percent of any birth cohort in any society commits 50 percent of the crime committed by members of that cohort.  There are also suggestions in the literature that the criminal justice authorities know who these six percent are when they are very young.  They are in trouble early and often.  

    See Marvin Wolfgang, Robert F. Figlio, and Thorstein Sellin, Delinquency in a Birth Cohort (Chicago: University of Chicago Press, 1972) and Paul Tracy, Marvin Wolfgang, and Robert Figlio, Delinquency Careers in Two Birth Cohorts (Washington, DC: U.S. Department of Justice, Office of Juvenile Justice and Delinquency Prevention, National Institute for Juvenile Justice and Delinquency Prevention, 1985).  

Question:

1.  What are the implications for criminal justice policy of the finding that six percent of a birth cohort commits 50 percent of that cohort's crime?  How does this fit with the Donohue & Levitt finding on the connection between legalized abortion and crime?  

2.  Assuming that the six percent of malefactors could be identified at an early age, would you suggest special monitoring of that group?  Why or why not?  

Back to Index of Sections on Criminal Law

Crime and Social Interaction

The sociological literature lays great stress on the fact that the decision to commit crime may be importantly determined by the social interactions within certain subgroups within society.  The fact that young males commit a very large percentage of crime -- e.g., 50 percent of property crimes -- and that young males occupy a particular subculture within society that highly values conforming behavior is one casual bit of evidence in favor of this sociological view.  

    Recently economists have become extremely interested in social interaction as the source of behavior.  In an important recent article -- "Crime and Social Interactions," Quarterly Journal of Economics (May, 1996): 507-45 -- Edward Glaeser, Bruce Sacerdote, and Jose Scheinkman explore the importance of social factors in determining crime rates.  They begin with the observation that, although most investigators focus on broad trends in crime rates, they are struck by the "astoundingly high variance [in those rates] across time and space."  For example, within the United States the homicide rate fell by half between 1933 and 1961, and homicide rates vary greatly across countries.  ("[Those rates] range from 6.1 homicides per million Japan, to 12.6 homicides per million in Sweden, to 98.0 homicides per million in the United States in 1990.")  In seeking to explain this variance, the economic theory of criminal behavior might look first to differences in economic and social conditions.  However, crime rates vary dramatically over short distances within which the underlying economic conditions would seem to be relatively close.  For instance, "East Point, Georgia, has a crime rate of 0.092 crimes per capita.  El Dorado, Arkansas, which has more unemployment, less education, more poverty, and lower per capita income, has a crime rate of 0.039 crimes per capita.  The 51st precinct of New York City has 0.046 crimes per capita while the wealthier 49th precinct has 0.116 crimes per capita."   (508)  Glaeser, Sacerdote, and Scheinkman "find that less than 30 percent of the variation in cross-city or cross-precinct crime rates can be explained by differences in local area attributes."  (508)  

    The authors develop a model of social interaction in which there are two kinds of agents: "(1) agents who influence and are influenced by their neighbors; and (2) agents who influence their neighbors, but who cannot themselves be influenced ('fixed agents')."  (509)   Fixed agents are those who are not connected to their neighbors.  "[T]he number of fixed agents can be seen as a metaphor for the forces that slow social interaction, such as strong parents, formal schooling, or any information that counters peer influence."  (509)  The model suggests that the greater the percentage of fixed agents in an area, the less binding are social interactions, and, therefore, the higher will be certain crime rates.  

    Using a data set on a variety of different crimes in the United States in 1970 and 1985 and in New York City in 1985, Glaeser, Sacerdote, and Scheinkman find a statistically significant connection between the degree of social interaction and the level of certain crimes.  Generally speaking, they find that their is a very low level of social interaction in areas where there are high violent crimes (such as arson, murder, and rape) and higher levels of social interaction in areas where property crime are high.  They interpret these results to mean that property crimes are likely to be high in areas where there is a high behavioral covariance among potential criminals and that violent crimes are high where there is a low behavioral covariance.  

Back to Index of Sections on Criminal Law

Incapacitation or Deterrence?

One of the empirical puzzles of criminal studies is whether incarceration reduces the crime rate through a general deterrence effect or through incapacitation of those who are confined.  (We discuss the difference between general deterrence and specific deterrence on pp. 460-63 in the third edition of Law and Economics.)   The matter is still an open one, but there are important new studies on punishment and crime rates appearing frequently.  One important recent piece is Steven D. Levitt's "The Effect of Prison Population Size on Crime Rates: Evidence from Prison Overcrowding Litigation," Quarterly Journal of Economics (May, 1996): 319-353.  Levitt begins by noting that the prison population in the United States tripled between the late 1970s and the mid-1990s.  By 1994 there were over one million people in prisons in this country, and all levels of government were spending over $40 billion per year on incarceration.  

    Levitt directs his attention toward measuring the incapacitating effects of imprisonment.  He concludes, "Incarcerating one additional prisoner reduces the number of crimes committed by approximately fifteen per year, a number in close accordance with the level of criminal activity reported by the median prisoner in surveys."  (348)  

Question: 

    Re-read the material in the box on p. 471 describing Donohue and Siegleman's calculation of the socially optimal amount of imprisonment.  Would you alter those authors' conclusions on the basis of Levitt's finding about the specific deterrence effect of imprisonment?   

Back to Index of Sections on Criminal Law

"Broken Windows"

Recall the notion of a public good: something that has the properties of nonrivalrous consumption and high costs of excluding nonpaying beneficiaries ("free riders").  National defense is an example.  

    Public goods present a collective action problem -- societies must figure out a way to provide for the provision of those public goods when it is in everyone's interest not to pay for them.  Consider public order in an urban neighborhood as a public good.  (Can you explain in what sense this is a public good?)  So long as people in a neighborhood contribute to the maintenance of public order, there will be public order.  This may consist of paying attention to who is out and about, scaring away bad actors from the neighborhood, picking up litter from public places, and the like.  And very importantly it may also consist of maintaining one's own property.  Unfortunately, it may be the case that no one has a strong enough incentive to incur these costs when the benefits accrue to other people.  When the residents of a neighborhood quit caring about investing in order within their own neighborhood, they public order suffers.  Street crime increases, and other social problems begin to rise.  

    These thoughts lie at the heart of a very important hypothesis in criminal justice studies -- that of James Q. Wilson and George L. Kelling in an article entitled "Broken Windows" in The Atlantic Monthly (March, 1982).  Wilson  and Kelling contend that when broken windows are not fixed, this (and other failures to maintain property) are a sure signal that public order is decreasing or gone.  Some people will take the fact that private and public property are not being maintained as evidence that other social norms are not being enforced either.  High crime rates and other anti-social behaviors are sure to follow.  

Questions:

    Assume that the "broken windows" hypothesis is an important predictor of social disorder in urban neighborhoods.  What policies could help to prevent the increase in public disorder that follows from unrepaired broken windows?   

Back to Index of Sections on Criminal Law

Unintended Costs of "Three Strikes" Laws

The perception in the early and mid-1990s was that crime was increasingly significantly and that further steps needed to be taken to deter criminal activity.  (We now know -- see, for example, the entry above on "Legalized Abortion and Crime" -- that this perception was mistaken.  Crime had started to decline about 1991.)   Approximately half of the 50 states responded by passing "three strikes" laws, which provided for significantly more severe sanctions than would otherwise be due upon a person's conviction of a third felony.  For example, in 1994 California provided for a sentence of 25 years to life in prison for conviction of a third felony, regardless of the felony.  Let us leave to one side the legal and economic analysis of the deterrent impact, if any, of these "three strikes" laws and focus on an unintended consequence of these laws -- court congestion.  

    By late 1995 Los Angeles County simply could not cope with the flood of business that the "three strikes" law created.  California law requires that criminal trials commence within 60 days of filing.  Because of the greater jeopardy attending every felony conviction, more defendants insisted on trials, rather than plea-bargaining to a lesser charge and no trial.  The result was that in 1995 there was a 35 percent increase in criminal trials in Los Angeles County.  In response, the County had to double its jury summonses from 2 million to 4 million per year.  San Diego County saw an increase from 2,700 criminal cases per month before "three strikes" to about 4,000; Santa Clara County had a 50 percent increase in felony trials.  

    Not surprisingly, there is, as a result of these increases in criminal trials, a shortage of judges and courtrooms.  Los Angeles County had to bring 20 judges out of retirement to help deal with the increase in criminal trials.  

    Finally, California, like most states, allows a larger number of peremptory challenges to jurors (dismissals of potential jurors without the attorney's having to explain the dismissal) in cases involving life imprisonment or the death penalty.  (These challenges are in addition to the 10 challenges for cause that the State allows in these cases.)  However, California's peremptory challenges -- 20 -- in capital cases and those in which life imprisonment is possible are twice the number allowed in other states.  

    These factors constitute significant costs that must be weighed against whatever benefits the "three strikes" laws confer.  Do you have a sense for whether these costs exceed or are less than the benefits?  Can you think of ways in which the states can manage the increase in the number of criminal trials that follows from the "three strikes" laws?  

See "Where's the jury?," The Economist, p. 26 (December 16, 1995).  

Back to Index of Sections on Criminal Law

Illinois' Moratorium on the Death Penalty and an Update on New Procedures in Capital Cases in Illinois

On January 31, 2000, Governor George Ryan of Illinois declared a moratorium on all executions of the death penalty in the State of Illinois.  The reason for this extraordinary step was the revelation that 13 men who had been condemned to death since 1977 had been recently exonerated of the murders for which they had been convicted and set free.  Since 1977 Illinois has executed 12 men.  While the Illinois experience with exoneration of condemned men is extraordinary, it is not singular.  85 people in more than 20 states have been released from death sentences.  None of these people has been released on a technicality.  Rather, they have been released because new evidence has come to light (such as DNA evidence or recantation of a witness' testimony) that either proves them innocent or makes it doubtful that they would have been convicted under a standard allowing for reasonable doubt.  (You can find much information on these cases and on the death penalty generally at the Web site of the Death Penalty Information Center, a group that opposes the death penalty: www.essential.org/dpic/.)   

    Since 1976 when the U.S. Supreme Court reinstated the constitutionality of the death penalty, 610 people have been put to death by the various states.  The pace at which states are executing prisoners is increasing, with 98 having been killed in 1999, the highest total in the last 48 years.  

    Governor Ryan has appointed a distinguished group to review Illinois' procedures for imposing the death penalty and has asked them to report their findings and recommendations to him as soon as possible.  

    On January 23, 2001, the Chief Justice of the Supreme Court of Illinois, Moses W. Harrison II, announced that the Court has adopted new rules in capital cases.  The rules are to take effect in March, 2001.  (The Supreme Court of Illinois took these actions independently of the recommendations, if any, that Governor Ryan's panel might later issue or that the Illinois General Assembly might enact.)  The Court's new rules were suggested after two years of study by a committee of 17 judges.  

    One of the central flaws said to exist in the current Illinois system is that criminal defendants in capital cases sometimes receive less competent legal representation than the seriousness of their cases warrants.  The Chicago Tribune's study that uncovered 13 innocent men who had been condemned to death also found, for example, that 33 men in Illinois who had been sentenced to death had been represented by lawyers who had at one time been disbarred or suspended.  To correct this problem, the Illinois Supreme Court will require that two lawyers be appointed for every indigent defendant in a capital case.  Moreover, any lawyer in a capital case, whether appointed or retained privately, "must be certified by the [Illinois] Supreme Court as a member of the Capital Litigation Trial Bar.  Lead lawyers must have five years of experience in court and eight felony jury trials, including two murder trials, before taking on a death penalty case.  Their co-counsels must have three years of experience and five felony jury trials." 

    The new rules also create special duties for prosecutors and judges in capital cases.  For instance, the rules require prosecutors to hand over to defense lawyers any evidence that might exonerate their client, such as a statement that someone else committed the crime; to label clearly any evidence that might be mitigating, so that it does not get lost in the welter of information and testimony; and to hand over any material on DNA evidence, such as information on discrepancies in testing or a record of laboratory errors.  The new rules also require that judges in capital cases have had training every two years on capital cases and that they remind prosecutors in those cases that their duty is "to seek justice, not merely to convict."   

Back to Index of Sections on Criminal Law

Latest Incarceration Figures

    In the textbook we reported in Chapter 12 that the population of all jails and prisons in the United States is above 1 million.  In fact, in mid-2000 that population reached 2 million.  The annual cost of incarcerating this number of people is approximately $40 billion.  

    The current rate of imprisonment in the U.S. is 690 prisoners per 100,000 people.  That rate is the highest in the world, surpassing Russia.  As reported by Stuart Taylor in The National Journal of October 21, 2000, that rate is five times the imprisonment rate in the United Kingdom; six times that in Canada, Australia, and Spain; seven times that in France and Italy; and 17 times that in Japan.  

    Roughly half those 2 million people in prison are "serving time for small-time drug deals and other nonviolent crimes, such as stealing cars."  (Taylor, National Journal)  

Back to Index of Sections on Criminal Law

Virtual Pornography?

The United States Supreme Court has agreed to hear a case in the Fall, 2001, regarding whether digital depictions of children (or child-like animations) displayed on computer screens violate the Child Pornography Prevention Act of 1996.  That law takes accounts of computer sophistication by criminalizing the creation or possession of digitized images of children in sexual settings.  Congress felt that virtual child pornography hurt real children by increasing the demand for child pornography generally -- perhaps leading to the greater sexual exploitation of real children.   

    The constitutional issue presented is whether the Act's proscription on certain created digital images violates the First Amendment.   Three federal appellate courts have decided that this aspect of the Act does not violate the First Amendment on the ground that Congress could balance the cost of digitized virtual child pornography (the possible adverse consequences for real children) against the benefits of fostering creative expression and decide that the costs outweighed the benefits, without violating the Amendment.  However, a fourth federal appellate court, in California, struck down the law.  The court held that the Act criminalized "foul figments of creative technology that do not involve any human victims."  The central issue presented to the Supreme Court is whether the "justifications offered for the law are real or a pretext for suppressing troubling speech."  Generally speaking, the Court has always rejected the notion that repugnant ideas may be banned because of the consequences that will occur if they are adopted.  

    Does economics help to resolve this issue?  Should the Court allow Congress to make tradeoffs between the costs and benefits of suppressing certain kinds of speech?  If so, under what circumstances -- an imminent danger, a danger to vulnerable groups (such as children), public health and welfare, something else?  Or would it be better for the Court to forbid Congress to make these tradeoffs on the ground, perhaps, that the benefits of free speech always exceed the costs or the ground that the government ought not to judge which ideas are repugnant and which not?  

Back to Index of Sections on Criminal Law

© Copyright 2000, Robert D. Cooter & Thomas S. Ulen.  All rights reserved.