Crime in Cities Brendan O’Flaherty & Rajiv Sethi Frontiers of Urban Economics, Columbia University O’Flaherty & Sethi Crime in Cities Columbia University 1 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 2 / 55 Crime and Space What are crimes? O’Flaherty & Sethi Crime in Cities Columbia University 3 / 55 Crime and Space What are crimes? Crimes are activities that governments have threatened to punish Significant variation in coverage across time and space Examples: blasphemy, sodomy, prostitution, alcohol consumption O’Flaherty & Sethi Crime in Cities Columbia University 3 / 55 Crime and Space What are crimes? Crimes are activities that governments have threatened to punish Significant variation in coverage across time and space Examples: blasphemy, sodomy, prostitution, alcohol consumption Crimes with a spatial component O’Flaherty & Sethi Crime in Cities Columbia University 3 / 55 Crime and Space What are crimes? Crimes are activities that governments have threatened to punish Significant variation in coverage across time and space Examples: blasphemy, sodomy, prostitution, alcohol consumption Crimes with a spatial component Focus on index crimes; mala in se Murder, rape, robbery, assault, larceny, burglary, motor vehicle theft Other street crime (such as drug selling) also has spatial component We omit fraud, embezzlement, cybercrime, terrorism O’Flaherty & Sethi Crime in Cities Columbia University 3 / 55 Crime and Space What are crimes? Crimes are activities that governments have threatened to punish Significant variation in coverage across time and space Examples: blasphemy, sodomy, prostitution, alcohol consumption Crimes with a spatial component Focus on index crimes; mala in se Murder, rape, robbery, assault, larceny, burglary, motor vehicle theft Other street crime (such as drug selling) also has spatial component We omit fraud, embezzlement, cybercrime, terrorism Most personal crimes also have a strategic component O’Flaherty & Sethi Crime in Cities Columbia University 3 / 55 Some Data O’Flaherty & Sethi Crime in Cities Columbia University 4 / 55 Crime Rate (per 100,000 inhabitants), 1986=100 140 120 100 80 60 40 20 2012 2010 2008 2006 2004 2002 2000 1998 1996 1994 1992 1990 1988 1986 1984 1982 1980 1978 1976 1974 1972 1970 1968 1966 1964 1962 1960 0 Years Murder O’Flaherty & Sethi Robbery Crime in Cities Burglary MV Theft Columbia University 5 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 6 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 7 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 8 / 55 Crime and Location O’Flaherty & Sethi Crime in Cities Columbia University 9 / 55 Crime and Location Inter-Metropolitan Variation Robbery, motor vehicle theft, murder; more prevalent in larger cities Possible effects of density: more disputes, easier victim selection Rape, burglary, larceny appear uncorrelated with city size O’Flaherty & Sethi Crime in Cities Columbia University 9 / 55 Crime and Location Inter-Metropolitan Variation Robbery, motor vehicle theft, murder; more prevalent in larger cities Possible effects of density: more disputes, easier victim selection Rape, burglary, larceny appear uncorrelated with city size Intra-Metropolitan Variation Crime is highly concentrated within cities (much more so than across) Murder, robbery, motor vehicle theft more concentrated than poverty Burglary and theft are much less concentrated O’Flaherty & Sethi Crime in Cities Columbia University 9 / 55 Crime and Location Inter-Metropolitan Variation Robbery, motor vehicle theft, murder; more prevalent in larger cities Possible effects of density: more disputes, easier victim selection Rape, burglary, larceny appear uncorrelated with city size Intra-Metropolitan Variation Crime is highly concentrated within cities (much more so than across) Murder, robbery, motor vehicle theft more concentrated than poverty Burglary and theft are much less concentrated Within city hot spots are sites of disproportionate incidence Sherman’s (1989) Minneapolis study: 3% of locations, 50% of calls Crime concentrated at intersections/addresses, not neighborhoods Stable concentration despite adjustment by victims, police, businesses O’Flaherty & Sethi Crime in Cities Columbia University 9 / 55 Theoretical Approaches O’Flaherty & Sethi Crime in Cities Columbia University 10 / 55 Characteristics versus Incentives Cesare Lombroso (1878): criminals are born different Cesare Beccaria’s (1764): rewards and punishments determine crime O’Flaherty & Sethi Crime in Cities Columbia University 11 / 55 Characteristics versus Incentives Cesare Lombroso (1878): criminals are born different Cesare Beccaria’s (1764): rewards and punishments determine crime Lombroso argued that characteristics were physically identifiable Modern versions include environment and time-varying characteristics O’Flaherty & Sethi Crime in Cities Columbia University 11 / 55 Characteristics versus Incentives Cesare Lombroso (1878): criminals are born different Cesare Beccaria’s (1764): rewards and punishments determine crime Lombroso argued that characteristics were physically identifiable Modern versions include environment and time-varying characteristics Two claims of criminogenic characteristics: strong and weak Weak: relatively stable traits predict who commits crime Strong: prevalence of characteristics affects volume of crime O’Flaherty & Sethi Crime in Cities Columbia University 11 / 55 Which Characteristics Matter? O’Flaherty & Sethi Crime in Cities Columbia University 12 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) Family Structure (single parent households) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) Family Structure (single parent households) Legality of Abortion (reduced births 4-5%, increased pregnancies) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) Family Structure (single parent households) Legality of Abortion (reduced births 4-5%, increased pregnancies) Education (preschool and school quality, high school completion) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) Family Structure (single parent households) Legality of Abortion (reduced births 4-5%, increased pregnancies) Education (preschool and school quality, high school completion) Personality Traits (hyper-vigilance to threat cues, hostile attribution) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) Family Structure (single parent households) Legality of Abortion (reduced births 4-5%, increased pregnancies) Education (preschool and school quality, high school completion) Personality Traits (hyper-vigilance to threat cues, hostile attribution) Physiological Traits (brain structure and function) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) Family Structure (single parent households) Legality of Abortion (reduced births 4-5%, increased pregnancies) Education (preschool and school quality, high school completion) Personality Traits (hyper-vigilance to threat cues, hostile attribution) Physiological Traits (brain structure and function) In utero experience (smoking, alcohol, hunger) O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Characteristics Correlated with Offending Age and Gender (inmates and arrestees mostly male and young) Childhood Lead Exposure (via impulsiveness, aggression, IQ, ADHD) Psychological disorders (schizophrenia, major depressive disorder) Family Structure (single parent households) Legality of Abortion (reduced births 4-5%, increased pregnancies) Education (preschool and school quality, high school completion) Personality Traits (hyper-vigilance to threat cues, hostile attribution) Physiological Traits (brain structure and function) In utero experience (smoking, alcohol, hunger) Race and Ethnicity effects not accounted for by characteristic distributions O’Flaherty & Sethi Crime in Cities Columbia University 13 / 55 Incentives O’Flaherty & Sethi Crime in Cities Columbia University 14 / 55 The Incentive Approach Cesare Beccaria’s (1738-1794): On Crimes and Punishments Crime depends on rewards and punishments Expected punishment should be set to equal expected reward O’Flaherty & Sethi Crime in Cities Columbia University 15 / 55 The Incentive Approach Cesare Beccaria’s (1738-1794): On Crimes and Punishments Crime depends on rewards and punishments Expected punishment should be set to equal expected reward Approach revived by Becker (1968) Effectiveness of punishment depends on certainty and severity O’Flaherty & Sethi Crime in Cities Columbia University 15 / 55 The Incentive Approach Cesare Beccaria’s (1738-1794): On Crimes and Punishments Crime depends on rewards and punishments Expected punishment should be set to equal expected reward Approach revived by Becker (1968) Effectiveness of punishment depends on certainty and severity Beccaria: certainty is more effective (hope dispels apprehension) Beccaria also suggested celerity or swiftness of punishment matters O’Flaherty & Sethi Crime in Cities Columbia University 15 / 55 The Incentive Approach Cesare Beccaria’s (1738-1794): On Crimes and Punishments Crime depends on rewards and punishments Expected punishment should be set to equal expected reward Approach revived by Becker (1968) Effectiveness of punishment depends on certainty and severity Beccaria: certainty is more effective (hope dispels apprehension) Beccaria also suggested celerity or swiftness of punishment matters Punishments can reduce crime through deterrence or incapacitation Incarceration operates through both effects O’Flaherty & Sethi Crime in Cities Columbia University 15 / 55 Deterrence Effects of Certainty and Severity O’Flaherty & Sethi Crime in Cities Columbia University 16 / 55 Deterrence Effects of Certainty and Severity The Argentina mobilization (Di Tella and Schargrodsky, 2004) Argentina 1994: Bombing of Jewish Community Center Police presence increased at all Jewish and Muslim institutions Car thefts fell by 75% on affected blocks No evidence of displacement to other locations O’Flaherty & Sethi Crime in Cities Columbia University 16 / 55 Deterrence Effects of Certainty and Severity The Argentina mobilization (Di Tella and Schargrodsky, 2004) Argentina 1994: Bombing of Jewish Community Center Police presence increased at all Jewish and Muslim institutions Car thefts fell by 75% on affected blocks No evidence of displacement to other locations The Italian Collective Pardon (Drago et al., 2009) A 2006 prison release in Italy induced heterogeneity in severity Overcrowding resulted in early release Reoffenders faced restoration of remaining sentences Those facing more severe punishments found less likely to re-offend O’Flaherty & Sethi Crime in Cities Columbia University 16 / 55 Race and Ethnicity Large racial disparities in arrests, incarceration, and victimizations Not due to lower likelihood of punishment conditional on offending Opportunity costs may be part of the explanation But other factors are in play, especially for homicide and robbery O’Flaherty & Sethi Crime in Cities Columbia University 17 / 55 Race and Ethnicity Large racial disparities in arrests, incarceration, and victimizations Not due to lower likelihood of punishment conditional on offending Opportunity costs may be part of the explanation But other factors are in play, especially for homicide and robbery Need to consider strategic incentives arising from categorization O’Flaherty & Sethi Crime in Cities Columbia University 17 / 55 Race and Ethnicity Large racial disparities in arrests, incarceration, and victimizations Not due to lower likelihood of punishment conditional on offending Opportunity costs may be part of the explanation But other factors are in play, especially for homicide and robbery Need to consider strategic incentives arising from categorization Crimes are strategically and informationally distinct O’Flaherty & Sethi Crime in Cities Columbia University 17 / 55 Robbery O’Flaherty & Sethi Crime in Cities Columbia University 18 / 55 Strategic Considerations Robbery involves victim selection, resistance, forced compliance O’Flaherty & Sethi Crime in Cities Columbia University 19 / 55 Strategic Considerations Robbery involves victim selection, resistance, forced compliance Unobserved heterogeneity in both victim and offender populations O’Flaherty & Sethi Crime in Cities Columbia University 19 / 55 Strategic Considerations Robbery involves victim selection, resistance, forced compliance Unobserved heterogeneity in both victim and offender populations Visible markers used to infer likelihood of resistance and violence Those least likely to resist targeted at highest rates Likelihood of resistance, violence will be identity contingent O’Flaherty & Sethi Crime in Cities Columbia University 19 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 20 / 55 Deterrence and Selection More effective deterrence (or better labor market) lowers crime O’Flaherty & Sethi Crime in Cities Columbia University 21 / 55 Deterrence and Selection More effective deterrence (or better labor market) lowers crime But who exit differ systematically from those who remain O’Flaherty & Sethi Crime in Cities Columbia University 21 / 55 Deterrence and Selection More effective deterrence (or better labor market) lowers crime But who exit differ systematically from those who remain For robbery, those who remain more likely to force compliance O’Flaherty & Sethi Crime in Cities Columbia University 21 / 55 Deterrence and Selection More effective deterrence (or better labor market) lowers crime But who exit differ systematically from those who remain For robbery, those who remain more likely to force compliance Prediction: robberies are more violent when less frequent Evidence from NCVS supports this O’Flaherty & Sethi Crime in Cities Columbia University 21 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 22 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 23 / 55 Racial Stereotypes and Robbery Victim resistance is itself a form of deterrence O’Flaherty & Sethi Crime in Cities Columbia University 24 / 55 Racial Stereotypes and Robbery Victim resistance is itself a form of deterrence (Wright and Decker, 1997): I rob mostly whites... I usually don’t have no problem [with resistance], none at all. [Whites] got this stereotype, this myth, that a black person with a gun or knife is like Idi Amin or Hussein. And [a] person [who believes] that will do anything [you say] O’Flaherty & Sethi Crime in Cities Columbia University 24 / 55 Racial Stereotypes and Robbery Victim resistance is itself a form of deterrence (Wright and Decker, 1997): I rob mostly whites... I usually don’t have no problem [with resistance], none at all. [Whites] got this stereotype, this myth, that a black person with a gun or knife is like Idi Amin or Hussein. And [a] person [who believes] that will do anything [you say] Whites accept the fact that they’ve been robbed... some blacks would rather die than give you they bucks and you damn near have to be killing [them] to get it O’Flaherty & Sethi Crime in Cities Columbia University 24 / 55 Racial Stereotypes and Robbery Victim resistance is itself a form of deterrence (Wright and Decker, 1997): I rob mostly whites... I usually don’t have no problem [with resistance], none at all. [Whites] got this stereotype, this myth, that a black person with a gun or knife is like Idi Amin or Hussein. And [a] person [who believes] that will do anything [you say] Whites accept the fact that they’ve been robbed... some blacks would rather die than give you they bucks and you damn near have to be killing [them] to get it Victim selection not based on anticipated cash holdings: most white people have about two dollars on them, and credit cards, something like that whites, they have credit cards and checkbooks on them... they get robbed, they cancel it all they got is plastic and checks O’Flaherty & Sethi Crime in Cities Columbia University 24 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 25 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 26 / 55 Implications of Victim Selection O’Flaherty & Sethi Crime in Cities Columbia University 27 / 55 Implications of Victim Selection Rates of Offending: Stereotypes of violence lower resistance This raises returns to offending Racial disparities in robbery offending differ from burglary or theft O’Flaherty & Sethi Crime in Cities Columbia University 27 / 55 Implications of Victim Selection Rates of Offending: Stereotypes of violence lower resistance This raises returns to offending Racial disparities in robbery offending differ from burglary or theft Violence Conditional on Resistance Black victims avoided by less desperate offenders Offender pool faced by white victims is less desperate on average Conditional on resistance, whites face violence less often O’Flaherty & Sethi Crime in Cities Columbia University 27 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 28 / 55 Differential Victimization and Segregation O’Flaherty & Sethi Crime in Cities Columbia University 29 / 55 Differential Victimization and Segregation Standard approaches to segregation: preferences or discrimination Differential victimization provides alternative explanation O’Flaherty & Sethi Crime in Cities Columbia University 29 / 55 Differential Victimization and Segregation Standard approaches to segregation: preferences or discrimination Differential victimization provides alternative explanation Whites rationally resist at lower rates, thus targeted more often Conditional on income, will choose structurally safer locations O’Flaherty & Sethi Crime in Cities Columbia University 29 / 55 Differential Victimization and Segregation Standard approaches to segregation: preferences or discrimination Differential victimization provides alternative explanation Whites rationally resist at lower rates, thus targeted more often Conditional on income, will choose structurally safer locations Racial disparities in victimization rates highest at intermediate income O’Flaherty & Sethi Crime in Cities Columbia University 29 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 30 / 55 Homicide O’Flaherty & Sethi Crime in Cities Columbia University 31 / 55 Homicide The only serious crime potentially motivated by preemption O’Flaherty & Sethi Crime in Cities Columbia University 32 / 55 Homicide The only serious crime potentially motivated by preemption If I go downstairs to investigate a noise at night, with a gun in my hand, and find myself face to face with a burglar who has a gun in his hand, there is a danger of an outcome that neither of us desires. Even if he prefers to just leave quietly, and I wish him to, there is danger that he may think I want to shoot, and shoot first. Worse, there is danger that he may think that I think he wants to shoot. Or he may think that I think he thinks I want to shoot. And so on. O’Flaherty & Sethi Crime in Cities Columbia University 32 / 55 Homicide The only serious crime potentially motivated by preemption If I go downstairs to investigate a noise at night, with a gun in my hand, and find myself face to face with a burglar who has a gun in his hand, there is a danger of an outcome that neither of us desires. Even if he prefers to just leave quietly, and I wish him to, there is danger that he may think I want to shoot, and shoot first. Worse, there is danger that he may think that I think he wants to shoot. Or he may think that I think he thinks I want to shoot. And so on. “Self-Defense” is ambiguous, when one is only trying to preclude being shot in self-defense. Schelling (1960) O’Flaherty & Sethi Crime in Cities Columbia University 32 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 33 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 34 / 55 Homicide Homicide rates depend on investments in lethality Newark: murders rose while shootings fell O’Flaherty & Sethi Crime in Cities Columbia University 35 / 55 Homicide Homicide rates depend on investments in lethality Newark: murders rose while shootings fell Need model of endogenous lethality with preemptive motive O’Flaherty & Sethi Crime in Cities Columbia University 35 / 55 Homicide Homicide rates depend on investments in lethality Newark: murders rose while shootings fell Need model of endogenous lethality with preemptive motive Three lethality levels: unarmed, low, high with different costs Unobserved heterogeneity in costs of killing Preemptive motive implies strategic complementarity in lethality Bayes-Nash equilibria have threshold structure (cost maps to lethality) Can have multiple equilibria and a murder multiplier O’Flaherty & Sethi Crime in Cities Columbia University 35 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 36 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 37 / 55 Homicide When killing is common, climate of fear prevails O’Flaherty & Sethi Crime in Cities Columbia University 38 / 55 Homicide When killing is common, climate of fear prevails Preemptive motive looms large, feedback to high homicide rates O’Flaherty & Sethi Crime in Cities Columbia University 38 / 55 Homicide When killing is common, climate of fear prevails Preemptive motive looms large, feedback to high homicide rates Small reductions hard to sustain, large reductions possible But require coordinated change in expectations Example: Operation Ceasefire O’Flaherty & Sethi Crime in Cities Columbia University 38 / 55 Racial Disparities in Victimization and Offending Fear depends on observable characteristics of others O’Flaherty & Sethi Crime in Cities Columbia University 39 / 55 Racial Disparities in Victimization and Offending Fear depends on observable characteristics of others Individuals who are feared will be preemptively killed more often And individuals with more fear will kill more often O’Flaherty & Sethi Crime in Cities Columbia University 39 / 55 Racial Disparities in Victimization and Offending Fear depends on observable characteristics of others Individuals who are feared will be preemptively killed more often And individuals with more fear will kill more often Historically, penalties lower if victim is black Raises incentives for pre-emptive killing Effect is especially strong if both parties fear being killed pre-emptively O’Flaherty & Sethi Crime in Cities Columbia University 39 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 40 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 41 / 55 Stand-your-Ground Laws Broaden notion of justifiable homicide Makes threatened individuals more dangerous Hence more likely to be killed preemptively Evidence seems to support this (Cheng and Hoekstra, 2013) O’Flaherty & Sethi Crime in Cities Columbia University 42 / 55 Street Vice O’Flaherty & Sethi Crime in Cities Columbia University 43 / 55 Street Vice Prostitution, illegal gambling, drug selling O’Flaherty & Sethi Crime in Cities Columbia University 44 / 55 Street Vice Prostitution, illegal gambling, drug selling Diffuse demand but highly concentrated supply Models of spatial competition can be adapted to this Fixed costs of protection, transportation carries risks O’Flaherty & Sethi Crime in Cities Columbia University 44 / 55 Street Vice Prostitution, illegal gambling, drug selling Diffuse demand but highly concentrated supply Models of spatial competition can be adapted to this Fixed costs of protection, transportation carries risks Greater demand density implies more sellers per unit distance Central cities will have lower prices and higher seller density Competition from city lowers suburban prices and seller density Strong enough effect can eliminate suburban markets O’Flaherty & Sethi Crime in Cities Columbia University 44 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 45 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 46 / 55 Race and Street Vice What accounts for racial disparities in exposure to street vice? O’Flaherty & Sethi Crime in Cities Columbia University 47 / 55 Race and Street Vice What accounts for racial disparities in exposure to street vice? Drug selling generates negative externalities: robbery, homicide Courts not used for dispute resolution; pervasive threat of violence Drug sellers are attractive robbery targets Marginal costs of killing lower for drug sellers Negative externalities cause exit of the more affluent non-users O’Flaherty & Sethi Crime in Cities Columbia University 47 / 55 Race and Street Vice What accounts for racial disparities in exposure to street vice? Drug selling generates negative externalities: robbery, homicide Courts not used for dispute resolution; pervasive threat of violence Drug sellers are attractive robbery targets Marginal costs of killing lower for drug sellers Negative externalities cause exit of the more affluent non-users Preferences over racial composition can then give rise to segregation Street vice correlated with race through neighborhood dynamics Not because vice originates in or moves to black neighborhoods O’Flaherty & Sethi Crime in Cities Columbia University 47 / 55 Some Lessons for Law Enforcement Street vice requires coordinated expectations between buyers, sellers Most desirable locations will be occupied by best protected sellers With well settled expectations, violence can be limited Disruption by law enforcement can trigger battles for best locations Law enforcement should target least lucrative locations first O’Flaherty & Sethi Crime in Cities Columbia University 48 / 55 Police Killings O’Flaherty & Sethi Crime in Cities Columbia University 49 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 50 / 55 O’Flaherty & Sethi Crime in Cities Columbia University 51 / 55 Safe and Risky Encounters Is equality of arrest and victimization rates the right benchmark? O’Flaherty & Sethi Crime in Cities Columbia University 52 / 55 Safe and Risky Encounters Is equality of arrest and victimization rates the right benchmark? Only if objectively threatening encounters have equal frequency O’Flaherty & Sethi Crime in Cities Columbia University 52 / 55 Safe and Risky Encounters Is equality of arrest and victimization rates the right benchmark? Only if objectively threatening encounters have equal frequency Suppose equal frequency of risky encounters and no police bias Proportion unarmed among victims should match proportion armed Guardian and Washington Post data can be used to test this O’Flaherty & Sethi Crime in Cities Columbia University 52 / 55 Washington Post 2015 Data 400 350 300 250 200 150 100 50 0 gun O’Flaherty & Sethi knife vehicle Crime in Cities toy weapon unarmed Columbia University 53 / 55 The Guardian 2015 Data 350 300 250 200 150 100 50 0 gun O’Flaherty & Sethi knife Crime in Cities vehicle unarmed Columbia University 54 / 55 Mass Incarceration Why is mass incarceration tolerated? O’Flaherty & Sethi Crime in Cities Columbia University 55 / 55 Mass Incarceration Why is mass incarceration tolerated? Could it be because of its racial character? O’Flaherty & Sethi Crime in Cities Columbia University 55 / 55 Mass Incarceration Why is mass incarceration tolerated? Could it be because of its racial character? Essentialist interpretations of equilibrium phenomena O’Flaherty & Sethi Crime in Cities Columbia University 55 / 55 Mass Incarceration Why is mass incarceration tolerated? Could it be because of its racial character? Essentialist interpretations of equilibrium phenomena Facilitated by cognitive load, historical superstructure of beliefs O’Flaherty & Sethi Crime in Cities Columbia University 55 / 55 Mass Incarceration Why is mass incarceration tolerated? Could it be because of its racial character? Essentialist interpretations of equilibrium phenomena Facilitated by cognitive load, historical superstructure of beliefs See Glenn Loury’s Du Bois and Tanner lectures O’Flaherty & Sethi Crime in Cities Columbia University 55 / 55