Racial Disparity in Consent Searches and Dog Sniff Searches

An analysis of Illinois traffic stop data from 2013

 Introduction

Throughout Illinois, there is a dramatic racial disparity in police use of so-called “consent searches” and dog sniff searches during routine traffic stops. This is shown by a decade of government data about consent searches collected and published under the Illinois Traffic Stop Statistical Study Act of 2003 (“Study Act”), and the first two years of such dat

a regarding dog sniff searches.

For example, the recently published 2013 data shows that Illinois State Police (“ISP”) troopers are 2½ times more likely to consent search Hispanic motorists compared to white motorists, but white motorists are 2½ times more likely than Hispanic motorists to be found with contraband during such ISP searches. Likewise, ISP troopers are more than twice as likely to dog sniff Hispanic motorists compared to white motorists, yet white motorists are 64% more likely than Hispanic motorists to be found with contraband during a trooper’s search based on a dog alert. These disparities are a clear sign that the ISP’s threshold to conduct consent searches and dog sniff searches is far lower for Hispanic motorists than for white motorists. Such racial disparate impact against black and Hispanic motorists likewise exists statewide, and in stops by the Chicago Police Department and other local and county police agencies.[1]

This ACLU of Illinois report begins with a brief discussion of the Study Act, upon which this report is based. It then presents findings of substantial and ongoing racial disparate impact in the use of consent searches and dog sniff searches during routine traffic stops.

The Study Act

The Study Act requires all police officers in Illinois to document all of their traffic stops, including motorist race and what happened. It also requires all police agencies in Illinois to report their stops data to the Illinois Department of Transportation (“IDOT”).  It then requires IDOT to publish an annual data report. See 625 ILCS 5/11-212. IDOT has published ten years of Study Act data on its website.[2]

The Study Act has twice been expanded to capture new kinds of traffic stop data. In 2006, in response to Study Act data regarding racial disparity in consent searches, it was expanded to require disclosure of whether a consent search yielded contraband, and whether a motorist declined consent to search. See Public Act 94-997. In 2011, in response to Study Act data regarding racial disparity in canine sniffs, it was expanded to document whether a dog sniff occurred, whether a dog alerted, whether a dog alert caused a search by an officer, and whether contraband was discovered.  See Public Act 97-0469.[3]

This report begins with the data stated in the Study Act, and makes simple calculations as shown in the attached exhibits.

Findings and discussion

I. Consent searches

A. The inherent hazards of consent searches

A consent search occurs when a police officer asks for a motorists’ permission to search an automobile during a routine traffic stop, whether or not there are any lawful grounds for the search, simply based on an officer’s “hunch” that the motorist is carrying contraband. Consent searches inherently raise three serious civil rights and civil liberties concerns.

First, in many cases, the motorist’s supposed “consent” to search is not truly voluntary. Consent is often granted on an isolated roadside in a one-on-one encounter with an armed law enforcement official. This setting is inherently coercive. Many civilians believe they must grant consent. Other civilians fear the consequences of refusing to grant consent, such as the issuance of extra traffic citations, or the delay caused by further interrogation or bringing a drug-sniffing dog to the scene. Thus, the overwhelming majority of motorists consent to a search when asked.

Second, once consent is granted, the result is an intrusive and publicly humiliating search of one’s car and/or person. See Terry v. Ohio, 392 U.S. 1, 24-25 (1968) (describing a pat-down frisk of one’s body as a “severe” intrusion, and as “annoying, frightening, and perhaps humiliating”); Florida v. J.L., 529 U.S. 266, 272 (2000) (describing such frisks as “intrusive” and “embarrassing”).

Third, because the decision whether to request consent to search is typically based on the subjective “hunch” of individual police officers, consent searches are inherently susceptible to bias, conscious or otherwise. From a management perspective, consent searches are particularly troublesome. Since they are subjective, they are not subject to meaningful supervisory review.            

B. Statewide consent searches

Statewide in 2013, black and Hispanic motorists were nearly twice as likely as white motorists to have their vehicles consent searched during traffic stops. Specifically, black motorists were 95% more likely, and Hispanic motorists were 89% more likely. There are similar disparities every year such data has been collected starting in 2004. See Exh. 3 (analysis of statewide consent search rates).

On the other hand, when police in Illinois performed a consent search in 2013, white motorists were far more likely than minority motorists to be found with contraband. Specifically, white motorists were 49% more likely than black motorists, and 56% more likely than Hispanic motorists. Again, there are similar disparities every year such data has been collected starting in 2004. See Exh. 4 (analysis of statewide consent search hit rates). Such “hit rate” disparities show that “a lower standard of proof was applied to searches of minorities than to searches of Caucasians.” See Police Executive Research Forum, By the numbers: A guide for analyzing race data from vehicle stops (2004) at p. 274 (“PERF manual”).[4]

C. ISP consent searches

In 2013, the ISP conducted 2,124 consent searches. This was more than any other police agency in Illinois, and 9% of all consent searches performed in the state. See Exh. 5 (analysis of ISP consent search rates); Exh. 3 (analysis of statewide consent search rates); Exh. 1 (IDOT report of 2013 Study Act data) at p. 11.

Hispanic motorists were 2.49 times more likely than white motorists to have their vehicles consent searched during traffic stops by ISP troopers. But white motorists were 2.48 times more likely than Hispanic motorists to be found with contraband during such searches. There are similar disparities every year such data has been collected starting in 2004. See Exh. 5 (analysis of ISP consent search rates); Exh. 6 (analysis of ISP consent search hit rates).

D. CPD consents searches

In 2013, the CPD conducted 1,390 consent searches. Other than the ISP, this was more than any other police agency in Illinois. See Exh. 7 (analysis of CPD consent search rates); Exh. 1 (IDOT report of 2013 Study Act data) at p. 11.

Black and Hispanic motorists were more than four times more likely than white motorists to have their vehicles consent searched during traffic stops by the CPD in 2013. Specifically, black motorists were 4.74 times more likely, and Hispanic motorists were 4.09 times more likely. There are similar disparities every year such data has been collected starting in 2004. See Exh. 7 (analysis of CPD consent search rates).

On the other hand, when CPD officers performed a consent search in 2013, white motorists were far more likely than minority motorists to be found with contraband. Specifically, white motorists were 2.1 times more likely than black motorists, and 1.86 times more likely than Hispanic motorists. Again, there are similar disparities every year such data has been collected starting in 2004. See Exh. 8 (analysis of CPD consent search hit rates).

E. Consent searches by other police agencies

Eleven of the nearly 1,000 police agencies in Illinois conducted more than one-third of all the state’s consent searches in 2013. After the ISP and the CPD, the following nine agencies conducted the most consent searches: Aurora (1,157), Springfield (896), McHenry County (596), Lake County (419), Grundy County (410), Rockford (408), Jacksonville (388), Posen (383), and Belleville (358). Exh. 1 (IDOT report of 2013 Study Act data) at p. 11. Racial disparity resulted from the consent searches by many of these police agencies in Illinois. For example:

Aurora.  Black and Hispanic motorists were far more likely than white motorists to be consent searched (2.36 and 1.63 times more often), and white motorists were far more likely than black and Hispanic motorists to be found with contraband (30% and 46% more). See Exh. 9 (analysis of Aurora consent search rates); Exh. 10 (analysis of Aurora consent search hit rates). Aurora is the second-most populous city in Illinois, with 200,000 people.

Springfield.  Black motorists were 3.4 times more likely than white motorists to be consent searched, and white motorists were 89% more likely than black motorists to be found with contraband. See Exh. 11 (analysis of Springfield consent search rates); Exh. 12 (analysis of Springfield consent search hit rates). Springfield is the sixth-most populous city in Illinois, with 117,000 people.

Lake County.  Hispanic motorists were twice as likely as white motorists to be consent searched, and white motorists were 40% more likely than Hispanic motorists to be found with contraband. See Exh. 13 (analysis of Lake County consent search rates); Exh. 14 (analysis of Lake County consent search hit rates). Lake County is the third-most populous county in Illinois, with 700,000 people.

Rockford.  Black motorists were 69% more likely than white motorists to be consent searched, and white motorists were 67% more likely than black motorists to be found with contraband. See Exh. 15 (analysis of Rockford consent search rates); Exh. 16 (analysis of Rockford consent search hit rates). Rockford is the third-most populous city in Illinois, with 150,000 people.

F. The solution: ban consent searches

Consent searches should be prohibited. The General Assembly should enact legislation banning the practice statewide. Policymakers for individual police agencies should prohibit the practice, too. Consent searches are coercive and invasive of the privacy of motorists of all races. Moreover, consent searches will always cause racial disparity, given the very high level of officer subjectivity in deciding when to request consent, and the very high frequency that civilians grant consent.

II. Dog sniff searches

A. The inherent hazards of dog sniff searches

A dog sniff search occurs when a police officer directs a police dog to sniff a vehicle for drugs during a routine traffic stops, simply based on an officer’s “hunch” that the motorist is carrying drugs. A dog alert ordinarily provides a lawful basis for police officers to manually search the interior of the vehicle. Dog sniff searches inherently raise three serious civil rights and civil liberties concerns.

First, dog sniffs are menacing, especially for minority motorists, in light of historical abuses committed with police dogs. Dog sniffs also are humiliating, taking place in full view of passing motorists, friends and strangers alike, many of whom probably conclude that the people subjected to dog sniffs must be guilty of something. Full car searches based on false dog alerts are even more frightening and embarrassing.

Second, dog sniffs are all too frequently unreliable. As shown below, dog alerts are wrong as often as they are right. Thousands of innocent motorists are subjected to lengthy, invasive, and humiliating vehicle searches by police officers as a result of erroneous dog alerts.

Third, there is a substantial racial disparity in erroneous dog alerts, as shown below. Perhaps this is the “Clever Hans Effect,” named for the horse widely believed to be capable of arithmetic, but actually capable only of reading the unintentional emotional reactions of his human audiences while he slowly stamped his hoof. See “Clever hounds,” The Economist (Feb. 15, 2011).[5]

B. Statewide dog sniff searches

Statewide in 2013, no contraband was found during 40% of the officer searches performed in response to a dog alert. Stated differently, false alerts by police dogs caused police officers to manually search the vehicles of 1,715 innocent motorists. See Exh. 17 (analysis of statewide dog sniff hit rates).

There was also a statewide racial disparity in 2013. Black motorists were 55% more likely than white motorists to be subjected to a dog sniff. Yet white motorists were 14% more likely than black motorists to be found with contraband during officer searches performed in response to a dog alert. See Exh. 18 (analysis of statewide dog sniff rates); Exh. 17 (analysis of statewide dog sniff hit rates).

C. ISP dog sniff searches

In 2013, no contraband was found during 50% of the trooper searches performed in response to the alert of an ISP dog. Thus, 381 innocent motorists were subjected to a full search of their cars by ISP troopers because ISP dogs were wrong as often as they were right. See Exh. 19 (analysis of ISP dog sniff hit rates).

Even worse is the racial disparity. Hispanic motorists were 2.16 times more likely than white motorists to be subjected to a dog sniff search by ISP troopers in 2013. Yet white motorists were 64% more likely than Hispanic motorists to be found with contraband during a trooper search prompted by the alert of an ISP dog. Stated differently, while white motorists had contraband more than one-half of the time (53%), Hispanic motorists had contraband only one-third of the time (33%). This ISP racial disparity was similar in 2012. See Exh. 20 (analysis of ISP dog sniff rates); Exh. 19 (analysis of ISP dog sniff hit rates).

D. The solution: limit and regulate dog sniff searches

Given the foregoing problems, the ACLU of Illinois recommends three changes regarding dog sniff searches for drugs during routine traffic stops.

First, police officers should be required to have reasonable suspicion of criminal activity before subjecting motorists to a dog sniff for drugs. While the Illinois Supreme Court held that this standard was required by the U.S. Constitution, the U.S. Supreme Court subsequently reversed and held it was not, and then the Illinois Supreme Court held it was also not required by the Illinois Constitution.[6] The dissents of U.S. Supreme Court Justices Souter and Ginsburg emphasized the fallibility of drug sniffing dogs, and the fact that that dog sniffs make a traffic stop lengthier and more embarrassing, intimidating, and adversarial.[7] In 2005, legislation to mandate this standard (H.B. 1557) was sponsored by most members of the Illinois House Black Caucus. Without this more objective standard, too many police officers use hunches to decide which cars to sniff, and those hunches inevitably cause racial disparate impact.

Second, police agencies should annually use the Study Act data to identify dogs with unusually high false positives. Such dogs might need further training, or retirement. Information about this dog-by-dog review process should be monitored by a statewide regulatory body, and should be made publicly available.

Third, training of dogs and handlers should be improved. In response to a 2011 study showing racial disparity in dog sniff searches,[8] new Illinois legislation mandated that all state and local police dogs must be trained by programs that meet the minimum certification requirements set by the Illinois Law Enforcement Training Standards Board. See Public Act 97-469, codified at 50 ILCS 705/10.12. The troubling 2013 data – including both the high rate of false positives and the substantial racial disparate impact – show that more needs to be done to improve this training.

[1] While this report focuses on consent searches and dog sniff searches, many police agencies in Illinois also have racial disparities in the initial decisions to commence traffic stops. Statewide in 2013, minority motorists were stopped 22% more frequently than expected based on the minority driver benchmarks established by the statistical consultant retained by the Illinois Department of Transportation (“IDOT”). Nearly 120 Illinois police agencies, comprising 13% of all such agencies, stopped minority drivers at least twice as frequently as expected based on their respective IDOT benchmarks. See Exh. 1 (IDOT report of 2013 Study Act data) at pp. 4-5. Moreover, there was a statewide racial disparity in the duration of traffic stops in 2013: the mean durations are 11 minutes for white drivers, 12 minutes for black drivers, and 13 minutes for Hispanic drivers. See Exh. 2 (IDOT data sheet for all of Illinois in 2013).
[2] See http://www.idot.illinois.gov/transportation-system/local-transportation-partners/law-enforcement/illinois-traffic-stop-study.
[3] For all its strengths, the Study Act should be improved in two ways. First, it should be made permanent. Under a necessary four-year extension enacted earlier this year, the Study Act is now scheduled to sunset in July 2019. See Public Act 98-686. Second, the Study Act should be expanded to cover sidewalk stops as well as traffic stops.
[4] See http://www.policeforum.org/assets/docs/Free_Online_Documents/Racially-Biased_Policing/by%20the%20numbers%20-%20a%20guide%20for%20analyzing%20race%20data%20from%20vehicle%20stops%202004.pdf.
[5] See http://www.economist.com/blogs/babbage/2011/02/animal_behaviour/print.
[6] People v. Caballes, 207 Ill. 2d 504 (2003); Illinois v. Caballes, 541 U.S. 972 (2004); People v. Caballes, 221 Ill. 2d 282 (2006).
[7] 541 U.S. at 411-13 (Souter, J., dissenting); id. at 421-22 (Ginsburg, J., dissenting). See also http://www.aclu.org/files/FilesPDFs/caballes.pdf (the ACLU amicus brief before the U.S. Supreme Court in Caballes, making similar arguments).
[8] See http://articles.chicagotribune.com/2011-01-06/news/ct-met-canine-officers-20110105_1_drug-sniffing-dogs-alex-rothacker-drug-dog.