Analyzing Stops, Citations, and Searches in
Washington and Beyond
Mario L. Barnes
Professor of Law and Senior Associate Dean for Academic Affairs,
University of California, Irvine, School of Law

Robert S. Chang
Professor of Law, and
Executive Director, Fred T. Korematsu Center for Law and Equality

Seattle University School of Law Legal Paper Series # 12-25
Published in the Seattle University Law Review, Vol. 35 (2012)

This paper can be downloaded without charge from
the Social Science Research Network Electronic Paper Collection

Electronic copy available at: http://ssrn.com/abstract=2072434

 Analyzing Stops, Citations, and Searches in
Washington and Beyond
Mario L. Barnes∗ & Robert S. Chang∗∗
I. INTRODUCTION
Racial disproportionality1 in the criminal justice system is a fact.2
But the fact of racial disproportionality is the beginning and not the end
of the conversation. The fact that blacks are overrepresented in stop,3
arrest,4 charge,5 pretrial detention,6 conviction,7 and incarceration8 statis∗
Professor of Law and Senior Associate Dean for Academic Affairs, University of California, Irvine, School of Law.
∗∗
Professor of Law and Associate Dean for Research and Faculty Development; Executive Director,
Fred T. Korematsu Center for Law and Equality, Seattle University School of Law.
1. We follow the convention adopted in the Preliminary Report on Race and Washington’s
Criminal Justice System:
In this Report, we use disproportionality to refer to a discrepancy between reference
groups’ representation in the general population and in criminal justice institutions. In
contrast, we use disparity when similarly situated groups of individuals are treated differently within those institutions, or to refer to overrepresentation of particular groups in the
criminal justice system that stems from criminal justice practices or policies.
Task Force on Race & the Criminal Justice Sys., Preliminary Report on Race and Washington’s
Criminal Justice System, 35 SEATTLE U. L. REV. 623, 630 (2012) [hereinafter Task Force Report].
2. See generally Alfred Blumstein, On the Racial Disproportionality of United States’ Prison
Populations, 73 J. CRIM. L. & CRIMINOLOGY 1259, 1259 (1982) (“Although blacks comprise roughly one-eighth of the population, they represent about one-half of the prison population.”); AM. BAR
ASS’N, JUSTICE KENNEDY COMM’N, REPORT WITH RECOMMENDATIONS TO THE ABA HOUSE OF
DELEGATES 48 (2004) (“A report released by the Bureau of Justice Statistics found that a black male
had a 1 in 3 chance of being imprisoned during his lifetime, compared to a 1 in 6 chance for a Latino
male and a 1 in 17 chance for a white male.”), http://www.americanbar.org/content/dam/
aba/publishing/criminal_justice_section_newsletter/crimjust_kennedy_JusticeKennedyCommission
ReportsFinal.authcheckdam.pdf; MARK MAUER & RYAN S. KING, THE SENTENCING PROJECT,
UNEVEN JUSTICE: STATE RATES OF INCARCERATION BY RACE AND ETHNICITY (2007),
http://www.sentencingproject.org/doc/publications/rd_stateratesofincbyraceandethnicity.pdf
(discussing incarceration disproportionalities).
3. See, e.g., Angela Y. Davis, Race, Cops, and Traffic Stops, 51 U. MIAMI L. REV. 425, 431
(1997) (“Empirical evidence suggests that race is frequently the defining factor in pretextual traffic
stops.”).
4. Cf. Alfred Blumstein, Racial Disproportionality of U.S. Prison Populations Revisited, 64 U.
COLO. L. REV. 743 (1993) (discussing proportion of prison racial disproportionality that might be
accounted for by racial disproportionality at arrest).
5. See, e.g., Task Force Report, supra note 1, at 641 fig.1.

673

Electronic copy available at: http://ssrn.com/abstract=2072434

 674

Seattle University Law Review

[Vol. 35:673

tics demonstrates only correlation and not causation.9 A number of commentators caution that disproportionality10 and the overrepresentation of
blacks, Native-Americans, and Hispanics in Washington State’s prisons
do not prove racial discrimination.11 Further, the fact of disproportionality at each stage of criminal justice processing does not prove that racial
discrimination occurs at each particular stage.12 For example, the observed disproportionality at imprisonment might merely be a downstream
artifact of disproportionality at conviction, which might in turn be a

6. See Janet L. Lauritsen & Robert J. Sampson, Minorities, Crime, and Criminal Justice, in
THE HANDBOOK OF CRIME AND PUNISHMENT 58, 71 (Michael H. Tonry ed., 1998) (discussing the
relative lack of empirical research in this area, with a few prior studies tending “to show that the
direct influence of race on pretrial release is insignificant once a defendant’s dangerousness to the
community . . . and prior history of failing to appear at trial are controlled,” but elaborating that
“race is related to bail decision making in complex, interactive ways,” with one study finding that
“dangerousness and severity had stronger influences on bail decisions for whites” but that “in the
main they suggest that white defendants receive better returns on their resources.” (citation omitted)).
7. See Henry Greely et al., Family Ties: The Use of DNA Offender Databases to Catch Offenders’ Kin, 34 J.L. MED. & ETHICS 248, 258 (2006). (“African-Americans constitute about thirteen
percent of the U.S. population, or about thirty-eight million people. In an average year, over forty
percent of people convicted of felonies in the United States are African-American.”).
8. See, e.g., MAUER & KING, supra note 2; BUREAU OF JUSTICE STATISTICS, U.S. DEP’T OF
JUSTICE, PRISONERS IN 2010 7 tbl.14 (2012), http://bjs.ojp.usdoj.gov/content/pub/pdf/p10.pdf (As of
December 31, 2010, in state and federal prisons, “black non-Hispanic males had an imprisonment
rate (3,074 per 100,000 U.S. black male residents) that was nearly 7 times higher than white nonHispanic males (459 per 100,000),” and 7.3% of black males between the ages of 30–34 were incarcerated.).
9. Latino and Native-American disproportionality also exists within the criminal justice system, but those populations are not as consistently studied, and there are problems sometimes with
data collection. For example, in Washington State, many local jails do not collect ethnic demographic information, which likely leads to an undercount of Latinos and an overcount of whites and probably skews the reported Latino-White disproportionality to be lower than what it actually is. See
Task Force Report, supra note 1, at 639−40. For a general discussion about correlation and causation, see Brett Houlding & Simon P. Wilson, Considerations on the UK Re-Arrest Hazard Data
Analysis, 10 LAW, PROBABILITY & RISK 303, 320 (2011) (“Confounding occurs when correlation is
mistaken for causation, and such occurrences lead to spurious causal relationships.”).
10. See, e.g., Clayton Mosher et al., The Importance of Context in Understanding Biased Policing: State Patrol Traffic Citations in Washington State, 9 POLICE PRAC. & RES. 43, 44 (2008); Michael R. Smith, Depoliticizing Racial Profiling: Suggestions for the Limited Use and Management of
Race in Police Decision-Making, 15 GEO. MASON U. C.R. L.J. 219, 239 (2005).
11. Overrepresentation is demonstrated by comparing the prison population with the state
population. Compare WASH. STATE DEP’T OF CORR., QUARTERLY FACT CARD (2011), http://
www.doc.wa.gov/aboutdoc/docs/msFactCard_003.pdf, with U.S. CENSUS BUREAU, STATE &
COUNTY QUICK FACTS: WASHINGTON (2010), http://quickfacts.census.gov/qfd/states/53000.html.
12. Cf. Blumstein, supra note 2, at 1261 (exploring the “degree to
which . . . [disproportionality] is likely to have emerged as a consequence of racial discrimination . . . compared to the alternative explanation . . . of disproportionate involvement in criminal
activity, and particularly in the kind of criminal activity that is most likely to lead to imprisonment
and to longer sentences”).

Electronic copy available at: http://ssrn.com/abstract=2072434

 2012]

Analyzing Stops, Citations, and Searches

675

downstream artifact of pretrial detention, charge, arrest, or stop.13 Closer
analysis of each stage is required in order to determine whether observed
disproportionality is a product of legally relevant factors as opposed to
suggesting more strongly that race plays an illegitimate role.
This Article seeks to examine more closely the disproportionality
with regard to traffic stops, citations, and searches. We focus on three
reports produced by a team of researchers from Washington State University (WSU) that examine Washington State Patrol traffic stops, citations, and searches.14

II. ACCOUNTING FOR RACIALLY DISPARATE CITATION RATES
AMONG THE WASHINGTON STATE PATROL
In its most recent review of Washington State Patrol traffic stops, a
group of researchers from WSU concluded the following with regard to
stops and citations: “[W]hen racial differences in compliance with traffic
and safety laws are statistically controlled for, there is no evidence of
systematic racial bias on the part of the Washington State Patrol at the
level of which drivers are issued citations.”15 With regard to searches, the
researchers concluded that, although race continues to correlate with rate
of searches, there is
no evidence that these disparities at the bivariate level are the result
of intentional or purposeful discrimination, and thus we find no evidence of intentional “racial profiling” (evidence of purposeful or intentional discrimination is generally the first step required by the
federal courts when attempting to prove racial discrimination under

13. Id. (concluding that 80% of observed disproportionality at imprisonment for certain serious
crimes is likely an artifact of differential involvement in crime as measured by arrest). We note
though that Blumstein relies on national statistics and that there is significant variation at the state
level. See Robert Crutchfield, George Bridges & Susan Pitchford, Analytical and Aggregation Biases in Analysis of Imprisonment: Reconciling Discrepancies in Studies of Racial Disparity, 31 J. RES.
CRIME & DELINQUENCY 166, 179–89 (1994) (analyzing issue at the state level and observing that in
1982, only 20% of the Black-White disproportionality in Washington State prisons was an artifact of
disproportionality at arrest).
14. NICHOLAS P. LOVRICH ET AL., DIV. OF GOV’TL STUDIES & SERVS., WASH. STATE UNIV.,
REPORT TO THE WASHINGTON STATE PATROL 1 (2007) [hereinafter 2007 WSP REPORT], available
at http://www.wsp.wa.gov/publications/reports/wsu_2007_report.pdf; NICHOLAS P. LOVRICH ET AL.,
DIV. OF GOV’TL STUDIES & SERVS., WASH. STATE UNIV., ANALYSIS OF TRAFFIC STOP DATA
COLLECTED BY THE WASHINGTON STATE PATROL 1 (2005) [hereinafter 2005 WSP REPORT], available at http://www.policeforum.org/library/racially-biased-policing/supplemental-resources/wsu_20
05_report[1].pdf; NICHOLAS P. LOVRICH ET AL., DIV. OF GOV’TL STUDIES & SERVS., WASH. STATE
UNIV., WSP TRAFFIC STOP DATA ANALYSIS PROJECT REPORT 1 (2003) [hereinafter 2003 WSP
REPORT], available at http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.102.7223&rep=rep
1&type=pdf.
15. 2007 WSP REPORT, supra note 14, at 38 (emphasis removed).

 676

Seattle University Law Review

[Vol. 35:673

the Equal Protection [C]lause of the Fourteenth Amendment of the
U.S. Constitution).16

We do not take issue with the WSU researchers’ conclusion in the 2003
report, which stated, “The Washington State Patrol is one of a very few
agencies so far studied which does not exhibit a pattern of disproportionality at the stop level.”17 We also agree that the evidence does not support a finding of widespread intentional discrimination on the part of the
Washington State Patrol. We disagree, however, with WSU’s methodology for accounting for racial disproportionality at the citation level, and
we assert that there is a logical fallacy by which they discount racial disproportionality with regard to searches.
We note, though, that traffic stops conducted by the Washington
State Patrol constitute only a fraction of the stops that are made by law
enforcement agencies throughout Washington State. We applaud the
Washington State Patrol for collecting the data and for commissioning
reports by a team of researchers from WSU to analyze it. We encourage
other law enforcement units throughout the state to do the same.
A. Washington State Patrol Citations
First, it is refreshing that the three reports done by the WSU researchers take seriously the requirement for a nuanced study of the
meaning of race within the policing context. Recently, sociolegal scholars have called for empirical studies to be more sophisticated in their
considerations of race and racialized data.18 Law scholars have also started to organize projects that meld critical and empirical approaches to the
study of race.19 Along these lines, we appreciate the researchers’ at-

16. Id. at 49.
17. 2003 WSP REPORT, supra note 14, at 41 (emphasis removed). We do note, as they do, that
there are some Autonomous Patrol Areas (APAs) where significant disproportionalities exist. Id. We
suggest follow up work to examine more closely whether race-neutral factors can account for these
disproportionalities.
18. See, e.g., Laura E. Gómez, A Tale of Two Genres: On the Real and Ideal Links Between
Law and Society and Critical Race Theory, in THE BLACKWELL COMPANION TO LAW AND SOCIETY
455 (Austin Sarat ed., 2004) (opining that there is little to be gained from treating race as “an easily
measured independent variable” within an empirical study); Laura E. Gómez, Presidential Address,
45 LAW & SOC’Y REV. (forthcoming 2012); Richard Lempert, A Personal Odyssey Toward a
Theme: Race and Equality in the United States: 1948–2009, 44 LAW & SOC’Y REV. 431, 458 (2010)
(calling for those doing empirical work in law to “critically examin[e] the modern jurisprudence on
race”).
19. One such project is the Critical Race Theory and Empirical Methods Working Group
formed at UC Hastings Law School, which brings together critical legal scholars and social scientists
to explore methodologies for studying race. On deploying social science data related to policing, see
Eric J. Miller, What Criminal Law and Procedure Can Learn From Criminology: Putting the Practice into Theory, 7 OHIO ST. J. CRIM. L. 31 (2009) (discussing how courts need to be more careful in

 2012]

Analyzing Stops, Citations, and Searches

677

tempts to question the significance of the ostensibly racially disparate
statistics for traffic citations in Washington State. In particular, the researchers should be commended for their attention to refining their
methodology and exploring the benefits of multivariate analysis.20 Again,
it is helpful and encouraging that the data from the three separate time
periods reveal that, unlike earlier studies in other jurisdictions,21 there is
no discernible bias with regard to traffic stops (contacts). There were,
however, racially disparate results for average citation rates,22 average
number of violations per contact,23 and the seriousness of the violations
for which citations were issued.24 Interestingly, in these latter two data
their reliance on social science data to support policing practices); L. Song Richardson, Arrest Efficiency and the Fourth Amendment, 95 MINN. L. REV. 2035 (2011).
20. The researchers, for example, broke out East Indians as a distinct group in the 2007 report.
2007 WSP REPORT, supra note 14, at 3. They also were responsive to data suggesting overrepresentation for African-Americans in certain locations and/or at specific times. 2005 WSP REPORT, supra
note 14, at 11.
21. See, e.g., Clayton Mosher & J. Mitchell Pickerill, Methodological Issues in Biased Policing
Research with Applications to the Washington State Patrol, 35 SEATTLE U. L. REV. 769, 772−74
(2012) (discussing studies of racial profiling in San Francisco, New York City, Illinois, Kansas, and
Minnesota); IAN AYRES & JONATHAN BOROWSKY, ACLU OF SOUTHERN CAL., A STUDY OF
RACIALLY DISPARATE OUTCOMES IN THE LOS ANGELES POLICE DEPARTMENT (2008) (looking at
traffic stops in Los Angeles for an eleven-month period between 2003–2004, and finding that blacks
and Hispanics were stopped and searched by Los Angeles Police Department officers significantly
more frequently than whites), http://www.aclu-sc.org/downloads/2/681086.pdf; DAVID A. HARRIS,
ACLU, DRIVING WHILE BLACK: RACIAL PROFILING ON OUR NATION’S HIGHWAYS (1999) (looking
at data reflecting racial profiling of Hispanics in Illinois and blacks in Pennsylvania and Maryland),
http://www.aclu.org/racial-justice/driving-while-black-racial-profiling-our-nations-highways; 2005
WSP REPORT, supra note 14, at 27–28.
22. 2007 WSP REPORT, supra note 14, at 24 (“Black drivers were somewhat more likely to be
issued traffic citations . . . than were White motorists in 23 of 34 APAs, Native Americans were
more likely to be issued citations than Whites in 29 APAs, Asians were likewise more likely than
Whites to be issued citations in 20 APAs, East Indians were likely to suffer the same fate in 20
APAs, and finally Hispanics were more likely to receive citations in 29 APAs across the state.”);
2005 WSP REPORT, supra note 14, at 29 (“The findings reported . . . that Blacks were more likely to
be issued citations than Whites in 27 of 40 APAs, Native-Americans were more likely to be issued
citations than Whites in 38 APAs, Asians in 25 APAs, Hispanics in 39 APAs, and East Indians in 26
APAs.”); 2003 WSP REPORT, supra note 14, at 79 (“Blacks received a greater number of citations in
12 APAs, Native Americans received a greater number of citations in 15 APAs (and a lower number
of citations in one APA), Asians received a significantly higher number of citations in 15 APAs,
while Hispanics received a significantly greater number of citations in 29 APAs.”).
23. For the 2001–2002 data, average citation rates per contact were as follows: Asians (1.71),
Whites (1.74), Blacks (1.94), Hispanics (1.98), and Native-Americans (2.05). See Mosher et al.,
supra note 10, at 46; see also 2007 WSP REPORT, supra note 14, at 26 (discussing the greater average citation rates per contacts for blacks, Hispanics, and Native-Americans); 2005 WSP REPORT,
supra note 14, at 30–31 (same).
24. For the 2001–2002 data, citation seriousness rates were as follows: Asians (.14), Whites
(.19), Blacks (.31), Hispanics (.33), and Native-Americans (.45). Mosher et al., supra note 10, at 46.
The 2005 and 2007 reports contained similar seriousness results. See 2005 WSP REPORT, supra note
14, at 32 (“East Indian[s] . . . have the lowest average seriousness score calculated at .09, followed
by Asian drivers computed at .13 and White drivers figured at .17. In contrast, the average serious-

 678

Seattle University Law Review

[Vol. 35:673

categories, Asians and East Indians did not have a greater than average
number of citations per contact or more serious citations than whites, but
blacks, Hispanics, and Native-Americans did. Perhaps, even more ironically, when the studies applied multivariate analysis, including regression analyses that factored in interaction terms for race, citation disparities substantially dissipated for most groups and Autonomous Patrol
Areas (APAs), but not for Asians.25
Breaking out the distinctions between citations, average citations
per contact, and seriousness of violations among different racial groups
is potentially very helpful for those committed to ensuring particularized,
antiracist policing. But the researchers’ ultimate determination that bias
may not explain disproportionate citation for various minority groups—
when the number of violations and violation seriousness is taken into
account—is problematic given that the data often vary for particular racial groups and that alternate explanations for the racially disparate variances within the reports have limited potential.26
With regard to our analysis of the citation data, however, we
acknowledge both a limit and a theoretical commitment that shapes our
perspective and understanding. First, given the vastness of the data collected and the myriad types of dissimilar findings that are experienced
across the various racial groups, it is important to be measured in our
criticism. This is especially true given that a number of our concerns
stem not from the findings of the three reports, but from a search for potentially alternate explanations for racial disparities. Second, we think it
is important to mention that while we do accept the challenge to assess
the research on its own merits, we do so with a perspective that additionness score for Hispanic drivers was .27, for Black drivers it was .29, and for Native American drivers the much higher rate of .43 was calculated.”); see also 2007 WSP REPORT, supra note 14, at 28–
29 (reporting similar results).
25. See Mosher et al., supra note 10, at 51 (discussing 2001–2002 data); 2007 WSP REPORT,
supra note 14, at 30–35; 2005 WSP REPORT, supra note 14, at 38–40.
26. Here we agree with prominent researcher of racial profiling, Professor David A. Harris. In
his assessment of the Lamberth 2002 New Jersey Turnpike study, which determined that speeding
by black drivers led them to be stopped more often than members of other racial groups, Professor
Harris opined:
Since everyone violates some aspect of the incredibly detailed traffic code during even
the shortest drive, police must therefore decide whom they will stop. Police in New Jersey (and everywhere else) must use discretion when they make this decision. The fact
that some groups may be over-represented among people who violate one particular kind
of traffic law hardly means that they are the only ones who could (or do) get stopped.
Speeding is only one possible violation among many. The bottom line is simple. Police
officers have always known that the traffic code is law enforcement’s friend. It allows an
officer to pull over virtually any driver, almost any time, because everyone breaks the
traffic laws.
David A. Harris, The Reality of the Racial Disparity in Criminal Justice: The Significance of Data
Collection, 66 LAW & CONTEMP. PROBS. 71, 95 (2003).

 2012]

Analyzing Stops, Citations, and Searches

679

ally brings a critical race—rather than purely empirical—focus to bear in
analyzing the data and methods.
Initially, one of our concerns relates to how the traffic stops might
have been coded for race and racial identification. The 2005 report data,
however, referenced the use of audits that confirmed the consistency of
the racial identification in the study.27 A second concern about the way
the reports consider race results from one of the explanations for disproportionate citation rates. The researchers essentially rely on claims related to driver behavior across race and ethnicity to explain differences in
citation rates. For example, in the 2007 report, when the average number
of violations and violation seriousness is controlled for, and in the single
recorded violation condition, racial disparities in citations across APAs
substantially disappear for all groups except Asians (for both controls in
5 APAs).28 The report surmises that the higher rate for Asians in a number of APAs is largely explainable as a result of more contacts due to
speeding violations.29 The researchers also refer to data in the 2003 and
2005 reports indicating that certain minority groups wear seat belts less
often30 and have differential rates of compliance with traffic and safety
laws.31
Furthermore, the reports cite data from other states that indicate certain minority groups may have different behaviors with respect to speeding and other forms of aggressive driving.32 Only one of the studies, a
1999 seat belt study for traffic fatalities and alcohol-related incidents,
involved additional data from Washington State.33 Moreover, there were
multiple speeding studies that reached different conclusions with regard
to the existence of race-based speeding differences.34 These types of behavioral data, if more robust, could significantly assist researchers in
more definitively explaining the existence of citation differences across
racial groups.

27. 2005 WSP REPORT, supra note 14, at 4–8.
28. 2007 WSP REPORT, supra note 14, at 33.
29. Id.
30. Mosher et al., supra note 10, at 45; 2003 WSP REPORT, supra note 14, at 22–30, 32.
31. 2005 WSP REPORT, supra note 14, at 34.
32. Id. at 27–29.
33. 2003 WSP REPORT, supra note 14, at 36–37.
34. The Lamberth studies in New Jersey (1996) and Maryland (1997) found no racial difference in speeding. See 2005 WSP REPORT, supra note 14, at 27, 92. The Johnson study of the New
Jersey Turnpike in 2001 found that blacks were more likely to speed in areas where the speed limit
was 65 miles per hour, but not 55 miles per hour. See id. at 10, 92. An ACLU report on profiling
from the 1990s also indicated that, at the time, there was little proof of racially specific driving habits. See HARRIS, supra note 21 (“[T]here is no study or data that supports the view that racial minorities violate traffic laws in any greater number than whites.”).

 680

Seattle University Law Review

[Vol. 35:673

Currently, however, the cited research provides more questions than
answers for the types of citation disparities discovered in Washington
State. In a discussion within the 2003 report, the researchers looked at
data from external studies where the research was seven years old or older, but only one of these studies took place within Washington State.35
For these external studies, it is not clear to us whether the researchers
used multivariate regression analyses or precisely how race was considered. From the critical race perspective, this ambiguity produces at least
two concerns. First, with the exception of the Engel study36 cited in the
2003 report, very few of the studies seem to account for relevant intrarace distinctions. An important question is whether the other cited studies
treat race as a monolith when making claims about driving behavior
without routinely asking about the significance of categories such as
gender and age. This call for intersectional analysis of data is consistent
with a primary concern within the critical race discourse: that law be sensitive to overlapping and reinforcing forms of subordination in operation
for persons who simultaneously inhabit multiple minority identity categories.37
While the Washington reports do account for multiple factors,38
they also do not provide intersectional data broken out specifically for
the citation data in each of the three studied periods and for identity categories such that one could analyze results at the intersection of motorists’
race, gender, and age.39 Such data would be helpful in assessing just how
35. 2003 WSP REPORT, supra note 14, at 118–23.
36. Id. at 118.
37. See, e.g., Kimberlé Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black
Feminist Critique of Antidiscrimination Doctrine, Feminist Theory, and Antiracist Politics, 1989 U.
CHI. LEGAL F. 139; Angela P. Harris, Race and Essentialism in Feminist Legal Theory, 42 STAN. L.
REV. 581 (1990).
38. For example, the reports demonstrate that women as a group received fewer citations, and
younger drivers received a greater number of citations. See Mosher et al., supra note 10, at 49. These
findings are consistent across the reports. 2007 WSP REPORT, supra note 14, at 30; 2005 WSP
REPORT, supra note 14, at 38; 2003 WSP REPORT, supra note 14, at 101.
39. There is, however, some different treatment of multifactor analysis in the reports. The 2005
and 2007 reports include gender and age, as well as race data, but provide no significant intersectional data. See 2005 WSP REPORT, supra note 14, at 36–38; 2007 WSP REPORT, supra note 14, at
29–30. In the separate article reviewing the 2000–2001 data, the researchers also reported on the
effects of race alone on citation rather than race mixed with interaction terms. Mosher et al., supra
note 10, at 50 tbl.3. But when one reviews the 2003 report itself, there are references to statistics at
the intersection of race and gender and race and age for contact data. See 2003 WSP REPORT supra
note 14, at 48–50. For example, for contacts, “Approximately 30% of White drivers and Asian drivers stopped by WSP troopers were females, while approximately 25% of Black drivers and only
18.6% of Hispanic drivers contacted were female in gender. These statewide gender differences in
the proportion of those contacted by WSP officers analyzed by race and ethnicity are fairly consistent across the 40 APAs in Washington state.” Id. at 48. It would have been very helpful to have
such data for the citation stage. This failure to carefully look at the identity intersections produces
criminology studies with confounding results. See also Edward L. Glaeser & Bruce Sacerdote, The

 2012]

Analyzing Stops, Citations, and Searches

681

much race and gender—both separately and together—may figure into
citation decisions during stops. For instance, the reports indicated that
women are stopped and cited less than men.40 As stated earlier, some
racial groups, however, have higher and more serious citation counts.
The next logical question then is whether women of color have citation
counts more consistent with female gender numbers or with male race
numbers.
There is a question about the role racial bias plays in constructing
an understanding of driver behavior. A recent study at the University of
Kansas discusses this question.41 The study focused on a survey of 2500
drivers within the Kansas City metro area and two different types of traffic stops. The researchers distinguished between traffic-safety stops,
which pertain to serious traffic-safety violations, and investigatory stops,
which involve minor violations such as driving a few miles over the
posted speed limit. The researchers found the following:
[For] traffic safety stops, police do not stop African Americans at
higher rates than whites and do not carry out investigatory intrusions of African Americans at higher rates than whites. But in investigatory stops, police stop African American drivers at rates
dramatically higher than whites and more frequently subject African
Americans than whites to such investigatory intrusions as questioning and searches.42

The data suggest that the type of violation may dictate when race is a
consideration. This finding is consistent with the unhelpful Supreme
Court jurisprudence in this area. In the well-known case of Whren v.
United States, the Court determined that where the police had reasonable
cause to believe an actual traffic violation had been committed, it did not

Determinants of Punishment: Deterrence, Incapacitation and Vengeance, HARV. INST. OF ECON.
RES. (2000) (study of accidental vehicular killings where defendants were punished more harshly for
killing women and less harshly for killing blacks, but there was no data on black women victims),
available at http://www.economics.harvard.edu/pub/hier/2000/HIER1894.pdf.
40. See 2007 WSP REPORT, supra note 14, at 30 (“WSP traffic stop data for the period of November 2005 through September 2006 indicated with respect to gender that males were significantly
more likely to be issued a citation in 21 of 34 APAs.”); 2005 WSP REPORT, supra note 14, at 38
(“[F]emales were significantly less likely to be issued a citation in 31 of 40 APAs.”); 2003 WSP
REPORT, supra note 14, at 74 (“Gender also has a strong effect on citation; females are significantly
less likely to be cited in 30 of the 40 APAs.”).
41. Charles R. Epp et al., Inventing the Investigatory Stop, in PULLED OVER: HOW RACE
MATTERS IN POLICE STOPS (manuscript on file with authors) (forthcoming 2013); Donald P. HaiderMarkel et al., Constructing Distrust: The Consequences of African-American Encounters with Police
(discussing the University of Kansas research and how discretionary police stops affect driver distrust of police), available at http://2012sppconference.blogs.rice.edu/files/2012/02/Haider-Markelet-al-SPPC-2012.pdf.
42. Epp et al., supra note 41.

 682

Seattle University Law Review

[Vol. 35:673

matter that the police also had other improper motivations for the stop.43
Such a rule essentially empowers police to profile based on race just as
long as officers can articulate any nominal violation of the law.44
Based on the study from the University of Kansas researchers and
the holding in Whren, one might expect there to be higher numbers of
stops (and potentially citations) for blacks for low-grade speeding offenses and seat belt violations (in jurisdictions where seat belt offenses
are labeled as a safety or investigatory offense). Study results and precedents such as these, then, generally support the contention that police
bias in enforcement for legitimate but very minor traffic violations may
also account for some of what studies capture as racialized driver behavior (e.g., claims that blacks or other persons of color are more serious
offenders or more frequent offenders for certain types of offenses).45
The three Washington State Patrol reports find an increasing number of serious violations for citations issued to certain minorities without
breaking out a typology similar to that done in the University of Kansas
study. Instead, in the 2007 report, for example, there are data for APAs 6
(South Seattle) and 11 (Yakima) that indicated overrepresentation of
blacks when compared to whites in APA 6 (and APA 23).46 In addition,
data suggest disparities for Hispanics in APA 11 when compared to
whites where the benchmarks were related to daylight stops, radar use,
and collision data.47 Data from the 2005 report also indicated disparities
for Hispanics in the same APAs.48
In those APAs, the researchers attempted to discern the importance
of location and the reason for the stop. For blacks, the report found some
difference in both location of the stops and reasons for the stops. The
researchers, however, indicated that the location differences did not indicate systemic targeting and made no reference to the significance that
blacks are less likely to be contacted via radar patrols and are more likely

43. Whren v. United States, 517 U.S. 806 (1996).
44. For analyses and criticisms of the Whren opinion, see David A. Harris, “Driving While
Black” and All Other Traffic Offenses: The Supreme Court and Pretextual Traffic Stops, 87 J. CRIM.
L. & CRIMINOLOGY 544 (1997); Kevin R. Johnson, How Racial Profiling in America Became the
Law of the Land: United States v. Bignoni-Ponce and Whren v. United States and the Need for Truly
Rebellious Lawyering, 98 GEO. L.J. 1005 (2010).
45. For discussions on the importance of using benchmarks in racial profiling data collection,
see Harris, supra note 26, at 84–86; Mosher & Pickerill, supra note 21, at 779−82. Cf. 2005 WSP
REPORT, supra note 14, at 10–11 (discussing differing viewpoints on benchmarks and discussing the
report’s selection of archival, rather than observational, benchmarks).
46. 2007 WSP REPORT, supra note 14, at 17–22.
47. Id. at 20. The 2007 report performed this same detailed analysis for APA 23 (Kelso) and
found that blacks were again more likely to be stopped on I-5, and they were more likely to be contacted for speeding (through combined categories of speeding) compared to whites. Id. at 22–23.
48. 2005 WSP REPORT, supra note 14, at 38–40.

 2012]

Analyzing Stops, Citations, and Searches

683

to be targeted due to lane and HOV violations.49 With regard to violation
difference in general, the report suggests that stop data should be viewed
cautiously because they may contain errors.50 For APA 11, the data indicated that Hispanics were less likely to be cited for radar speeding violations and more likely to be cited for lane violations.51 Measures such as
these indicate a need for a more nuanced study to determine when, in
fact, race-based differences may be evidence of bias. A more particularized study also would require a discussion of when claims about difference should be discussed as a function of officer behavior, in addition to
or instead of driver behavior, for certain groups and certain offenses.
Some of the focus group data gathered by the researchers further demonstrate a need for such analysis.
One of the more disturbing citation results is consistent across the
reports—minorities are more likely to be cited than whites, and the disparity results from differential rates of violating traffic laws. But the researchers’ description of focus group comments from the 2005 report
complicates this understanding. The description of the focus group comments starts with the premise that racial differences in traffic law compliance exist among drivers52 and that officers are aware of these differences.53 To the extent differential offending may be partially informed by
license and insurance offenses, which implicate socioeconomic class,
what is reflected as a race-based violation potentially arises as a function
of minorities being overrepresented among the poor.54 While this point
about class is helpful as an explanation for the connection between driver
identity and violation behavior, claims that ostensibly implicate links
between traffic violation and identity-based stereotypes are more problematic. For example, consider the following excerpt:
Troopers and sergeants taking part in the focus groups also offered
useful suggestions with respect to explaining the higher rate of citation documented for Asian drivers. Some officers volunteered the
comment that Asian drivers were not as “accomplished” in their
driving skills as others and they were disproportionately responsible
for accidents. As one WSP sergeant noted, “I’d take ten white drivers, ten black drivers, ten Asian, ten Hispanic drivers . . . I would

49. 2007 WSP REPORT, supra note 14, at 19.
50. 2005 WSP REPORT, supra note 14, at 6–7.
51. Id. at 34–35 (presenting quantitative data of differential rates of citation among minority
groups for various offenses related to driving under the influence, seat belt violations, failure to
maintain insurance, and license offenses).
52. Id. at 35.
53. Id.
54. Id. at 34–36.

 684

Seattle University Law Review

[Vol. 35:673

say the highest percentage of lower driving skills lies within the
Asian.”55

The comment is disturbing because it taps into what many will regard as
a stereotype—a “false or misleading association[] between a group and
an attribute that [is] held by their subjects in a rigid manner, resistant to
counterevidence.”56—with regard to Asians and poor driving.57 Moreover, without referencing the problematic nature of the statement, the report refers to the comment as “useful.”58 Another focus group participant
indicated that ‘“[t]here’s a large population of Asians in our district . . . . Unfortunately, they’re not always the best drivers . . . they
cause a huge amount of the accidents.’”59 Again, rather than ask about
the potential shaping function on the behavior of police officers who
subscribe to such beliefs, the report instead merely confirms that Asians
had “the largest disparity between self-initiated and accident contacts for
any of the five minority groups.”60 There were also comments claiming
that Asian drivers were more argumentative when stopped and that there
were language barriers.61
Comments of this nature were not made about only Asians. In addressing the driving habits of Native-Americans and Hispanics, officers
opined,
Every fifth person that we stop that is Native American will either
be suspended, have a warrant, or be drunk . . . if we stop a white
person, it would probably be every twentieth car we contact, at
night-time, that would be suspended, have a warrant, or be drunk.62

Another trooper commented as follows: “I remember when I first got
here, a lot of the vehicles, especially in the Hispanic community, either
55. Id. at 43–44.
56. Lawrence Blum, Stereotypes and Stereotyping: A Moral Analysis, 33 PHIL. PAPERS 251,
288 (2004).
57. This notion of stereotype implicates the understanding of a false belief about a group. See
Kwame A. Appiah, Stereotypes and the Shaping of Identity, in PREJUDICIAL APPEARANCES: THE
LOGIC OF ANTIDISCRIMINATION LAW 64 (Robert Post ed., 2001). Another definition of stereotype
raises the issue of “false positives” between some ascribed group behavior and a nonconforming
group member. Professor Appiah has also described this belief about stereotyping as follows:
[T]he idea of ascribing to an individual a property in the belief that it is a characteristic of
some social group to which she belongs, where there is indeed a statistical correlation between that property and being a member of the group, but where, in fact, she does not
have that property.
Id. at 63.
58. 2005 WSP REPORT, supra note 14, at 43.
59. Id. at 44.
60. Id.
61. Id.
62. Id. at 35.

 2012]

Analyzing Stops, Citations, and Searches

685

they were not licensed or driving on a suspended license. I mean, a majority of them.”63
One might argue that there is minimal significance of the focus
group comments given that data reflect differential citation among members of various racial groups, and there are no significant racialized disparities for contacts. Moreover, though these comments reflect prevalent
stereotypes, just because one references a stereotype does not mean that
one will act on that prejudice.64 The problem is that the report presupposes that such comments are proof of racial differences in traffic violations.
The comments could, however, be regarded as reflective of attitudes that
inform the pre- and post-stop expectations of officers.65 Even a finding
that there is no significant evidence of profiling in contacts does not absolutely lead to a conclusion that there is no bias in the number of citations.
By engaging in oversimplified race-based stereotyping, the Washington State Patrol undermines faith in the assertion that the racially disparate citation data are not, in some way, evidence of bias. The researchers’ somewhat uncritical assessment of the meaning of the focus group
commentary also raises a concern about whether sufficient attempts have
been made to investigate explanations for citation differences not tied to
stereotypes. It is not absurd to question whether the researchers would at
least consider the effect of race bias by patrols. The researchers do, at
times, make other such normative-leaning claims that are not supported
by robust data. For example:
Although we do not have strong data to support this speculation, it
is possible that the high citation rate for Asian drivers is related to
the fact that younger Asians are driving at higher speeds than
Whites, and thus are more susceptible to citation. This high citation
rate for Asians may also be related to the WSP attempting to deter
the “street racing” phenomenon, which has become a problem at the

63. Id.
64. See Blum, supra note 57, at 254 n.3 (“Someone might regard . . . Asians as bad drivers, yet
not feel negatively toward those groups. Moreover, even if a stereotype is negatively evaluatively
charged, for a particular carrier of that stereotype, this charge need not always trigger the corresponding negative affect.”); see also Appiah, supra note 57, at 67 (“The problem is not stereotyping
but bigotry—that is unjustified hatred or contempt . . . . Bigotry often leads to stereotyping and stereotyping can lead to bigotry. But each can stand alone.”). Professor Appiah has, however, made a
stronger claim in this area that “there is no reason to suppose that normative stereotypes as such
must be wrong, or that public actions grounded on them are to be criticized, even where they involve
differences in treatment that are judged to be invidious.” Appiah, supra note 57, at 65.
65. Cf. Robert J. Smith & Justin D. Levinson, The Impact of Implicit Racial Bias on the Exercise of Prosecutorial Discretion, 35 SEATTLE U. L. REV. 795, 797 (2012) (arguing that “implicit
racial attitudes and stereotypes skew prosecutorial decisions in a range of racially biased ways”).

 686

Seattle University Law Review

[Vol. 35:673

national level and in the Pacific Northwest . . . and which has been
associated with young Asian males.66

Focus group data of the kind collected in the 2005 report produce at
least two other issues. First, given the persistence of the differential citing of Asian drivers and comments implicating Asian driver stereotypes
among the Washington State Patrol, one normative question that arises is
whether Asians in Washington State, like blacks in some other jurisdictions, suffer the consequences of being the most “otherized” minority
population.67 This question seems especially relevant where the 2005
report says, “[I]t is important to reiterate that when racial differences in
compliance with traffic and safety legislation are statistically controlled
for, there is not a single autonomous patrol area in which Blacks, Native
Americans, Hispanics, or East Indians are significantly more likely to be
issued citations.”68 Second, one wonders how such a finding would inform data analysis of the current statistics and research design moving
forward. Much like the researchers, one could look at the citation data
and focus group information and decide that, in the absence of greater
proof of bias, the most likely explanation for disparate citation practices
is driver behavior. In the alternative, where one brings a critical skepticism to racially disparate results even where proof of animus is ambiguous,69 one could decide that at this point it would be premature to eliminate bias as being at least partially responsible for the consistent citation
disparities. Further, one embracing this latter view of the data would also
be a bit more hesitant to declare that the findings across the reports produce little evidence of systemic problems. At bottom, then, the citation
results are opaque enough to support multiple views on the operation of
bias.
B. Seeking a Deeper Understanding of Methodological Choice
Our final concern, which is related to the strength of claims discounting racial bias as an explanation for higher citation rates for certain
minorities, pertains to the methodology used by the researchers. First,
related to all three data sets, researchers sought to explain the existence
of bias by reviewing results for citations with only one recorded viola66. 2007 WSP REPORT, supra note 14, at 74.
67. Lenese C. Herbert, O.P.P.: How “Occupy’s” Race-Based Privilege May Improve Fourth
Amendment Jurisprudence For All, 35 SEATTLE U. L. REV. 727, 738 (2012) (noting that “the disfavored status as nonwhites erroneously relegates [blacks] to the low status of The Other, the outsider”).
68. 2003 WSP REPORT, supra note 14, at 45 (emphasis removed).
69. See Smith & Levinson, supra note 65, at 801 (arguing that situational cues, rather than
animus, often trigger racial stereotypes resulting in unintentional biased decision-making).

 2012]

Analyzing Stops, Citations, and Searches

687

tion. For example, the 2003 report found the following results when there
was only one recorded violation condition:
Th[e] table reveals that Black drivers who have committed only a
single violation are significantly less likely to be cited in two
APAs . . . and are significantly more likely to be cited in three
APAs . . . . Native-American drivers who committed only a single
violation are significantly less likely to receive a citation in four
APAs . . . and are significantly more likely to be cited in four
APAs . . . . Hispanics who have committed a single violation are
significantly more likely to be cited in seven APAs . . . . Asians who
have committed a single violation are more likely to be cited in 14
APAs.70

In the 2007 report, in the one recorded violation condition, the researchers also found that blacks, Hispanics, and Native-Americans were not
more likely to be cited in any of the thirty-four APAs, East Indians were
significantly more likely to be cited in two APAs, and Asians were significantly more likely to be cited in five APAs.71
We question why the researchers find the one recorded violation
condition so helpful in a data set where the most concerning racially disparate result has to do with multiple citations and citation seriousness. In
other words, how is it that this data point should be considered in light of
the other data of racial disparity? Is the one recorded violation condition
helpful to explain what is going on in the more complex contacts? It
seems to be an overstatement to make the claim that reduced racial disparities in the one recorded violation condition is supported by the claim
that “they [the data] do not indicate the operation of systemic bias in citing minorities who have only a single violation recorded by the WSP.”72
We argue that one question in particular requires more consideration: at
what threshold, and in what violation conditions, should one be comfortable speaking to the existence of structural or systemic forms of bias?
Finally, the researchers make a second claim with regard to potential bias that is problematic when analyzing the citation “not issued”
condition. They assert that there is no bias or “piling on” evident in this
condition because minorities who receive no citations do not have higher
average violations and do not have violation seriousness numbers that are
much lower than the averages for these measures in the noncitation condition for whites. Essentially, the researchers are constructing bias in the
noncitation condition as evident only where minorities would need to be
far more “innocent” than whites in order not to be cited. In the 2003 re70. 2003 WSP REPORT, supra note 14, at 84–85.
71. 2007 WSP REPORT, supra note 14, at 37.
72. Id.

 688

Seattle University Law Review

[Vol. 35:673

port, for each minority group in all the APAs, violation numbers for the
noncited were similar to or higher than the numbers of noncited whites.73
This claim leads to two questions. First, why would disadvantageous treatment in violations be assessed as a function of the noncitation
condition, especially where troubling issues seem to be in the cited condition? Second, are there other ways we could think about analyzing race
differentials through a different lens other than “bolstering” or piling on
violations per citation? For example, the report has already noted the potential socioeconomic explanation for the racialized violation disparity.
Another explanation, at least for certain drivers, could pertain to police
behavior once they have made a decision to stop a car or issue a citation.
These questions are meant neither to be exhaustive nor to suggest fundamental flaws with the citation data and methodology. We merely seek
to fully understand the connection between the non-, one-, and multiplecitation conditions, and to explore whether there are any other measures
or conditions that would help us further assess the racial disparities observed in the broadest portions of the citation data.

III. WASHINGTON STATE PATROL SEARCHES
Although the Washington State Patrol data divide searches into
seven categories, the WSU researchers chose to group them into three
general categories: No Search; Low-Discretion Search (Search Incident
to Arrest, Impound/Inventory Search, Warrant Search); and HighDiscretion Search (Consent Search, K9 Search, Terry (Pat Down)
Search).74 This strikes us as a generally useful way to differentiate
searches because of the difficulty of tracking trends or patterns within the
seven search categories; we are curious, however, about what the more
differentiated data would show, especially with regard to the different
kinds of high-discretion searches. We were troubled by the way the 2003
report discusses the lack of data regarding the frequency of trooper requests to conduct a search that is refused by the driver in the absence of
probable cause. Without any data, the following hypothetical is posed:
If, hypothetically, White drivers and Asian drivers who refuse to
consent searches more often than Hispanics, Blacks and Native
Americans, the disparities in discretionary searches would be diminished and we would conclude that troopers do not use their own dis-

73. 2003 WSP REPORT, supra note 14, at 84–86.
74. 2007 WSP REPORT, supra note 14, at 42; 2005 WSP REPORT, supra note 14, at 50; 2003
WSP REPORT, supra note 14, at 92 (This report characterized the Low-Discretion Search category as
Nondiscretionary Searches).

 2012]

Analyzing Stops, Citations, and Searches

689

cretion to target any particular racial or ethnic groups. This is an
empirical question for which we presently have no data, however.75

One recent study found no difference between races or levels of income
with regard to whether a person felt free to terminate their encounters
with the police.76 The subsequent 2005 report stated that “troopers consistently told us that very few people actually refuse when asked to consent to a search.”77 Therefore, for the researchers to suggest, without data, that they lack information that might reveal the problem does not really exist is unsupported and misleading.78
During the 2003, 2005, and 2007 reports, searches occurred in
3.5%, 3.5%, and 3.3% of all Washington State Patrol contacts with drivers in the respective periods.79 Of these searches, the following table
breaks down the frequency by type of search—low discretion or high
discretion (using the typology in the reports).
Table 1. Type of search80

Low discretion
High discretion

Mar. 2002–
Oct. 2002
77%
23%

July 2003–
June 2004
87%
13%

Nov. 2005–
Sept. 2006
88%
12%

One immediate observation is that, as a proportion of total searches,
high-discretion searches dropped from 23% during the first observation
period to 13% and 12% in the later two periods. Given some of the problems (e.g., improperly coded searches, including the search field left
blank) noted by the researchers in the 2003 and 2005 reports,81 it is not

75. 2003 WSP REPORT, supra note 14, at 95.
76. David K. Kessler, Free to Leave? An Empirical Look at the Fourth Amendment’s Seizure
Standard, 99 J. CRIM. L. & CRIMINOLOGY 51, 53 (2009).
77. 2005 WSP REPORT, supra note 14, at 68.
78. Id.; 2003 WSP REPORT, supra note 14, at 95.
79. 2003 WSP REPORT, supra note 14, at 93 tbl.S–1; 2005 WSP REPORT, supra note 14, at 51
tbl.S–1; 2007 WSP REPORT, supra note 14, at 44 tbl.18.
80. Because the data come from different length time periods, we include in the table the frequency of search expressed as a percentage of total observations. For the period of March 2002
through October 2002, the total observations were 677,514; low-discretion searches were 18,062;
and high-discretion searches were 5331. 2003 WSP REPORT, supra note 14, at 93 tbl.S–1. From June
2003 through June 2004, total observations were 1,103,121; low-discretion searches were 32,552;
and high-discretion searches were 4965. 2005 WSP REPORT, supra note 14, at 51 tbl.S–1.
81. See 2005 WSP REPORT, supra note 14, at 47–48 (discussing data in the 2003 and 2005
reports).

 690

Seattle University Law Review

[Vol. 35:673

apparent that anything conclusive can be stated with regard to why the
proportion of high-discretion searches has dropped.
Table 2. Frequency of search by race, expressed as a percentage82

White
Black
NativeAmerican
Hispanic
Asian/Pacific
East Indian

Mar. 2002–
Oct. 2002
Low83 High
2.5
.4
6.6
1.0

July 2003–
June 2004
Low
High
2.6
.4
5.6
1.0

Nov. 2005–
Sept. 2006
Low
High
2.6
.3
5.3
.7

12.9

2.1

11.8

1.6

12.7

1.4

5.7

1.0

.9

.3

5.6
2.3
1.0

.9
.4
.2

5.4
2.3
.7

.9
.3
.1

Table 3. Disproportionality ratio in comparison to white drivers84

White
Black
NativeAmerican
Hispanic
Asian/Pacific
East Indian

Mar. 2002–
Oct. 2002
Low
High
1.0
1.0
2.6
2.5

July 2003–
June 2004
Low
High
1.0
1.0
2.2
2.5

Nov. 2005–
Sept. 2006
Low
High
1.0
1.0
2.0
2.3

5.2

5.3

4.5

4.0

4.9

4.7

2.3

2.5

.4

.8

2.2
.9
.4

2.3
1
.5

2.1
.9
.3

3
1
.3

For low-discretion searches during the three respective time periods, blacks were 2.6, 2.2, and 2.0 times more likely to be subject to a
low-discretion search than whites; Native-Americans were 5.2, 4.5, and
4.9 times more likely to be subject to a low-discretion search than
whites; Hispanics were 2.3, 2.2, and 2.1 times more likely to be subject
82. 2007 WSP REPORT, supra note 14, at 44 tbl.19; 2005 WSP REPORT, supra note 14, at 52
tbl.S–2; 2003 WSP REPORT, supra note 14, at 95 tbl.S–3.
83. The later 2005 and 2007 reports substituted “low discretion” for “nondiscretionary searches,” which was used in the 2003 report. See 2005 WSP REPORT, supra note 14, at 50; 2003 WSP
REPORT, supra note 14, at 92.
84. 2003 WSP REPORT, supra note 14, at 95 tbl.S–3; 2005 WSP REPORT, supra note 14, at 52
tbl.S–2; 2007 WSP REPORT, supra note 14, at 44 tbl.19.

 2012]

Analyzing Stops, Citations, and Searches

691

to a low-discretion search than whites; Asian/Pacific Islanders were
searched about the same rate as whites; and East Indians were searched
at lower rates than whites.85
For high-discretion searches during the three respective time periods, blacks were 2.5, 2.5, and 2.3 times more likely to be subject to a
high-discretion search than whites; Native-Americans were 5.3, 4.0, and
4.7 times more likely to be subject to a high-discretion search than
whites; Hispanics were 2.5, 2.3, and 3 times more likely to be subject to
a high-discretion search than whites; Asian/Pacific Islanders were
searched about the same rate as whites; and East Indians were searched
at lower rates than whites.86
The 2003 report stated:
It is noteworthy, however, that the influence of race would not seem
to depend on whether a trooper conducts a nondiscretionary or discretionary search. The coefficients for Black, Hispanic and Native
American drivers remain positive and at about the same magnitude
for both categories of searches. This finding of no difference in
rate of search of minority drivers between discretionary and
nondiscretionary searches suggests that where WSP officers
have the most discretion in choosing to conduct a search, they
do not act any differently toward different racial groups than
when they act with no (or with little) discretion. This finding in
turn suggests that while there appear to be systematic disparities in
the probability that these three minority groups will be searched
compared to Whites and other racial groups, those disparities do not
seem to be the result of the intentional bias of troopers.87

The 2005 and 2007 reports repeated the assertion that because the magnitude of the disproportionality is roughly the same for low- and highdiscretion searches, it reflected consistent trooper behavior toward different racial groups regardless of search, which is inconsistent with a
finding of bias.88
This comparison between the conclusions drawn based on nondiscretionary or low discretion and high discretion is mistaken. The very
reason for differentiating between the two is that, for low-discretion
searches, presumably very little discretion is allowed because there is a
triggering event (e.g., arrest, impound/inventory, or warrant) that leads to
the search. The events or circumstances that trigger the low-discretion
85. 2003 WSP REPORT, supra note 14, at 95 tbl.S–3; 2005 WSP REPORT, supra note 14, at 52
tbl.S–2; 2007 WSP REPORT, supra note 14, at 44 tbl.19.
86. 2003 WSP REPORT, supra note 14, at 95 tbl.S–3; 2005 WSP REPORT, supra note 14, at 52
tbl.S–2; 2007 WSP REPORT, supra note 14, at 44 tbl.19.
87. 2003 WSP REPORT, supra note 14, at 101.
88. 2007 WSP REPORT, supra note 14, at 49–51; 2005 WSP REPORT, supra note 14, at 52.

 692

Seattle University Law Review

[Vol. 35:673

searches presumably are not connected to race. If someone is stopped for
a DUI and arrested, and there is no suitable unimpaired driver, a search is
supposed to occur. In contrast, high-discretion searches occur in the absence of these specific triggering events and require further investigation.89
The 2005 and 2007 reports included a more detailed analysis that
tried to control for more factors in order to see if the racial disproportionalities can be accounted for by nonracial factors.90 The 2007 report
predicted search rates for eighteen- and fifty-year-old drivers for “stops
involving a white male police officer, in the daytime, on an interstate,
with one (non-serious) violation.”91
Table 4. Predicted search rate for eighteen-year-old male driver,
non-Latino white male officer, daytime, interstate, one nonserious violation.92

Native-American
Latino
Black
Non-Latino White
Asian/PI
East Indian

Low-discretion search
3.1%
1.1%
1.2%
.8%
.6%
.1%

High-discretion search
1.6%
.7%
.6%
.4%
.25%
.1%

For high-discretion searches, in comparison to an eighteen-year-old
white male driver, the eighteen-year-old Native-American male driver is
four times more likely to be searched; the eighteen-year-old black male
driver is 1.5 times more likely to be searched; the eighteen-year-old Latino male driver is 1.75 times more likely to be searched; and AsianAmericans, Pacific Islanders, and East Indians are less likely to be
searched than whites.93
When age, seriousness of the violation, race of the officer, time of
day, and location of the stop are factored in, race remains an unaccounted
for factor that correlates with a heightened probability of search. The fact
that the magnitude of search probability differences is roughly equal for
low- and high-discretion searches does not support the conclusion “that
89. We also think further work needs to be done with regard to the operation of discretion even
with what are characterized as low-discretion searches.
90. 2007 WSP REPORT, supra note 14, at 46–51; 2005 WSP REPORT, supra note 14, at 56–64.
91. 2007 WSP REPORT, supra note 14, at 48.
92. Id.
93. Id.

 2012]

Analyzing Stops, Citations, and Searches

693

officers do not act differently based on race when they have higher levels
of discretion.”94

IV. CONCLUSION
We agree with the WSU researchers that the evidence does not support a finding of widespread intentional discrimination on the part of the
Washington State Patrol. But the persistence of the disproportionality by
race, especially for high-discretion searches, is consistent with the operation of implicit bias. This heightened probability of a search cannot be
justified by “hit rates” when the likelihood of finding contraband is either
the same or lower for racial minorities than it is for whites.95
We commend the Washington State Patrol for not “engaging in disproportionate enforcement at the level of original driver contact” that are
found in many other jurisdictions.96 We also agree that disproportionality
by itself does not constitute evidence of purposeful or intentional discrimination. The unfortunate state of our antidiscrimination laws is that
much discrimination occurs for which there is no legal remedy.97 But
rather than rely on law’s failure to provide a remedy, the Washington
State Patrol should explore the ways that unconscious or implicit bias
can operate to produce the observed differences and institute policies and
trainings to mitigate this bias.98
It is not enough to tell the eighteen-year-old Native-American male
driver, stopped by a white male officer for a nonserious violation on the
interstate during the day, that we don’t know why he is four times more
likely to be searched than a similarly situated eighteen-year-old white
male driver.
We can do better. We must do better.

94. Id. at 49.
95. We find it odd that the 2007 Washington State Patrol Report emphasizes that the hit rate
for low-discretion searches is exactly the same for non-Hispanic whites, African-Americans, and
Native-Americans. Id. at 50. We think the more appropriate reference points are the hit rates for
high-discretion searches, where the hit rates are lower for African- Americans, Hispanics, and Native-Americans than for whites. Id. at 46.
96. 2003 WSP REPORT, supra note 14, at 47.
97. See Alan D. Freeman, Legitimizing Discrimination through Antidiscrimination Law: A
Critical Review of Supreme Court Doctrine, 62 MINN. L. REV. 1049 (1978).
98. Cf. Andrea D. Lyon, Race Bias and the Importance of Consciousness for Criminal Defense
Attorneys, 35 SEATTLE U. L. REV. 755, 762 (2012) (contending that criminal defense attorneys
should explore the ways unconscious and implicit bias may affect the best interests of their clients).