Crime Readings

Question 1

Raphael (2011) - Punishment

Answer 1

17% increase in incarceration due to crime rates, other 83% due to incarceration policy

US stats 1984-2002

Question 2

Levitt (1998) - Punishment

Answer 2

Murder not affected by either deterrence or incapacitation

Incapacitation effects significant for rape and robbery (and larger than deterrence), not significant for other crimes

Property crimes (burglary, larceny and auto theft) and assault, deterrence effects much larger than incapacitation, explaining 75% of overall effect

Question 3

Mueller-Smith (2015) - Punishment

Answer 3

Each year of incarceration leads to 3.6% decline in employment for felony offenders.

4.5% decline for misdemeanour offenders (bigger fall due to larger original P(employment)

Question 4

Bayer et al. (2009) - Punishment

Answer 4

Peer effects in juveniles increases reoffending

Question 5

Bhuller et al. (2018) - Punishment

Answer 5

Incarcerations decrease P(reoffending) within 5 years and number of offences over same period.

Norwegian data where emphasis on rehabilitation and education.

Question 6

Ehrlich (1975) - Punishment

Answer 6

Death penalty lead to reduction of 7-8 homicides per execution. Cited in Gregg discussion which ended moratorium in 1976.

Later the methodology and results completely refuted, e.g. D&W (2006)

Question 7

Donohue & Wolfers (2005)

Answer 7

Death penalty used so rarely that its effect on homicides cannot be disentangled from yr-to-yr changes in homicide rates.

Not just problem of significance, positive or negative result is extremely sensitive to very small changes in econometric specifications.

Question 8

Lochner and Moretti (2004) - Ed, Un and JC - Incarceration effects

Answer 8

Extra year education reduces P(incarceration) 0.1% for whites, and 0.37% for African-Americans

Implying as much as 23% of the 1980 Black-White incarceration gap was due to the average gap in education levels.

Question 9

Lochner and Moretti (2004) - Ed, Un and JC - Savings

Answer 9

A 1% increase in high school completion rate of all men aged 20-60 would save the US as much as $1.4bn per year in reduced costs from crime incurred by victims and society at large.

This is $2,100/graduate/year, which is 14-26% of the yearly private return to high school graduation.

Question 10

Deming (2011) - Ed, Un and JC

Answer 10

School quality lottery between local school and another with expected median quality.

Reduces crime, concentrated with black males and high-risk youth who committed 50% less crime.

Nearly all reduction in crime after completion of school, and persists for 4-7 years after random assignment.

Question 11

Raphael and Winter-Ebmer (2001) - Ed, Un and JC

Answer 11

US Data from 1971 to 1997

Use IV of military defence contract spending to instrument for employment rates as they are bound to be correlated.

Also include state-specific measure of oil price shocks (intensity employment in manufacturing X change in relative price of oil).

Significant effect on unemployment levels on property (socioeconomic) crimes, where valuable property is stolen in absence of work.

Find fall in US unemployment from 7.4% to 4.9% between 1992-1997 can account for slightly more than 40% of the decline in overall property crime rate.

Question 12

Farrington et al. (1986) - Ed, Un and JC

Answer 12

Also found unemployment had positive effect on crimes with material gain.

Question 13

Mocan et al. (2005) - Ed, Un and JC

Answer 13

Model where individual builds up human or criminal capital, which affects incentives of legitimate/criminal work in next period.

Suggests likelihood of being criminal next period is increased by criminality this period - persistence.

Concern in a recession that a dip into criminality for a couple periods may leads to build up of criminal capital, which makes it more profitable to remain a criminal when you wouldn’t have been one in a good job market.

Question 14

Bell et al. (2017) - Ed, Un and JC

Answer 14

Significant + effect of entry unemployment rate on subsequent arrests for both property and violent crime.

In a recession (5% higher unemployment) is associated with a 5.7% increase in P(ever arrested).

Effect persistent, even 20 years later. After controlling for the unemployment rate they face now, poor initial unemployment rate at labour market entry increases the likelihood of criminality.

Question 15

Mancino et al. (2016) - Ed, Un and JC

Answer 15

Compare accumulated experience (human/criminal capital) and the previous period’s criminal decision.

Finds that whilst both significant effects, find that lagged criminal decision has much larger effect on p(criminal participation)

Good news for effectiveness of policy, temporary shock that pushes people out of crime can have lasting effects even if human capital is largely unaffected.

Question 16

Knowles et al. (2001) - Race and Racial Profiling

Answer 16

Guilt defined as finding drugs, black-white hit rate almost identical (0.34 vs 0.32) - Hispanic rate 0.11 is significantly lower and implies racial prejudice against Hispanics.

More recent evidence suggests race of race of officer matters

Question 17

Pierson et al. (2017) - Race and Racial Profiling

Answer 17

Problem of inframarginality

Wary when comparing hit rates: if a region applying a race-neutral threshold for searching such as p(guilty)>10%, different groups may have different make-ups which erroneously suggests officers are racist when they deliberately applied a race-neutral threshold.

Question 18

Simoiu et al. (2017) - Race and Racial Profiling

Answer 18

Traffic stops in North Carolina, estimate significantly higher P(guilt) thresholds for stopping white and Asian drivers, (15%/13% respectively) than for black and Hispanic drivers (7% and 6% respectively)

Question 19

Goel et al. (2016) - Race and Racial Profiling

Answer 19

Weapon stops NYC

49% black suspects less than 1% chance of weapon, 34% Hispanics, only 19% of white stops did.

Indicates a substantial fraction of stops, were conducted on the basis of relatively little evidence.

Question 20

Glaser and Sacerdote (2003) - Race and Racial Profiling

Answer 20

Use DUI killings to remove endogeneity of relationship between driver and victim, effectively random as they’re out of control.

Drivers who kill African-Americans receive sentences that are approx. 60% shorter than the sentences received by drivers who kill whites.

Question 21

Mustard (2001) - Race and Racial Profiling

Answer 21

Black and Hispanic defendants get longer sentences.

Question 22

Blume et al. (2004) - Race and Racial Profiling

Answer 22

Blacks who murder whites more likely to get death penalty than any other perpetrator/victim combination

People who murder black people less likely to get the death penalty.

Question 23

Mueller-Smith and Schnepel (2018) - Race and Racial Profiling

Answer 23

Pausing/terminating a defendant’s progress through criminal justice system halves reoffending rates, increases quarterly employment by 53% over 10 years

Particularly important is avoiding the lifelong mark of a felony

Question 24

Glaser et al. (1996) - Gangs and Social Interactions

Answer 24

Social interactions fairly negligible for murder, rape and arson
Larger for robbery, assault and burglary,
Largest for auto theft and larceny

Found crimes by younger criminals had more social interactions

Argue that presence of “strong families” interferes with the transmission of criminal choices across individuals

Question 25

Silverman (2004) - Gangs and Social Interactions

Answer 25

Otherwise decent people can get drawn into violence to build a reputation, in order to protect themselves in the future

Question 26

O'Flaherty and Sethi (2010) - Gangs and Social Interactions

Answer 26

Murders in Newark Killing is a strategic complementarity, with an increased likelihood of violence, the reward from acting first and killing your aggressors first is higher. Drives expectation of high murder rates to be self-fulfilling, "murders beget murders"

Question 27

Kelly et al. (1993) - Gangs and Social Interactions

Answer 27

Chinese organised crime in NYC Due to lack of faith in the police, communities tolerate gangs up to a point. Toleration of gangs due to risk of disrupting the current equilibrium will lead to a worse gang, i.e. one that doesn't respect social norms like not selling to children.

Question 28

Akerlof and Yellen (1993) - Gangs and Social Interactions

Answer 28

Increasing the communities willingness to report crimes will help deal with gangs.

Question 29

Levitt and Ventakesh (2000) - Gangs and Social Interactions

Answer 29

When comparing two areas with similar demographics, those with "higher socioeconomic status and neighbourhood cohesions" suffered markedly less gang and drug activity.

Question 30

Akerlof and Yellen (2013) - Gangs and Social Interactions

Answer 30

“Stress manipulation of social values is as important in the control of crime as harsh punishments and high public expenditures for police” “We argue that these traditional approaches to crime control may prove counterproductive in the long run if they undermine community values” “The willingness of local residents to reveal information to the police, we shall argue, depends partly on the behaviour of the local gang but also on the legitimacy of the police in the community”

Question 31

Evans et al. (2018) - Drugs and Drug Policy

Answer 31

Systemic crime impact of gang violence in the 80s persisted for decades: in 2010, close to 1000 extra murders due to crack emergence

Question 32

Becker et al. (2004) - BGM - model implications Drugs and Drug Policy

Answer 32

“Under certain assumptions, including inelastic demand, a tax to discourage legal drug production would reduce consumption by more than an efficient war on drugs” “With inelastic demand, resources are actually drawn into the drug business as enforcement reduces consumption.”

Question 33

Becker et al. (2004) - BGM - Real world findings Drugs and Drug Policy

Answer 33

Making goods illegal and punishing by imprisonment is more costly to higher income persons, and hence tends to reduce consumption more than consumption of lower income persons Greater political influence of higher education and income groups would explain why drugs are illegal rather than subject to sizeable monetary excise taxes. Also explains why punishment mainly imposed on suppliers rather than consumers as they’re more likely to be low-income

Question 34

Galenianos et al. (2012) - Model description - Drugs and Drug Policy

Answer 34

Theoretical Search Model, where moral hazard from unknown quality until after purchase, and search costs in finding a seller This causes rip-offs with no quality (distribution has bunch at 0 and a long right tail). Also causes long-term buyer relationships in order to avoid rip-offs.

Question 35

Galenianos et al. (2012) - Policy implications - Drugs and Drug Policy

Answer 35

Can leverage this moral hazard, by using market for lemons theory, by inducing sellers to dilute more they hope the market will collapse. Induce lower quality sales by reduced sentences on low quality, which effectively increases the wholesale price of drugs, but without associated cost of war on drugs. Easily implemented with minor changes to sentencing guidelines, also reduces expenditures as incarceration rates would be lower.

Question 36

Kilmer et al. (2010) - Drugs and Drug Policy

Answer 36

Marijuana elasticity = -0.54

Question 37

Jofre-Bonet and Petry(2008) - Drugs and Drug Policy

Answer 37

Heroin elasticity = -0.9 for addicts = INELASTIC | Cocaine elasticity = -0.9 to -1.05 for addicts = INELASTIC to ELASTIC

Question 38

Caulkins and Reuter (2006) - Drugs and Drug Policy

Answer 38

Despite majority dropping out, and over half relapsing within 5 years, find rehab is cost-effective when compared with jail which they estimated at time of writing is $30,000 While only reducing lifetime consumption by single digit percentages, high school interventions are effective as so cheap.

Question 39

Bureau of Justice Statistics (2004) - Drugs and Drug Policy

Answer 39

Pharmacological crime – 26% federal prisoners committed their current offence while under the influence of drugs Socioeconomic crime – 18% of federal prisoners said they committed their current offence to obtain money for drugs Unfortunately, this doesn’t tell us what percentage of drug users commit property crimes

Question 40

Bhaskar et al. (2017) - Drugs and Drug Policy

Answer 40

Surprisingly, online drug markets don’t seem to suffer from substantial moral hazard problems associated with search models

Question 41

Krueger and Malečková (2003) - Education/Poverty - Terrorism

Answer 41

Terrorism not linked to poverty and lack of education, terrorists are relatively well educated and come from middle classes, consistent with Lerner (1958). Terrorism is a form of political engagement where middle classes more likely to be engaged. Increasing levels of education in Palestine negatively correlated with support for a dialogue with Israel.

Question 42

Abadie (2006) - Terrorism

Answer 42

Reiterates poverty not significant for terrorism risk once characteristics like political freedom accounted for. Political freedom has non-monotonic effect on terrorism (partial horseshoe) - countries with intermediate levels of political freedoms are most prone to terrorism As experienced in Iraq, transitions from authoritarian regime to democracy may be accompanied by temporary increases in terrorism (analogous to destabilising gangs)

Question 43

Abadie and Gardeazabal (2003) - Terrorism

Answer 43

Look at ETA violence in Basque region of NE Spain, build a synthetic Basque region based on economic growth determinants before violence outbreak (85% Catalonia and 15% Madrid), predicts what would have happened in absence of violence Find that per capita GDP in the Basque region declined about 10% relative to the synthetic control region, with the gap widening in response to spikes in terrorist activity As expected, as the truce became credible they show Basque stocks outperformed non-Basque stocks. In reverse, at the end of the cease-fire, Basque stocks showed a negative performance relative to non-Basque stocks.

Question 44

Krueger & Laitin (2007) - Terrorism

Answer 44

Evidence countries with higher GDP/capita more likely to be target of international terrorism. Not necessarily in the rich country, quite commonly close to the origin country, e.g. directed at US embassies.

Question 45

Besley and Mueller (2012) - Terrorism

Answer 45

Estimates the value of peace on property prices in Northern Ireland, finding peace leads to increase in house price of 1.3%-3.5%. Effect more pronounced in Belfast, gains between 5.9% and 16.6%

Question 46

Krueger and Malečková (2003) - Civil Liberties - Terrorism

Answer 46

Whilst poorest countries do spawn more terrorists, effect not present once controlling for measures of civil liberties. Countries with weaker civil liberties are the origin of a larger number of international terrorist events.

Question 47

Krueger and Malečková (2003) - Religion - Terrorism

Answer 47

While they find evidence that increased affiliation with 4 main religions (Islam, Christianity, Buddhism and Hinduism) is associated with incidence of terrorism. F-test cannot reject null hypothesis that each of the 4 religious groups have the same effect, not surprising as we've seen terrorism throughout the world.

Question 48

Abadie and Gardeazabal (2003) - Criticism of previous studies -Terrorism

Answer 48

Argue previous cross-country studies of political unrest on economic growth is inappropriate due to the heterogeneity between the natures of political unrest in each country. Benefit of studying unrest in an otherwise affluent area is that temporal variation in prosperity not considered likely to feed back into terrorist activity which mitigates endogeneity issues.

Question 49

Mancino et al. (2016) - Peer Effects - Ed, Un and JC

Answer 49

Looking at adolescents and young adults who have already committed criminal offences School enrolment increases violent and drug-related crimes, thus questioning the more schooling-less crime hypothesis. Find non-cognitive skills significantly reduce criminal participation, whilst cognitive ones don't Find a range of childhood risk factors significant, at least for young people who are already criminals

Question 50

Grogger and Ridgeway (2006) - Race and Racial Profiling

Answer 50

Use theory of 'veil of darkness' to overcome problems of find a benchmark population to compare to, instead use the same population in the benchmark but compare day to night. Assume racial differences in traffic patterns, driving behaviour and exposure to law enforcement only vary with time of day, not time of year Focusing sample on same time of day from 5pm to 9pm, find no evidence of racial profiling in traffic stops in Oakland California.

Question 51

Miron and Zwiebel (1995) - murders

Answer 51

March-October 1988, vast majority of drug related murders in 1988 were economic compulsive or systemic in nature, due to the violence associated with the illegal market. In comparison, 0 murders of these kinds were reported in relation to alcohol.

Question 52

Miron and Zwiebel (1995) - conclusions

Answer 52

"Our conclusion is that a free market in drugs is likely to be a far superior policy to current policies of drug prohibition." "On net, the existing evidence suggests the social costs of drug prohibition are vastly greater than its benefits."

Question 53

Gostin (1991) - Drugs and Drug Policy

Answer 53

Restrictions on clean needles increased the spread of HIV. HIV infection rates are lower in cities or countries with needle exchanges or greater legal access to clean needles.

Question 54

Ayres and Waldfogel (1994) - Race and Racial Profiling

Answer 54

Find bail bondsmen offer lower rates for minorities, which implies racial bias in setting bail (too high)