Policy Considerations and a Preview of the Vehicle Stop Analysis Workshop

1
Policy Considerations and a Preview of the
Vehicle Stop Analysis Workshop

• Lorie A. Fridell, Ph.D.
• Director of Research
• PERF

2
Todays Presentation

• Discuss importance of policy in an agencys
  response to RBP
• Discuss various types of policies (advocating
  two)
• Provide preview of workshop on 3rd day
  Analysis/interpretation of vehicle stop data

3
PERFs 2001 Report Racial Biased Policing A
Principled Response

• PURPOSE Provide guidance to law enforcement
  agency executives in their response to RBP and
  perceptions of RBP
• FUNDING COPS Office
• AVAILABILITY www.policeforum.org

4
Agencies can respond to RBP in the following
realms

• Supervision/Accountability
• Policy
• Recruitment/Hiring
• Training/Education
• Minority Community Outreach
• Data Collection

5
A Key Theme Bias might manifest in various ways

• A few racist officers who act upon those biases
• The overwhelming number of well-meaning,
  professional officers many of whom might not be
  fully cognizant of the extent to which bias
  enters into their decision-making.
• Agency-level (not just individual-level)
  practices, policies, decisions.

6
Bad Apples(Racist officers who act on their
biases)

• Unlikely to be significantly impacted by policy
  and training
• If they were acting within current policy .
  wouldnt be bad apples.
• Some methods
• Early Warning Systems
• Effective Supervision
• Accountability through discipline.

7
Well-Meaning Officers

• Likely many are not fully cognizant of the extent
  to which race/ethnicity are used in their
  decision-making
• Policy to guide them on when race/ethnicity are
  and are not appropriate to use
• Training
• In that policy
• That facilitates an analytical understanding of
  RBP to promote correct behavior
• Leadership.

8
Institutional factors that contribute to the
problem

• What policies, enforcement strategies, reward
  structures, hiring/promotional activities, etc.
  might be contributing to RBP (or perceptions)?
• What internal racial tensions/bias exist?

9
Anti-Biased Policing Policy

10
Need Policies to Guide (well-meaning) Officers in
the Use of Race/Ethnicity in Making Decisions

• Focus Groups Different views among personnel
  within same agency regarding whether/how
  race/ethnicity can/should be used to make
  decisions
• This ambiguity creates great risk of biased
  policing activities.

11
Written policies must reflect definition of RBP

• Racially biased policing occurs when the police
  inappropriately consider race or ethnicity in
  deciding with whom and how to intervene in an
  enforcement capacity.

12
Find the line ..
Inappropriate Use of Race Based on
stereotypes, biases, etc.
Appropriate Use of Race Legally relevant
Appropriate Use of Race Legally relevant

13
Three Major Models

• Anti-Racial Profiling Policies
• Suspect-specific policies
• PERF Report Policy.

14
From most restrictive to least
15
Anti-Racial Profiling Policies

16
Anti-RP Policies

• Do not intervene (stop, arrest, search) solely
  on the basis of race
• Positive Convey a message
• But do not provide new guidance to personnel

17
Surely officers knew before such policies were
adopted that they could not intervene in a law
enforcement capacity SOLELY on basis of race.

• Must provide more meaningful guidance.

18
Suspect-Specific Policies

19
Suspect-Specific Policies

• Officers may not consider race or ethnicity of a
  person in the course of any law enforcement
  action
• UNLESS the officer is seeking to detain,
  apprehend, or otherwise be on the lookout for
• a SPECIFIC SUSPECT sought in connection with a
  SPECIFIC CRIME who has been identified or
  described in part by race or ethnicity.

20
Suspect-Specific Policies (Cont.)

• Example If looking for a suspect--reliable
  information indicates male, 58, lean,
  long-haired, pierced nose, and is ASIAN
• ASIAN can be considered (along with the other
  demographics, evidence) in developing RS or PC to
  detain/arrest.

21
In agencies with Suspect-Specific policies..

• These circumstances are the ONLY ones in which
  race/ethnicity can be used to make decisions.

22
PERF Report Policy

23
PERF Report Policy

• Encompasses the Suspect-Specific provision
• But allows for additional uses of race beyond
• specific known suspect
• specific crime
• Based on S-S principles.

24
Has both 4th and 14th Amendment provisions

25
4th Amendment Provision

• Officers shall not consider race/ethnicity to
  establish reasonable suspicion or probable cause
  EXCEPT ..

26
Exception

• Officers may take into account the reported
  race/ethnicity of a potential suspect(s) based on
  trustworthy, locally-relevant information that
  links a person or persons of a specific
  race/ethnicity to a particular unlawful
  incident(s).

27
PERF Report Policy

• Disallows race used as a general indicator for
  criminal behavior
• Disallows use of stereotypes/biases
• Allows for the consideration of race AS ONE
  FACTOR in making law enforcement decisions IF
• trustworthy and locally relevant information
• links specific suspected unlawful activity to a
  person or persons of a particular race/ethnicity
  
• Relies on Descriptions of actual suspects, not
  general Predictions of who may be involved in
  crime.

28
The Principles Underlying PERF Report Policy

29
Principle Race/ethnicity should be treated like
other demographic descriptors

• Police can use race/ethnicity as one factor in
  the same way that they use age, gender, etc. to
  establish RS/PC.

30
Opposing ArgumentRace/Ethnicity Are Different

• Race/ethnicity are among a group of factors that
  have heightened constitutional protection (others
  include, e.g., religion, gender)
• Does this mean we should also give gender this
  special status in guiding police behavior? ..

31
Race/Ethnicity are Different(Cont.)

• No, race/ethnicity are different (particularly in
  the law enforcement context), because
• We have heightened community concern about the
  use of race/ethnicity (not gender).
• We have prejudices in society vis a vis
  race/ethnicity that provide for the potential
  abuse.
• Relatedly, we have a history of actual abuse
  and/or perceived abuse on the part of police vis
  a vis race/ethnicity (not gender).

32

• Again, our principle Race/ethnicity should be
  treated like other demographic descriptors

33
Principle We use (or should use) demographic
information in policing in the manner articulated
in the PERF Policy

• That is, the parameters on the use of
  race/ethnicity are the same ones that do (or
  should) apply to other demographics (e.g.,
  gender, age).

34
That is

• Information on Demographic A (e.g., age) can be
  considered as one factor
• IF trustworthy, locally relevant information
• Links specific suspected unlawful activity to a
  person or persons who manifest Demographic A
  (e.g., age).

35
Opposing Argument

• We should focus only on behavior, not
  demographics
• (At least outside of a suspect-specific
  description.)

36
Example Using Demographic, Age

• Graffiti problem at particular location
• Credible witnesses describe several perpetrators
• 54 white male juvenile, red hair, blue
  sweatshirt and khakis, tattoo
• 53 black female juvenile, black hair, jeans and
  red t-shirt
• 5 1 Asian male juvenile, tattoo, Redskins
  jacket and jeans
• Etc

37
Example (Cont.)

• Officer is charged with stopping this behavior
• She will use multiple factors to establish
  reasonable suspicion prior to any detention
• Is juvenile relevant to her activity??
• PERF Report Policy Yes

38
Multiple factors and juvenile can be one of
them..
39
Because, we have

• Trustworthy, locally relevant information
  (witnesses)
• that links a person or persons of a specific
  demographic (juvenile)
• to particular unlawful incidents (graffiti in a
  particular location).

40
Example Race/Ethnicity CAN be considered

• A number of middle school students have reported
  that adult, Hispanic men are selling guns to
  students in the area immediately surrounding the
  school.
• Officer could consider citizen ethnicity around
  the school as ONE factor in totality weapons
  charge. (Just as they could consider adult, men.)
• They would need other information to produce the
  totality of circumstances. Could not act just
  upon the ethnicity info.

41
Note, suspect-specific policy would allow police
to consider ethnicity here only

• To intervene with someone who matched multiple
  descriptors provided by a witness Adult, male,
  Hispanic, red t-shirt, carrying paper bag.
• Hispanic could be used if part of an overall
  description of an individual
• and then police could only intervene with an
  individual matching that description.

42
Use of Race/Ethnicity Not OK

• Officer sees poorly dressed young African
  American male walking in an upper-class white
  neighborhood
• This is NOT a sufficient basis for a detention
  AND, further, should not be used as a basis for a
  pretext stop.
• That is, the policy precludes the
  race-out-of-place stops.

43
Importantly
Only talking about using race/ethnicity as ONE
factor among multiple factors in establishing RS
or PC
44
Difference as applied to 9-11 events

• Suspect-Specific Policies
• Middle-Eastern descent can be used (as one factor
  in a set of factors) to establish RS/PC
• when seeking a particular known suspect involved
  in the terrorism
• if reliable information indicates the particular
  known suspect being sought is of Middle-Eastern
  descent.
• That is, looking for Person A. (Or looking for
  A, B and C.) Information indicates A is of
  Middle-Eastern descent. Can consider this
  information.

45
PERF Report Policy Application

• Middle Eastern descent can be used (as one factor
  in a set of factors) to establish RS/PC
• when seeking people involved in a terrorism
  episode, although NOT NECESSARILY PARTICULAR ONES
  that you know did particular things
• IF reliable information links persons of M-E
  descent to this particular unlawful incident(s).
  
• Looking for A,B,C, and ?????.

46
Example Continued

• Suspect-Specific If looking for Suspect A who
  we know did X and police have reliable
  information indicating Suspect A is of Middle
  Eastern descent, can use this factor as ONE .
• PERF REPORT If looking for suspects and police
  have reliable information indicating the suspects
  are of Middle Eastern descent, can use this
  factor as ONE..

47
Applying the PERF Provisions

• Trustworthy Information Numerous and reliable
  sources point to involvement of persons of Middle
  Eastern Descent
• Locally Relevant Information Does not have to
  be locally BASED/GENERATED, but it must be
  reasonable to believe that it is relevant to the
  local area.
• Multi-site attack on 9/11
• Reasonable concerns in many cities.

48
Link to specific suspected unlawful activity

• That is, cannot use Middle Eastern descent as a
  factor if investigating e.g., home burglaries
• The trustworthy, locally-relevant information
  linked persons of Middle Eastern descent to
  terrorist acts.

49
Applying PERF Policy
Other possible factors to establish Reasonable
Suspicion Reports of credible informants
Suspicious Behavior etc.
50
Again, and importantly,
Only talking about using race/ethnicity as ONE
factor among multiple factors in establishing RS
or PC
51
Also, again

• Relies on descriptions of suspects not
  predictions of criminality.

52
14th Amendment Provision

• Thus far, we have focused on the 4th Amendment
  provision of the policy
• Indicating when police can use race as one factor
  in a set of factors to establish RS or PC.

53
Need for this equal protection provision

• The 4th Amendment provision is necessary, but not
  sufficient.
• Officers may meet all 4th Amendment requirements
  of policy/law and still be biased in their
  treatment of citizens (e.g., deciding which
  lawbreakers they will detain, cite, arrest show
  respect).

54
Example of the Gap that the 14th Amendment
Provision Fills

• Officer X always abides by 4th A provisions
  (including, but not limited to race/ethnicity
  provisions in PERF policy)
• But s/he
• Always tickets African American DUI offenders,
  lets Caucasians off
• Treats Caucasians with respect, is verbally
  abusive towards Hispanics.

55
Need an additional provision to highlight equal
protection in all police activities.

56
14th Amendment Provision

• Except as provided above, race/ethnicity shall
  not be motivating factors in making law
  enforcement decisions.

57
We need this second provision to prohibit, for
instance

• Disproportionately arresting minorities (but not
  non-minorities) for noise violations because of
  their race/ethnicity
• Disproportionately targeting minorities for Whren
  stops, because of their race/ethnicity
• Treating persons with disrespect because of their
  race/ethnicity.

58
Sets up the but for test for officers

• Would I be engaging this particular person but
  for the fact that this person is Hispanic?
• Would I be asking this question of this person
  but for the fact that this person is African
  American?

59
Together, the two provisions

• Prohibit racially biased policing
• Tightly circumscribe use of race/ethnicity in
  making decisions
• Prompt officers to carefully consider their
  motives for engaging individuals.

60
Recommendation

• Should consider adopting a policy that guides
  your officers in the use of race/ethnicity to
  make law enforcement decisions
• Recommend it be at least as restrictive as PERFs
  or go further and adopt the suspect-specific
  model.

61
Increasingly, agencies have included in policy
Provisions to reduce PERCEPTIONS of RBP

• PERF Policy
• Be courteous and professional.
• Introduce self and state reason for stop (unless
  there is a reason not to).
• Keep detention as brief as possible and inform
  driver of reason for delays.
• Apologize or at least explain if officer
  determines that the reasonable suspicion was
  unfounded (e.g., after investigatory stop).

62
Info on PERF Report Policy

• Chapter 4 of PERF Report at www.policeforum.org
  (Racially Biased Policing)
• ALSO (At same location) Supplementary
  discussion paper regarding principles on which
  policy is based.

63
Workshop Analyzing and Interpreting Vehicle
Stop Data

64

• Wednesday 830-5
• Team
• Lorie Fridell, PERF
• Geoff Alpert, University of S. Carolina
• Robin Engel, University of Cincinnati
• Amy Farrell, Northeastern University
• David Harris, University of Toledo
• John Lamberth, Lamberth Associates

65
2001 PERF Report Covers

• Arguments for/against data collection
• Recommendations regarding
• What activities to target
• What data elements to collect
• Preliminary discussion of benchmarking
  (Comparison groups for analyzing data).

66
Purpose of Workshopmove into analysis/interpretat
ion

• Provide a conceptual overview of the purpose and
  challenges of benchmarking
• Provide scheme for assessing benchmarking quality
• Provide overview of various benchmarking methods,
  and assessment of their quality
• Discuss how to analyze search and disposition
  data.

67
Quality of analysis nationwide

• Appears that the large majority of vehicle stop
  reports produced nationwide do not meet social
  science standards
• Most draw conclusions that are not supported by
  the results.

68
Workshop team will

• Discuss the pitfalls of bad analyses and
• Provide options for stronger analyses
• or at least more responsible interpretation of
  results.

69
Some topics

70
Many of you understand intuitively that census
benchmarking cannot prove the existence or lack
of existence of RBP..

71
Data Police Stops of Males v. Females
72
With data on residential population added
73
Is this proof of gender-biased policing?

• NO. And census benchmarking cannot prove
  RACIALLY biased policing either.

74
Wednesday

• Well explain why.

75
Many folks are unaware of the many alternative
methods for benchmarking.

• Well tell you what they are.
• Will describe at least 7 major benchmarking
  methods.

76

• Unadjusted census benchmarking
• Adjusted census benchmarking
• Observation benchmarking
• Crash data benchmarking
• Internal benchmarking
• Blind versus Not Blind benchmarking
• Survey benchmarking

77
Well cover searches

• If detained minorities are searched more than
  non-minorities, should your agency be accused of
  RBP?

78
Example

• An agencys data indicates that 25 of detained
  minorities are searched.
• 15 of detained Caucasians are searched..

79
Percent of detained persons searched
80
Can these percentages be used to draw conclusions
about whether bias is a part of this agencys
practices?

• Answer Wednesday.

81
Whats all the fuss about search hit rates?

• Search Hit Rate
• Percentage of searches that result in a find of
  seizable material.

82
The Results from Hypothetical Agency A Search
Hit Rates by Racial/Ethnic Group
83
Is this a red flag for racial bias in this
agency?

• Yes, well tell you why this is so.

84
Fact or fiction?

• F or F? Even though census-benchmarking is weak,
  if the results show no disparity this means no
  RBP.
• F or F? Even if a benchmark is weak, it can be
  used as a viable baseline for future years.
• F or F? Even if a benchmark is weak, if you find
  disparity over and over in all precincts, this is
  evidence of RBP.
• Well tell you the answers.

85
Who should be interested?

• Every law enforcement practitioner and other
  stakeholders concerned about RBP
• You should be interested if
• You are collecting data
• You will be collecting data
• State legislation on data collection is
  contemplated (or could be)
• Some residents in your jurisdiction want your
  agency to collect data

86
You should be interested if (cont.)

• You want to know why so many of the reports out
  there are sub-standard
• You are curious about this critical LE issue
• You want to understand how social science can be
  used to help us understand policing.

87
Workshop content based on

• Forthcoming PERF how to guide on
  analysis/interpretation
• Entitled By the Numbers A Guide for Analyzing
  Race Data from Vehicle Stops
• Funded by COPS
• Volume I on PERF website in December.
• Extensive and sophisticated analyses of vehicle
  stop data conducted by the workshop team members.