Education, Crime, and Neighborhood

1
Education, Crime, and Neighborhood
2
Neighborhood Effects

• Early work sought to measure education and crime
  as neighborhood effects.
• IDEA. We could measure the impacts (and possibly
  the benefits) of improved education or safer
  neighborhoods.
• P a ? biSi ? cjEducj ? dkCrimek
• where Si refers to structural characteristics
  EDUCj to Educational characteristics CRIMEk to
  Crime characteristics.

3
Shorthand

• P a ? biSi ? cjEducj ? dkCrimek
• P a b1s1 b2s2 ...
• c1Educ1 c1Educ1
• d1Crime1 d2Crime2
• Presumably, the cj and dk tell us something about
  impacts of good schools and safe neighborhoods.

4
Neighborhood Effects

• P a ? biSi ? cjEducj ? dkCrimek
• Dubin and Goodman (1982) had 21 educational
  characteristics, and 12 crime characteristics.
• They were able to reduce these into

5
Neighborhood Effects

• 5 educational components
• Test scores (Stud. Perf.)
• Change in test score
• Teacher quality

• Change in Stud. Achieve. Test
• Staff Experience

• 3 crime components
• Violent crime
• Property crime
• Shopping center crime
• All of them mattered.

6
Neighborhood Effects

• Dubin/Goodman Coefficients, for Baltimore City
  (mean value 34,650).
• 1 unit ? in test scores ? 2253 ? in house price,
  about 7.
• For crime
• 1 unit ? in property crime ? 795 ?.
• 1 unit ? in violent crime ? 3143 ?.
• 1 unit ? in shopping center crime ? 3721 ?

7
Victimization
Crime rate per 100,000
8
(No Transcript)
9
Victimization

• Violence ?low income victims
• Property crime ? higher income victims
• Central city residents are more likely to be
  victims.
• Blacks are more likely to be victimized.

10
Costs of Crime

• Direct Real Costs
• Death, injuries
• Must calculate dollar values for injuries
• Must impute costs for death
• Direct Transfer Costs
• Theft, fraud
• These are transfers because wealth is being
  transfered rather than destroyed.

11
Costs of Crime

• Indirect Real Costs
• Prevention costs. If we didnt have robberies,
  we wouldnt need to put locks on our homes.
• Criminal justice costs. If we didnt have
  crimes, our criminal justice/public safety system
  (police, courts, firefighters) would be much
  smaller.

12
Source Freeman 1996
13
Crime as a Rational Activity

• Premise Expected benefits exceed expected costs.
• Return to crime exceeds return to other plausible
  occupations.
• First, expected benefits.

Expected Loot E (L) prob of success
Value of Loot
So, if prob. of success is 0.6, and loot is
1200, expected loot is E (L) 0.6 1200 720
14
Crime as a Rational Activity

• Next, expected costs. Assume one year sentence if
  caught.
• Simplest model looks at costs as the probability
  of going to prison for one year, and the costs of
  being in prison.
• Expected prob. of going to prison is
• Prob of getting caught
• Prob of getting sentenced.
• EXCEL

Expected Cost E (L) Probc Probs
Cost
So, if Probc is 0.3, and Probs is 0.2, and
foregone wages are 3,000, then expected costs
are 0.3 0.2 3000 0.06 3000 180
15
(No Transcript)
16
Who Commits Crimes?

• Some people are skillful. They have low
  probabilities of getting caught and convicted.
  Expected loot may be high expected costs are
  low.
• Some have low opportunity costs.
• Some dont like society much so theyre not too
  averse to doing crimes.

17
Supply Curve for Burglary
S (more aversion)

• How many burglaries are committed?
• Depends on return.
• Depends on aversion to crime.

S (some aversion)
Net Return
S (no aversion)
of burglaries
18
Why Did Crime Drop in the 1990s?
Reduction
19
Why Did Crime Fall in 1990s?

• Strong Economy. There were more jobs and higher
  wages, causing a 2 reduction in property crime.
• Demographics. A decrease in the share of
  population in the crime-prone years of 16-24.
• Police Techniques. Including community policing
  and more aggressive control of public nuisances.
• Increase in Police. Increase of about 14, at a
  cost of 8.5 billion per year.
• Decrease in Crack.
• Legalized Abortion (1974). It appears that crime
  rates are higher among children born to reluctant
  parents.