New Impoundment Report 072110

Report on Impoundment of Vehicles Driven by Undocumented Immigrants in the City of Santa Barbara
July 31, 2010 (typos and minor errors corrected from 7/23/10 version) Russell Trenholme The author wishes to thank Belen Seara, Executive Director of PUEBLO and PUEBLO Education Fund board members Greg Prieto and Arthur Stevens for help with this report. The content and conclusions are the sole responsibility of the author. There has been no independent proofreading, and there are probably some remaining errors. Let me know if you discover any. The Excel spreadsheets and statistical analysis discussed in the report may be obtained by emailing me, [email protected] Background According to documents provided to PUEBLO by the Santa Barbara Police Department (SBPD) under the Public Records Law, the department issued 2,911 citations for “unlicensed driver” (CVC 12500) from the beginning of 2007 until mid-August 2009. Almost all of these citations resulted in 30-day impoundments of the vehicles driven at the time of the citation. This is because the department has a “mandatory impound” or “No license, no car, no exception” policy which departmental documents state is the “cornerstone policy” of the department’s traffic section. The overwhelming majority of the impounded vehicles were driven by undocumented immigrants who, since 1993, have been prohibited from obtaining a California driver’s license. A 30-day impoundment is a very severe penalty. The driver is typically deprived of a means of getting to work for a month. In addition, the driver not only must pay a large fine for the 12500 citation, but also must pay over $1,500 in towing and impoundment charges plus an “administrative fee” to the department before the vehicle can be recovered. The impoundment program is viewed by most Latinos as aimed at their community. The SBPD cites the following law (CVC 14602.6 (a)(1)) to justify these impoundments: 14602.6. (a) (1) Whenever a peace officer determines that a person was driving a vehicle while his or her driving privilege was suspended or revoked, driving a vehicle while his or her driving privilege is restricted pursuant to Section 13352 or 23575 and the vehicle is not equipped with a functioning, certified interlock device, or driving a vehicle without ever having been issued a driver’s license, the peace officer may either immediately arrest that person and cause the removal and seizure of that vehicle or, if the vehicle is involved in a traffic collision, cause the removal and seizure of the vehicle without the necessity of arresting the person in accordance with Chapter 10 (commencing with Section 22650) of Division 11. A vehicle so impounded shall be impounded for 30 days.

2 CVC 14602.6 is part of the “Safe Streets Act” of 1994 which was intended to severely punish drunk drivers and similar serious traffic offenders whose licenses had been revoked or suspended (guilty of violating CVC 14601). Note the wording that stipulates that the law also applies to unlicensed drivers who have never been issued a license. It does not apply to unlicensed drivers who previously had been issued valid licenses, thus it does not apply to all unlicensed (12500) drivers. In 1994, almost all Hispanic immigrants living in California, including undocumented immigrants, were in possession of valid California drivers licenses which were available independently of immigration status until 1993. At that time, the ranks of never-licensed drivers consisted principally of those who, for being underage or possessing physical or mental handicaps, were deemed incapable of safe driving. However, after 1993, as the drivers licenses of undocumented immigrants came to expire year-by-year without possibility of renewal, and as new immigrants crossed the border without visas, the number of unlicensed immigrant drivers grew rapidly. The result was that, in recent years, the overwhelming majority of unlicensed (including never-licensed in California) drivers have been undocumented immigrants rather than the mentally or physically handicapped persons who constituted the unlicensed driver group prior to 1994. In fact, these undocumented immigrant drivers came to greatly exceed the group that was the real target of the law: the serious offenders (14601 drivers) whose licenses had been suspended or revoked, typically for drunk driving so that now less than 20% of impounds are of vehicles driven by suspended or revoked drivers. (These issues are discussed in more detail in Appendices III and IV). This report lists five problems in the way the SBPD handles impoundments. These come out of an analysis of departmental procedures and actions as revealed by the documents turned over to PUEBLO. A list of recommendations follows the discussion of these problems. Appendices discuss the statistical analysis, legal issues, and the lack of foundation for the widely-held view that undocumented immigrants are dangerous drivers. Problem 1: The Santa Barbara Police are illegally impounding vehicles driven by previously-licensed drivers whose licenses have expired. Many undocumented immigrants entered the country legally with valid visas, obtained valid California licenses, but then overstayed their visas. As their licenses expire, with no possibility of renewal, these newly unlicensed drivers become subject to punishment for violation of CVC 12500. This is also true of undocumented immigrants who had obtained valid licenses prior to 1993. However, vehicles driven by such persons are not subject to impoundment under CVC 14602. The Santa Barbara Police Department “mandatory tow” policy is in violation of CVC 14602 which only sanctions 30-day impounds of vehicles driven by “never licensed” drivers. In fact, the SBPD treats all unlicensed drivers (12500s), whether or not previously licensed, as subject to having their vehicle impounded for 30 days, and all

3 unlicensed drivers are treated the same as the serious offenders (14601s) whose licenses have been revoked or suspended. All departmental documents and forms reflect this error, treating a 12500 citation as automatic grounds for a 30-day impoundment. For example, the report sheets used at DUI checkpoints, and for various patrols, treat a 12500 violation as one of the two possible reasons for an impound (the other being 14601). This is not a gray area of legal interpretation. The State Attorney General issued a report critical of the Maywood Police Department, and the report explicitly criticizes the department for failing to check whether the driver had ever been issued a license. Problem 2: The Fourth Amendment requirement that the police only impound vehicles that are creating public safety issues is consistently ignored by the SBPD. The 9th Circuit Court of Appeals (which governs California) recently ruled that mandatory impoundment is in violation of the Fourth Amendment (Miranda v City of Cornelius http://caselaw.findlaw.com/us-9th-circuit/1075130.html ). An impoundment must be justified through the “community caretaking doctrine.” Police cannot simply claim to be following a state law (whose constitutionality is currently before the same court) because impoundment under CVC 14602.6 is discretionary (“may” not “shall” is the operative word). In the Miranda v Cornelius case, the vehicle driven by an unlicensed driver was safely parked in a private driveway when it was impounded. The Santa Barbara police have been known to impound a car safely parked in a private garage when the police officer did not even observe the unlicensed driver driving the car. This is clearly contrary to the Miranda decision, as the 9th Circuit Court’s comments in a recent case in Glendale, make clear: “Our recent decision in Miranda v. City of Cornelius, 429 F.3d 858, 865 (9th Cir.2005), held that the government has the power to seize vehicles if ‘the driver is unable to remove the vehicle from a public location without continuing its illegal operation.’ ” This failure to distinguish public and private locations is only one extreme example of the failure of the SBPD to follow the community caretaking doctrine in deciding whether an impound is justified. Legal issues are discussed further in Appendix III. Problem 3: There is strong evidence that immigrants are being illegally targeted by some officers on Saturation and Directed Patrols Number of 12500 Impounds: Because of the department’s mandatory impound policy, it is possible to estimate the number of 30-day impoundments by simply adding up the number of 14601 and 12500 citations issued. The department provided 12500 and 14601 citation data to PUEBLO for all of 2007 and 2008 and for the first seven and one half months of 2009 (see Appendix I). This data indicates that 599 vehicles driven by revoked or suspended drivers (14601s) were impounded whereas 2,911 vehicles driven by unlicensed drivers (12500s) were impounded (83%). When we look at the statistics on other categories of citations—citations for moving violations, for equipment violations, drunk driving arrests— we find that a 12500 citation is the most common citation given by the Santa Barbara Police Department over the past two years, exceeding the total of all the various moving and equipment citations combined. It appears that impounding vehicles driven by undocumented immigrants is the principal activity of the traffic

4 section. Moreover, the increase in 12500 citations is notable: from 912 in 2007 to 1377 in 2008 and a projected total of 1678 for 2009. This is a 51% increase from 2007 to 2008 and an 81% increase for the first seven months of 2009 over the first seven months of 2008 (the only months for which we received complete 2009 data). 1 Among the data turned over by the department are report sheets for each checkpoint conducted during 2007, 2008 and part of 2009, for each “DUI Saturation Patrol” conducted during 2008 and part of 2009, and for “Directed Patrols” conducted during the same period (these two types of patrols are discussed below). We were also given data summarizing, by month, all 12500 and 14601 citations during 2008 and most of 2009. 2008 is the only year for which we have comparable statistics for each category. The data shows that 1198 12500 citations were issued during 2008. However, the checkpoints, Saturation Patrols, and Directed Patrols, jointly total only 296 12500 citations. Thus 902, or 75%, of the 12500 citations issued (during 2008) are unaccounted for by the report sheets provided PUEBLO. I was told that these 12500 citations were issued during routine patrol activities for which no report sheets are filed. So our analysis is limited to what happened at checkpoints, or during Saturation and Directed Patrols as revealed by the report sheets we did receive. Sobriety Checkpoints: Checkpoints have been a focus of protests and complaints by Hispanic drivers because of their intrusiveness and because of the perception that their primary focus often appears to be on impoundment rather than drunk driving. Checkpoints were sanctioned by the California and U.S. Supreme Courts as justifiable exceptions to the Fourth Amendment prohibition on unwarranted searches and seizures because of the “crisis” posed by drunk drivers. Both courts recognized that “sobriety checkpoints” are prima facie in violation of the Fourth Amendment, but ruled that such checkpoints could be conducted provided adequate safeguards to minimize their intrusiveness were put into place. In its split decision, the U.S. Supreme Court accepted the checkpoint procedures used in the State of Michigan. According to these procedures, driver’s licenses were only checked when the driver of the vehicle stopped at the checkpoint showed evidence of intoxication. The California Supreme Court imposed a number of restrictions on how checkpoints are to be conducted (known as Ingersoll guidelines); for example, checkpoints were to be publicly announced in advance, stops had to be independent of officer discretion (for example, every car, or every other car, is stopped); vehicles had to be given an opportunity to turn around and avoid passing through the checkpoint, checkpoints were to be conducted at times and places where drunk driving has been a problem, etc.
1

The zeal with which the impoundment program has been pursued by the Santa Barbara Police Department is attested to by the expenditure of taxpayer money to promote the program through the distribution of “badges, stickers, ball caps, bumper stickers, T-shirts, coffee cups, press kits, and pamphlets,” as well as the production of television commercials designed to build public support for the program. Here is a specific comment from one police document: “The artwork will show a vehicle behind a tow truck with a motor officer and his motorcycle next to the truck. The inscription will say, “No LicenseNo vehicle Santa Barbara Police Department STOP team.” This logo was to be reproduced on 300 T-shirts, 50 Polo shirts, 100 Tote bags, 50 baseball hats, 2500 pens and a white banner which would read, “Santa Barbara DUI Checkpoint.”

5

A recent widely published study by The Investigative Reporting Program at U.C. Berkeley ( http://californiawatch.org/public-safety/car-seizures-dui-checkpoints-proveprofitable-cities-raise-legal-questions ) revealed that throughout California checkpoints have focused not on apprehending drunk drivers but on impounding vehicles driven by undocumented immigrants, and that these impoundments are an important source of funding for police departments through the imposition of “administrative fees.” The study found that in some jurisdictions impoundments of 12500-driven vehicles occur at seven times the rate at which drunk drivers are arrested. The administrative fees are typically used to fund overtime pay for police, thus creating an incentive to police officers to maximize impounds. This is also the case in Santa Barbara where administrative fees are used to fund police overtime for the Directed Patrols which show the highest rates of 12500 impoundments. In Santa Barbara, conformity to the Ingersoll guideline requiring prior notification of a checkpoint is perfunctory and ineffective. The department simply sends an email to major news organizations which ignore them: the notice of an upcoming checkpoint never reaches the public. In addition, checkpoints are set up in such a manner that it is very difficult for drivers to anticipate and avoid them. However, these failures have an interesting consequence. Since drivers enter checkpoints unknowingly without possibility of exiting, vehicles stopped at checkpoints represent a relatively unbiased sample of vehicles and drivers on the road at night during the hours in which checkpoints are conducted. In particular, the data received by PUEBLO on checkpoint stops provides good estimates of the percentage of 14601 (revoked or suspended) and 12500 (unlicensed) drivers who are driving at night. The checkpoint data indicate that these percentages are quite low: only 4/10 of 1% for 14601 drivers, and 1.2% for 12500 drivers. (In the case of 14601 drivers, only 46 vehicles out of 12,344 checkpoint stops; and only 155 12500-driven vehicles. The percentage of DUI drivers discovered is only 6/10 of 1%. Rather than focusing on the abuse of checkpoints (the principal focus of the UC Berkeley investigative report), this report focuses on impoundments resulting from patrol activity. Saturation Patrols and Directed Patrols and Potential Targeting of 12500 Drivers: Given these low percentages, it is surprising to find that 5.6% of the vehicles stopped during night-time “Saturation Patrols” received 12500 citations, as did 18.5% of the drivers stopped on night-time “Directed Patrols”.2 These high percentages suggest that patrol officers have found a way of identifying and targeting target unlicensed drivers. Since the vast majority of 12500 drivers are undocumented immigrants, the easiest way of targeting an unlicensed driver is to target an undocumented immigrant. This may be done by racial profiling—not easy during a late night patrol; by focusing patrol activity on immigrant neighborhoods; or by focusing attention on the type of vehicles typically driven by undocumented immigrants.
2

Directed Patrols were originally supported by state grants used to implement the department’s “STOP” program which targets unlicensed drivers. These patrols are now “self-funded” through the administrative fees collected from impoundments.

6

The latter method can be quite effective, since in a wealthy community like Santa Barbara, the overwhelming majority of older, low-value vehicles are driven by the poorest segment of the population: unlicensed immigrants. It is not only a question of poverty, however. Since undocumented immigrants know that they face a substantial risk of having their vehicles impounded for 30-days and the recovery cost exceeds $1500, they rationally conclude that it is better to drive a “junker” (“carcacha”) which can be abandoned in case of impoundment if they lack the funds to recover the vehicle. The Santa Barbara Police Department strongly denies profiling immigrants. But profiling is an emotionally-fraught term, with strong connotations of prejudice based on racial characteristics. Even the notorious Sheriff of Maricopa County in Arizona, Joe Arpaio, denies profiling. As just noted there is a much more effective means of identifying undocumented immigrant drivers. From a legal standpoint, the issue of identification of these drivers is besides-the-point. The Fourth Amendment has consistently been interpreted by the courts as allowing for a stop to check documents in one case: evidence of a probable violation of the law. Typically, the law involved in a traffic stop would be a traffic law: the driver is observed speeding, observed making an illegal turn, driving without required lights, etc. These infractions are classified on police reports as either moving or equipment violations. In addition, a vehicle may also be observed driving in a manner that suggests intoxication (the criteria are spelled out by the National Highway Traffic Safety Administration). It is not legal to apply a stricter standard to an older vehicle that the officer assumes belongs to an undocumented immigrant.3 If unlicensed drivers are being legally stopped at very high rates, then they must be committing traffic offenses at very high rates compared with other drivers. Despite all the stereotyping of the “dangerous unlicensed driver” by the police, this seems highly unlikely. The lack of any factual basis for these stereotypes is discussed in Appendix IV. In fact, a common observation is that unlicensed immigrant drivers are timid and careful because they are afraid of being stopped and having their vehicles seized and impounded for 30 days. But we don’t have to rely on anecdotal evidence. If unlicensed drivers— almost all of whom are undocumented immigrants—are really especially bad drivers who are stopped for legitimate reasons, then they would be receiving a proportionately large number of moving and equipment citations. Patrols with a high rate of 12500 impounds should be giving out traffic citations (for moving or equipment violations) at a similarly high rate. But, in fact, the patrols that impounded the highest percentage of 12500
3

The Federal Government is investigating the Maricopa, Arizona, Sheriff’s department for selective enforcement various laws, including minor traffic offenses. Almost all those detained are Hispanics. From a recent story: “Thursday's operation would mark the 17th time Arpaio has deployed hundreds of deputies and volunteer posse members in an area of the Valley to root out illegal immigrants. Deputies typically take a "zero tolerance" approach to traffic offenses…” http://www.azcentral.com/community/phoenix/articles/2010/07/29/20100729joe-arpaio-immigrationsweep.html#ixzz0vES2KfIg It is not legal to concentrate strict enforcement of traffic laws (e.g. driving slightly over the speed limit) on certain ethnic groups, regardless of the motivation. Traffic laws must be applied impartially.

7 vehicles actually gave out fewer moving or equipment citations! But if these 12500 drivers are not being stopped for traffic violations, then they are being stopped illegally. The high rate of 12500 impounds on many Saturation Patrols and Directed Patrols suggests that some police have found a way of targeting unlicensed immigrant drivers and unlawfully stopping them. PUEBLO received usable report sheets for 50 Saturation Patrols that took place during 2008 and the first six months of 2009. During a majority of the patrols (29 patrols), 332 vehicles were stopped, 58 citations were given for moving and equipment violations (17% of stops), 5 suspended or revoked (14601) drivers were cited, but not a single unlicensed (12500) driver was encountered. In statistical terms, the modal percentage of 12500 citations is zero, and the median percentage is also zero. This is what we would expect if undocumented immigrants are average drivers who form a very low percentage (1.2% if we accept checkpoint statistics) of night-time drivers—that is, if undocumented immigrants are not being illegally targeted for traffic stops during patrols. The police running these patrols appear to be acting in keeping with the law. The overall 6.5% rate of 12500 citations on Saturation Patrols came from the remaining 21 patrols which have a combined 12500 rate of 11.4%. These 21 patrols stopped 370 cars and cited 42 drivers for 12500 but instead of giving out more moving and equipment citations as would be expected if the stops were legal, they gave far fewer moving and equipment citations than the zero 12500 patrols: such citations occurred during only 10% of their stops versus 17% during the 29 patrols that found zero 12500s. It certainly, doesn’t look as if the 12500 drivers were stopped because of dangerous driving. The evidence from the Directed Patrol data is even stronger. We have usable data for Directed Patrols in 2008, and 102 in 2009 (through November). Altogether there were 539 12500 impoundments in 2,895 vehicle stops for a rate of more than 18.6%. But the distribution is highly skewed, with a relatively low peak and some extreme high 12500% outliers pushing up the mean (average). The modal (most common) percentage of 12500 stops is zero (the same as with Saturation Patrols). Twenty-two patrols with 277 stops gave 89 moving or equipment citations, cited 12 drivers for suspended or revoked license (14601), but did not encounter or cite a single unlicensed (12500) driver. In contrast, there were another 22 patrols with 184 stops that cited half or more of the vehicles they stopped for 12500 (102 in all). And these high 12500 percent patrols had a lower stop percentage for traffic violations than the 22 patrols that encountered zero 12500 drivers. The high 12500 patrols only impounded 7 cars for 14601. It is hard to believe that the cars of these unlicensed drivers were stopped for legal reasons (probable cause of a law violation) during these 22 high 12500% patrols. To get further insight into the high 12500 citation rates during Directed Patrols, we compiled statistics by individual officers. We focused on the 21 (of 38) officers who ran at least two patrols in order to have a sufficient basis for comparison. These “key”officers may be divided into four groups or clusters based on their 12500 citation rate (per stop). The first cluster (which forms the lowest quintile of the key consists of four officers with 12500 rates of 3% to 7%. The second cluster consists of 9 officers with 12500 citation

8 rates from 9% to 13%. The third cluster consists of 4 officers with 12500 citation rates from 17% to 23%. The last cluster consists of the four officers with the highest 12500 citation rates, ranging from 33% to 48%. This group represents the highest quintile of the distribution. In fact these four officers ran 19 of the 22 patrols with 12500 stop rates greater than 50%. If all the patrols were operated under the same conditions, including stop procedures, the distribution of the 12500 rates would approximate a normal distribution, and we can use standard statistical methods to calculate the probability of various stop rates. These calculations are explained in Appendix II. Further evidence that unlicensed drivers are being targeted on patrols comes from a comparison of 14601 (suspended or revoked) and 12500 (unlicensed) citation rates. Since neither offense is visible until the driver’s license is checked, in the absence of profiling or targeting we would expect the ratio of 12500 to 14601 citations to be relatively constant, or for the percentage of 14601s to increase relative to 12500s on patrols (because of the proven bad driving of revoked and suspended drivers and also because patrol officers carry hotsheets identifying them; looking for these drivers is a primary mission of Direct Patrols). But the opposite is the case. During checkpoints, which randomly sample the population of night drivers, there are three times as many 12500 citations as 14601 citations (1.2% versus 0.4%). In other words, 75% of impounds at checkpoints were 12500s, only 25% 14601s. But instead of a relative increase in 14601s as the result of hot sheets, 14601s actually declined to just 13% of the total during Directed Patrols, and 12500s climbed from 75% to 87%. Thus the ratio of 12500s to 14601s increased from 3:1 to almost 7:1. It is difficult to think of any explanation for the relative increase in in 12500 citations during Directed Patrols other than that immigrant drivers (who constitute the vast majority of unlicensed drivers) are being identified then illegally stopped for a drivers license check. The patrol report sheets thus provide strong evidence that some officers are not following legal stop procedures and that, by one means or another, these officers are targeting undocumented immigrants. In Appendix II, there is a more technical discussion of the statistical evidence for targeting. It is possible to show that the distribution of 12500 stops on Directed Patrols is so skewed that it is a statistical impossibility that certain officers could have legally stopped so many 12500 drivers by chance if the drivers were following consistent procedures. On the other hand, the statistics for the low 12500% officers are consistent with a normal distribution, and are highly probably given a 1.3% mean 12500 rate as indicated by checkpoint data. One possible reason for the high 12500 impoundment rate found during Directed Patrols is that the program is “self-funded”; that is, the overtime pay for the officers sent out on Directed Patrols is paid for by the administrative fee charged when impounded cars are recovered. If so, it is perhaps not surprising to discover that officers on Directed Patrols cite a far higher percentage of 12500 drivers than officers on other patrols (e.g., Saturation Patrols) or at sobriety checkpoints.

9 Problem 4: Officers with suspiciously high 12500 stop rates are sent out on Directed Patrols much more often than the officers with low 12500 stop rates If we look at the individual statistics for the four high 12500% officers just mentioned (the fifth quintile of the key officer distribution), we find that they are sent out on patrol much more than the officers with the lowest 12500 stop rates (the first quintile): 71 patrols in the case of the high 12500 officers compared with only 16 patrols for the low 12500 officers. Could this be the result of the department’s emphasis on impoundments? Or perhaps is it perhaps because the 12500% officers are producing the revenue (through the Administrative Fees collected when cars are recovered from impound) to pay the overtime of Directed Patrol officers?

10 Problem 5: Inadequate reporting serves to cover-up potential illegal impounds The evidence just presented from Saturated Patrol and Directed Patrol data strongly suggests that some patrol officers are targeting undocumented immigrant drivers for driver’s license checks and likely impounds. Eyewitness accounts are hard to obtain. Undocumented immigrants almost never contest impounds because fear of deportation causes them to remain silent. Better documentation is the best way for the department to identify over-zealous officers who are making illegal stops and would serve as a deterrent to abuse. Another problem is the lack of any departmental training or clear guidelines on legal vehicle stop procedures. PUEBLO asked the Santa Barbara Police Department for all documents relating to stop procedures; none were provided. It appears that there are no written guidelines or directives regarding permissible traffic stops. The report sheets are incomplete in important respects. First, they fail to record a reason for a stop that leads to an impound. The reports provide no information on where the stops occurred; in fact, even overall patrol locations (and checkpoint locations) are never indicated on the reports so there is no way of confirming that the locations were locations with high accident or drunk driving incidents as recommended by the National Highway Traffic Safety Administration. The State Attorney General’s investigative report on the Maywood Police Department contains language that applies word-for-word to the Santa Barbara Police Department: Review of Department impound records and other police reports revealed the repeated failure to articulate, with sufficient detail, the facts supporting reasonable suspicion for stops. When a traffic stop results in a citation and impound, it is standard police practice for the citing officer to note the initial reason for the traffic stop (the applicable code violation and short narrative) in the officer’s report. …few drivers who were subjected to impounds were cited for any purported violations that led to the stop; rather the majority of citations were issued solely for the violation of driving without a valid driver’s license. This lack of information regarding the reasonable suspicion required for the stops is highly irregular and supports a strong inference that the Maywood Police Department engaged in a pattern or practice of conducting unlawful traffic stops. (pp. 13-14) The Attorney General’s report also criticized the sketchy manner in which reports are completed, a problem that also characterizes Santa Barbara patrol reports: The Maywood Police Department has accepted poor and insufficient report writing by officers. (p. 24) Since immigrants are afraid to speak out, the best restraint on illegal stops is strong oversight by the department, and such oversight depends on complete documentation of

11 stops. Without this information, the department can simply claim that profiling is against department policy and that it has no knowledge of profiling. Because reports may also be falsified, the State Attorney General recommended that Maywood patrol cars be equipped with video cameras to record what took place during a traffic stop. The problem goes far beyond the deficiency of Saturation and Directed Patrol reports, however. Statistics reveal that (during 2008, the only year for which we have full statistics for all categories) about 75% of 12500 citations are unaccounted for by Checkpoint, Saturation Patrol, or Directed Patrol stops. These unaccounted for 902 12500 citations are not covered by any reports! According to my conversation with Sergeant Rivas, head of the Traffic Section, these missing impounds result from normal daytime patrol activity which is not documented by report sheets. It is essential that all traffic stops resulting in 12500 citations be fully documented; it is not sufficient to focus only of Saturation and Directed Patrols.

12

Specific Recommendations Relating to Reports, Forms, and Procedures 1. Create a new report form to be used for each 12500 or 14601 citation not arising from a checkpoint. The report should give the location of the stop and the legal justification for the stop. Normally, a legal probable cause results in a moving or equipment citation. In the case of an impoundment of a car driven by a driver cited for 12500, the report should note: 1. whether the 12500 driver offered any evidence of having a current or expired non-California driver’s license or an expired California driver’s license , or claimed to have had one, and how the decision to impound was arrived at despite this evidence, and 2. The reasons the car was towed rather than being turned over to a licensed driver or safely parked. 2. Install and use video cameras to record stops. This is far less costly than in the past, and in the short run, cameras could be installed in the cars of officers with suspiciously high rates of stops of 12500 drivers. 3. All checkpoint and Saturation or Directed Patrol reports should indicate the location of the checkpoint or of the area covered by the patrol. All 12500 or 14601 citations issued during a checkpoint that occur as a result of stops of vehicles outside the checkpoint itself should be individually documented as in recommendation 1 above. 4. Clear written guidelines should be provided as to the legally justifiable reasons for a traffic stop. Officers whose citation activity is abnormal with respect to the large number of 12500 citations or low numbers of moving and equipment citations relative to stops should be questioned as to their stop procedures and steps should be taken to ensure that all stops are legally justified.

13

Appendix I Discussion of Documents Received Under The Public Records Act
The statistics discussed in this report were complied from documents delivered to PUEBLO by the Santa Barbara Police Department in response to a request under the Public Records Act. These include reports for each checkpoint conducted during 2008 and part of 2009, for each “DUI Saturation Patrol” conducted during 2008 and part of 2009, and for “Directed Patrols” conducted during the same period. (We asked for all such reports, and if any are missing, they would be insufficient to affect the analysis.) Scanned samples of Checkpoint, Saturation Patrol, and Directed Patrol report sheets follow this introduction. A discussion of how discrepancies were handled follows the sample the sample of each type of report sheet. Finally, there is a scanned copy of Document 000224 giving the monthly statistics for 12500 and 14601 citations. All original documents are held in the PUEBLO office and copies of the original spreadsheets are available for further analysis.

14 Saturation Patrol Report Sheets Sample (0 12500s in 8 stops: 0 is the modal number of stops on Saturation Patrols)

Saturation patrol report sheets have an entry entitled “Total Number of Impound Tows” with the further comment “(Including DUIs arrested for 12500/14601).” Immediately below this entry are two entries, one for the number of 14601 tows and one for the number of 12500 tows. In principle, the numbers entered for the two types of tows should total to the listed total number of impound tows, but in four cases (000111, 000118, 000125, and 000137) the total is one less. Where there is an entry for a DUI arrest we assume that it involved either a 14601 or 12500 tow, but the officer failed to record

15 which probably explains the shortfall. We used the data separately entered for 14601 and 12500 tows in these cases and did not attempt to judge whether the unlisted impound was for 14601 or 12500. This results in a small underestimation of the total number of impounds. In one case, however, (000110), the total number of impound tows is given as 2, but no tows for 14601 or 12500 are given below, and in this case there was only one DUI arrest. This report (000110) was simply omitted as being too confusing for analysis. Thus, it is likely that a small number of impound tows (maximum 6) have been omitted due to careless recording. There is also an entry for the total number of 12500 and 12601 citations. This almost always equals the total of the number of 12500 and 14601 impounds entered. However, in three cases (000132, 000092, and 000103) there is one more entry in the citation total than the total of separately listed 14601s and 12500s. This may indicate that a citation was given but that the vehicle wasn’t towed. Since we are interested in impound tows, we used the separately listed numbers of 14601 and 12500 tows rather than the citation figures. Finally, in one case (000083), the total number of impound tows was given as 5, the total number of separately listed 12500 and 14601 tows is also 5, but only 4 citations were issued. However, there was one DUI arrest which probably explains the missing citation. In summary, in only one case did the discrepancies result in omitting the Saturation Patrol report.

16 Directed Patrol Report Sheets Sample (63% of stops were 12500; this is a Fifth Quintile officer)

17 A second Direct Patrol report sheet is reproduced because it graphically illustrates the pressure to produce impoundments. Patrol officer 7 says “Sorry!” ong side a “frowny face” just below his 0 tally of impounds. This despite the fact that he gave an aboveaverage percentage of moving and equipment citations. Officer 7 has typical 12500 stats, with an average at the median (12%) when the high outliers are excluded.

In general, the same protocol was used in analyzing Directed Patrol report sheets as for Saturation Patrol report sheets. The totals given separately for 14601 and 12500 impound tows are used, even if their total is less than the listed total of 14601/12500 citations. There are five cases like this: 000501, 000493, 000454, 000415, and 000343. It is possible that in some a citation was issued but the vehicle was not towed; or perhaps it is

18 simply sloppy record keeping. Since the principal interest is in impounds, the key numbers are the numbers given for each type of impound. Report 000446 shows 3 impound tows, but only 2 citations; this looks like sloppy is presumably sloppy record keeping. Report 000437 shows a total of 4 impound tows but the separate entries only total to 3; again this looks like is sloppy recording. Similarly with 000409 which shows a total of 2 tows, but 3 12500 tows. Since we have clear separate totals for 12500 and 14601 impounds, these reports were used in calculating statistics. The only report eliminated from the Directed Patrol analysis is 000433 which does not list the number of traffic stops making it impossible to calculate 12500s or 14601s as a percentage of stops.

19 Checkpoint Report Sheets Sample (less than 1% of the 540 checkpoint stops were 12500-driven vehicles)

One checkpoint report was eliminated due to confusing entries: 0000069 shows a total of 6 14601/12500 impound tows in 308 stops but only indicates only 1 12500 tow and no 14601 tows (under the entry that separates the two). Although the 1.9% overall impound rate is in general agreement with the rate found on other patrols, it was decided to drop the report since it is impossible to know how many of the remaining 5 tows were 14601s and how many 12500s.

20 There is also some indication from other data that checkpoints occurred on 5/6/07 and on 4/23/09 but we received no reports for those dates. Nevertheless, we have clear data on 37 checkpoints conducted from January 20, 2007 to July 19, 2009 and these form the basis of the analysis. Another potential source of confusion on the checkpoint report sheets is that there is an entry for total number of traffic stops and another entry for total number of vehicles through checkpoint. It turned out that the first entry refers to cars stopped outside of the checkpoint or pulled out of the checkpoint line because of some obvious violation. The second entry is for the actual number of cars through the checkpoint. In addition to the 12,344 vehicles indicated to have passed through the 37 checkpoints, the reports show that another 297 vehicles were stopped outside the checkpoints. Since the 12500s are not distinguished as to whether they was discovered in the checkpoint line or outside it, we felt compelled to include all 12500s, including 12500s arising outside of checkpoints, but we did not add 297 to the checkpoint total in calculating the frequency of 12500s. This has the effect of raising (we don’t know how much) the calculated percentage of 12500s discovered at checkpoint stops.

21 12500 and 14601 Citations by Month from January 2007 to Mid-August 2009

22

Appendix II Statistical Analysis
The data from the stat sheets was tabulated and entered onto spread sheets which were then analyzed using standard statistical functions provided with Excel. These spread sheets may be obtained by emailing the author of this report at [email protected]. Other spreadsheets may be available from Greg Prieto who is writing his dissertation at UCSB on related material. Greg may be contacted through PUEBLO. The estimate of the frequency of 12500 drivers on the road at night is based on the 1.3% (actually 0.012557) 12500 rate observed during 12,652 checkpoint stops. It assumes that checkpoint stops approximate a random sampling of the drivers on the road at night. To test this assumption, we ran the Excel Statistics Descriptive Data module on the data from the 34 checkpoints; these are treated as “samples” from the underlying population. The mean of these “samples”is a little higher than the overall mean, 1.6%. This results from 8 checkpoints (almost 25%) with much higher than expected 12500 rates. We ran a 99.9% confidence interval on this mean (which gives the range of means that occur by change with a probability greater than one in a thousand). This revealed that there is 99.9% probability that the population mean lies between 0.08% and 2.4%. Eight checkpoints fell outside the upper bound of this range, with one at 4% and one at 5% which indicates that these checkpoints are atypical—perhaps because they were located in or near immigrant neighborhoods or were conducted earlier in the day when the percentage of undocumented immigrant drivers is expected to be higher. The 1.3% figure is more accurate than the 1.46% figure because it is derived from the pool of all checkpoint stops, but the closeness of the two means and the very tight confidence interval support use of checkpoint data in estimating the percentage of 12500 drivers on the road at night. To better understand the statistical analysis of Directed Patrol stops that follows, we begin by reviewing the conditions that would obtain if the patrols were conducted in a legal and unbiased manner. The four key assumptions are: 1. Grant-funded patrols are operated under requirements that insure relative uniformity with respect to patrol areas and patrol mission; 2. Patrol officers apply a consistent set of standards to traffic stops; 3. These standards comply with state and federal law regarding legal traffic stops; 4. The same stop standards are applied to immigrant drivers as to other drivers.. Given these four assumptions (which are consistent with the claims of the Santa Barbara Police Department), the citations for 12500 given on a patrol will be distributed according to a binomial distribution. That is, the percentage of 12500 drivers on the road during the usual patrol hours (generally 10 P.M. to 4 A.M.) who commit traffic infractions justifying stops will form the mean of the distribution and the probabilities of various numbers of 12500s relative to a given number of stops during a patrol can be calculated statistically. Similarly, the probability of various numbers of 12500 citations relative to total stops by an individual officer can be calculated.

23 If 12500 drivers commit traffic infractions at a rate approximating that of all drivers, the mean should be close to the 1.3% frequency of 12500 drivers observed at checkpoints. The actual mean 12500% on Directed Patrols is much higher: 18.6%. If all these 12500 stops are legal then 12500 drivers are exceedingly dangerous drivers, committing moving and equipment violations at 12 times the rate expected from their 2.3% frequency on the road at night. Whether the stops are legal or not (that is, excluding assumptions 3 and 4), we can test to see whether patrols are run under uniform conditions. (These tests are independent of any assumptions about the true mean number of 12500 drivers on the road at night; then apply whether or not 18.6% approximates the true mean.) If patrols are run in accord with assumptions 1 and 2, then the distribution of 12500 stop percentages should approximate the textbook statistical model of selecting white and black balls (sight unseen) from a very large urn. This produces a binomial distribution (which approximates a normal distribution given the large numbers involved). The distribution would be symmetrical around the mean (which, if stops are legal, would be the percentage of 12500 drivers committing traffic infractions relative to all drivers committing traffic infractions that justify a stop). By testing to see whether the actual observed distributions fit the model of a binomial distribution, we can see whether the distribution is consistent with assumptions 1 and 2. We ran the standard Excel Descriptive Statistics Module on the 203 Directed Patrols for which we were given usable data for 2008 and 2009. The mean stop rate per patrol is 21.7%. (Again, this is the mean rate per patrol rather than the 18.6% frequency found from pooling all Directed Patrol stops. It is higher because the distribution is skewed high by a number of high 12500% patrols.) A binomial distribution is symmetrical about the mean, the median equals (or at least approximates) the mean and the mode, and the skewness measure will be low (below 0.344 according to statistical theory). In fact, the Excel Descriptive Statistics Data reveals that the actual distribution is highly positively skewed (0.798): the median is 0.1847 (compared with the mean of 21.7%), and the modal frequency is zero—there were more patrols with no 12500 stops than any other frequency. Compared with a symmetrical distribution, there were a disproportionate number of patrols with a very high rate (up to 75%) of stops that resulting in 12500 impoundments. This is confirmed by a histogram and by a plot of the z-values of the patrol stop rates . The graphs clearly reveal the skewness, with a number of patrols in the far right tail. These statistics are inconsistent with the assumption of a binomial distribution and indicate a lack of uniform stop procedures. It also suggests that the high 12500% outliers are illegally targeting of 12500 drivers for stops. And the fact that the modal 12500 stop rate is zero suggests that the 18.6% mean results from illegal stops of undocumented immigrant drivers. Further evidence against the assumption that 12500 drivers are much worse than average is discussed in IV. We then repeated the statistical tests by looking at the traffic stops conducted by the 21 “key”officers who ran patrols (officers who ran only one Directed Patrol were excluded). If these officers were following uniform procedures we would expect high and low 12500 rates to be randomly distributed among the officers. However, the distribution of

24 12500%s per stop, grouped by key officer, is even more skewed than the data for individual patrols. The mean (average percent of 12500s citations by officer) is 0.1679 (16.8%), but the median number is only 0.1176 (11.8%).4 The skewness is 1.2719 which again indicates an unexpectedly long right tail (i.e., some officers with improbably high 12500 rates). The z-value plot confirms this, with the 12500 rates of Fifth Quintile Officers 17, 18, 19, and 20 extreme outliers at 33%, 34%, 37%, and 48%, respectively. In fact, 19 of the 22 patrols with 12500 rates of 50% or greater were operated by one of these four officers. The histogram shows the outliers well to the right of the rest of the distribution. Again we tested to see whether this is just an unlikely statistical anomaly. The four Fifth Quintile officers had 33% or more 12500 stops. Assume that the distribution is a binomial distribution about the mean of 0.1679 (16.8%) for the 21 officers’ stop rates. (Grouping patrols by key officer produces a slightly different mean than the aggregate frequency of 18.6%.) We then used the binomial function BINOMDIST in Excel to calculate the probability that at least one of the 20 officers would have encountered k or more 12500 drivers among the n cars stopped on the officer’s patrols—where k is the number encountered by the officer and n is the number of stops. (The calculation uses the 0.1679 mean as the probability of encountering a 12500 driver using the supposedly standard procedures assumed for Directed Patrols. The calculations are explained in detail below.) The Fifth Quintile outliers are Officers 17, 18, 19, and 20. Assuming a binomial distribution (reflecting uniform conditions), what is the probability of one of the 21 key officers encountering as many as 62 12500 drivers in 190 stops? The answer is 0.000001528, less than one in 640,000. Are we to believe that Officer 17, with this probability against his 12500 stop rate, was conducting only legal stops? What is the chance of one of the 21 officers encountering as many as 164 12500 drivers in 488 stops? The answer is zero to beyond 9 decimal places, or less than one in a trillion (due to the large number of stops). But Officer 18 cited 164 drivers for 12500 during his 488 stops. What is the chance of one of the 21 officers encountering as many as 40 12500 drivers among 109 stops? The answer is 0.00001064 or less than one in 98 thousand. But Officer 19 (with a 37% rate of 12500s) did cite 40 drivers for 12500 during the 109 stops he made. And what is the chance that one of the 21 officers would encounter 20 12500 drivers among 42 stops? Because of the smaller sample size, the probability is a little larger, 0.000075585 or less that one in 13 thousand. Officer 20 encountered 20 12500 drivers during his 42 patrol stops.
4

The reason the mean for the key officers is lower than the overall mean is that the high 12500 officers made many more stops than average thus pushing the overall percentage above the mean for patrols sorted by officer. The means using the individual 12500 rates weighs all key officers equally, thereby reducing the disproportionate effect of the high 12500 officers.

25

And what is the probability that any selection of four officers out of 20 would jointly produce such extremely high 12500 rates? The odds against this happening are astronomical, and provide further grounds for rejecting the assumption that uniform procedures are followed during Directed Patrol stops. What about the four officers with the lowest 12500 citation rates (The First Quintile)? We calculated the probabilities that one of the 21 officers would produce 12500 rates as low or lower than those of each of the officers in First Quintile (Officers 1, 2, 26, and 3). Converted to percentages, these are (respectively) 18% (one in 6), 86% (almost one), 62% (one in 2), and 3% (1 in 33). These odds suggest that only Officer 3 exhibited a 12500 stop rate that suggests non-conformity to the hypothesized standard that produced the 16.8% average. However, the 12500 stop rates of these four officers are completely consistent with following legal standards for stops, with the 12500 stops forming a binomial distribution about the 1.3% mean found for checkpoint stops? If 1.3% is the true mean (given uniform and legal stop procedures and average infraction rates by 12500 drivers), then the 12500 stop percentages they encountered have probabilities all are extremely close to 1 (certainty); that is, it is almost certain that at least one officer out of 20 would encounter the number of 12500 drivers during n stops that were found by each of the four officers. On the other, the probability of the four highest 12500 officers obtaining the 12500 rates they did falls to zero to beyond nine decimal places if the true mean is close to 1.3%. Most of the statistics cited in this report are straight-forward Excel functions or were produced by the Correlation, Descriptive Statistics, or Histogram modules in the Excel Data Analysis Pak. Technical Discussion: The calculation of the probabilities associated with the possibility of an officer who follows a uniform set of procedures obtaining some of the extreme results found in the data were calculated using the POWER and BINOMDIST functions. (Many of the calculations were done using a Spanish-language version of Excel and the function names are in Spanish. However, their English translations are obvious.) We illustrate the procedure using the Directed Patrol stats for Officer 17 who cited 62 out of 190 cars stopped for a 12500 violation. First, we calculate the probability that an officer stopping 190 cars using standard procedures (with a mean group 12500 rate of 16.8% (0.1679) of stops) would not encounter as many as 62 12500 drivers. This is calculated by BINOMDIST(61, 190, 0.1679, TRUE). (61 rather than 62 is used because the function calculates the probability of encountering 61 or fewer 12500 drivers from a stop of 190 vehicles.) This probability is 0.999999927. Next, we calculate the probability of this occurring for all 21 officers. The probability of it occurring for the first officer is 0.999999927; for the second the same; and for both the first and the second, 0.999999927 2 , and so on, and for all 21, 0.99999992721. This is calculated by the function POWER: POWER(BINOMDIST(61, 190, 0.19255, TRUE), 21) which equals 0.99999848. Then the probability of at least one officer out of 21 encountering 62 125002 is just the probability of all 21 officers not encountering more than 63 12500s in 190 stops or 1-

26 0.99999848 = 0.99999270.000001528. Inverting, this is one chance in 654,270. The entire formula is 1-POWER(BINOMDIST(61,190,0.1679,TRUE),21) In the case of testing the possibility of a low 12500 stop rate, we are interested in the probability of one in 21 officers encountering a number of 12500s equal to or less than the number encountered on the stops. We begin by calculating this for a single officer (in this case, we use Officer 3’s stats): BINOMDIST(8, 116, 0.1679, TRUE) or 0.001456884. The probability of a single officer failing to encounter as few as 8 12500 drivers in 116 stops is 1-0.001456884 = 0.99854312. The calculation is: 1-BINOMDIST(8,116,0.1679,TRUE). But the failure has to occur 21 times to insure that no officer gets as few as 8 12500s in 116 stops. This is calculated by POWER((1BINOMDIST(8,116,0.1679,TRUE) 21)) and the probability associated with at least one officer getting as few as 8 12500s in 116 stops is 1 minus this number, or 0.030152914. These probabilities use the mean rate of 12500s found on Directed Patrols, 16.8%. The probabilities of various outcomes assumes the validity of that mean, which is over ten times higher than the mean percentage of 12500s found at checkpoints. The very low probabilities associated with this artificially high mean demonstrates the the likelihood of targeting is very high even if the percentage of 12500 drivers committing traffic offenses at night relative to the total number of offenses is as high as 16.8%--which seems highly unlikely. We also calculated various 12500 stop-rate probabilities using the mean percentage of 12500 drivers derived from checkpoint data, 1.3% (actually 0.012557). These reveal the four highest 12500 officers to be even more extreme outliers, with odds of the 12500 stop rates they produced to be less than 1 in 100 trillion. The analysis also shows that the four lowest 12500 officers produced results that would be expected of at least one of 21 officers given a 1.3% frequency of 12500 drivers on the road at night.

27

Appendix III Discussion of Legal Issues Relating to Impoundment
The principal abuse uncovered in Santa Barbara is lack of legal justification for many of the traffic stops that result in a 30-day car impoundment. The Fourth Amendment prohibition on unwarranted searches and seizures has been interpreted by the courts as requiring that police have “reasonable suspicion” to believe that a traffic law has been violated or other criminal conduct is occurring to justify a traffic stop. (United States v. Lopez-Soto (9th Cir. 2000) 205 F.3d 1101, 1104-05). In addition, CVC 14607.6 states that an officer “shall not stop a vehicle for the sole reason of determining whether the driver is properly licensed.” If State and Federal law were being followed in Santa Barbara, we would expect relatively few 12500 impoundments to result from traffic patrols, which legally can stop drivers and issue a citation only when there is evidence that a traffic law (or other law) has been violated. A major concern of the Santa Barbara Police Department has been driving under the influence (DUI). DUI drivers may be encountered at sobriety checkpoints, when they are involved in an accident, or when they violate a traffic law by careless or reckless driving. (The only arbitrary stops permitted by law are those occurring at “carefully regulated “sobriety checkpoints, permitted by a 6-3 decision of United States Supreme Court in 1993. The justices struggled with Fourth Amendment issues, but allowed sobriety checkpoints as an exception due to the “serious public danger of drunk driving.”) Careless driving that gives evidence of impairment, according to the National Highway Traffic Safety Administration Manual, falls under four categories: “problems in maintaining proper lane position, speed and braking problems, vigilance problems, and judgment problems.” When this driving behavior is observed, it is standard practice throughout the country for the driver to be cited for careless driving or some related moving offense, followed by administering field sobriety test or other means of determining whether he or she is legally DUI. However, in the City of Santa Barbara, report sheets generated during patrols show no correlation (in the case of Directed Patrols there is a negative correlation) between citations for moving or equipment violations or field sobriety tests administered and citations for driving without a license (12500). A careful analysis of statistics generated from the reports provides strong evidence that many cars are being stopped without evidence that a traffic law is being broken or other criminal conduct is occurring. And further statistical analysis establishes that these stops are the result of targeting immigrant drivers. History of CVC 14602.6 During the late 1980s and early 1990s, a national campaign to crack down on drunk driving resulted in the passage, in 1994, of legislation known as the Safe Streets Act. CVC 14602.6 is part of this legislation. Its intention was to stop those convicted of drunk driving and similarly serious offenses from driving after their licenses had been suspended or revoked. But also referred to drivers who have never been licensed. Here is the wording of 14602.6(a)(1):

28 14602.6. (a) (1) Whenever a peace officer determines that a person was driving a vehicle while his or her driving privilege was suspended or revoked, driving a vehicle while his or her driving privilege is restricted pursuant to Section 13352 or 23575 and the vehicle is not equipped with a functioning, certified interlock device, or driving a vehicle without ever having been issued a driver’s license, the peace officer may either immediately arrest that person and cause the removal and seizure of that vehicle or, if the vehicle is involved in a traffic collision, cause the removal and seizure of the vehicle ( )1 without the necessity of arresting the person in accordance with Chapter 10 (commencing with Section 22650) of Division 11. A vehicle so impounded shall be impounded for 30 days.

The reason drivers who have never been issued a driver’s license were included was because state law had spelled out conditions under which the state could refuse a license (CVC 12806): alcoholism or drug addiction; “lapses of consciousness,” and physical or mental disabilities that could affect the safe operation of a motor vehicle.” But during this same period a radical change in the pool of unlicensed drivers occurred. Fourth Amendment Considerations Regarding Impoundment CVC 14602.6 makes impoundment discretionary (an interpretation endorsed by the State Attorney General in the Maywood report). When a 12500 citation is issued, and the driver has no record of a previous serious traffic offense, and the car can be (or has been) safely parked, there is no justification for towing and 30-day impoundment. These actions, executed automatically, impose great hardship of members of the immigrant community, and poison community relations. Of course, if a vehicle constitutes a safety hazard, the police are authorized to remove it under the community caretaking doctrine. A careful legal discussion of the relevant issues constitutional issues is found in an opinion issued by the Legislative Counsel of California (Driver’s License: Vehicle Impoundment #0711154) dated May 29, 2007. The key conclusion is this: Based on the reasoning and holding of the court in Williams, the lack of any discussion of the community caretaking function is Salcero, the facts and discussion of justification in Benites, and the recent federal cases of Coccia and Miranda, we conclude that the community caretaking justification must be met in order for a peace office to validly exercise his or her statutory authority to impound a vehicle after citing the driver for being an unlicensed driver, when there is no basis for a seizure of the vehicle under the Fourth Amendment to the United States Constitution. After the Attorney General’s investigation of the Maywood Police Department, the department implemented the report’s recommendations and according to the Maywood Police Department, officers “are trained that to the maximum extent possible, that they exercise not to impound; and to park the car instead or to have a licensed driver take it.”

29 There is no reason the Santa Barbara Police Department could not adopt a similar policy which is in accord with opinions of both the Legislative Counsel and the State Attorney General. At this time (July 2010), the constitutionality of CVC 14602 itself is being challenged with the case of Salazar v City of Maywood et al before the 9th Circuit Court of Appeals. (The plaintiff’s brief is found at http://www.lccr.com/6%2024%2009%20Salazar%20v.%20City%20of%20Maywood%2 0amicus%20brief.pdf . 14602.6: Meaning of “without ever having been issued a driver’s license” The Santa Barbara Police Department impounds cars for 30-days citing the authority of CVC 14602.6. However, 14602.6 does not apply to unlicensed drivers who had ever been issued a driver’s license. It was written to explicitly exclude those who showed driving competence through having obtained a valid license in the past. Thus it clearly does not apply to undocumented immigrants who previously had driver’s licenses. This is a large group that includes those who had licenses before 1994 when undocumented immigrants were denied renewals, and also to those who entered the United States with a valid visa, obtained licenses, but could not renew them because their visas had expired. There is no evidence that the SBPD recognizes that the law used to justify 30-day impoundments (14602.6) does not apply to these drivers. In contrast, the City of Los Angeles recognizes that “an expired California license prevents 30-day seizure. If the driver has an expired California license, no matter how old, the vehicle is not subject to 30-day seizure.” In fact, the Attorney General’s report on Maywood severely criticized the department for towing and impounding cars of unlicensed drivers who had valid foreign licenses or valid or expired out-of-state licenses.

30

Appendix IV The Myth of the “Dangerous Unlicensed Driver”
Some categories of validly licensed drivers have demonstrably higher accident rates; for example, drivers over age 75 or under age 20. It is possible, that immigrants have more accidents, but despite the widespread perception that this is true, there is no evidence to back this opinion. No doubt that some persons are denied drivers licenses because the state have evidence that these persons, for reasons of physical or mental incapacity, may be potentially dangerous. Before 1994, such persons (along with underage drivers) formed the overwhelming majority of never-licensed drivers, and there is a study which confirms that this group does seem to have more accidents. But the group of unlicensed drivers is very different today than at that time. The unlicensed driver group is now dominated by undocumented immigrants not by these deemed to be potentially dangerous. Until 1993, immigrants whether or not legally in this country, could qualify for a California Driver’s license by passing the usual tests. In 1993, voter backlash against undocumented immigration led to passage of a law denying driver’s licenses o those who could not prove legal residence in the United States. As the licenses of undocumented immigrants expired, and as new immigrants arrived in the state, the numbers of unlicensed drivers rapidly increased. The legislative findings in support of the Safe Streets Act are found in CVC 14607.4. The principal justifications for impoundment were given as “of all drivers involved in fatal accidents, more than 20 percent are not licensed to drive” and “a driver with a suspended license is four times as likely to be involved in a fatal accident as a properly licensed driver.” The DMV website contains a 1997 report, “Unlicensed Driving: A Major California Safety Problem,” in which it is stated that 8.8% of drivers at sampled locations had suspended or revoked licenses and 3.3% had no record of any driver’s license.” However, the estimated percentages are taken from information gathered during the years 1987-1992—years when undocumented immigrants were eligible for drivers licenses. The statistics thus indicate that during 1987-1992, 73% of those driving without valid licenses had had their licenses suspended or revoked. The 17% that had no record of ever having been issued a driver’s license consisted of persons who could not obtain a license due to physical or mental impairments, to minors, or to persons who lacked a sense of social responsibility. We would expect such persons to be substandard, perhaps dangerous, drivers. We can call this group the “bad 12500s.” During the subsequent 15 years, as the driver’s licenses of undocumented immigrants expired, and as new immigrants ineligible for licenses arrived in California, the pool of unlicensed drivers (relative to drivers whose licenses had been suspended or revoked) sharply increased. We can get an idea of how much by looking at statistics from stops at checkpoints (since these approximate a random sample of cars being driven at the time).

31 PUEBLO was given data on 37 checkpoints involving 12,652 vehicles conducted in the city from 1/20/07 through 7/19/09. These provide a good estimate of the percentage of 14601 (revoked or suspended) drivers versus 12500 (unlicensed drivers) during the past few years. Checkpoint statistics reveal that only 25% of towed cars involve 14601 versus 75% for 12500. Thus the percentage of 14601s in the pool has decreased from 73% to 25% over the period since undocumented immigrants were denied licenses. Increased police attention to drunk driving and tougher suspension/revocation standards imposed by the Safe Streets legislation (as well as defining DUI at a lower blood alcohol level and increasing sobriety checkpoints) undoubtedly increased the pool of 14601s which makes the drop in the relative percentage of 14601s even more striking. Simple arithmetic suggests a four to five times increase in the size of the unlicensed driver pool, with undocumented immigrants now forming the overwhelming majority of 12500 drivers; the old “bad 12500” group may now be as low as 5%, with undocumented immigrants forming 95% of the 12500s. Given this radical change in the composition of the unlicensed driver pool since 1993 when undocumented immigrants were denied the right to a new license, it is both incorrect and unjust to cite accident statistics for the “bad 12500” unlicensed drivers from 1987-1992 to justify the impoundment of cars being driven by undocumented immigrants. There is simply no evidence that undocumented immigrant drivers constitute any particular hazard, and certainly nothing to equate the quality of their driving with that of those whose licenses have been revoked or suspended for drunk driving. Police departments have also tried to justify the importance of impounding immigrants’ cars by citing a 1997 study that claims that unlicensed drivers are dangerous and that impoundment is an effective deterrent (David J. DeYoung, “An Evaluation of the Specific Deterrent Effect of Vehicle Impoundment on Suspended, Revoked and Unlicensed Drivers in California.) Although DeYoung mentions the two groups (DWUdriving while unlicensed, and DWS—driving while suspended or revoked), he never discusses differences between the two. His principal conclusion is that impoundment under the new 1994 law is effective because DWU and DWS drivers as a group had higher rates of accidents and traffic convictions in 1994 (before the new law went into effect) than did drivers whose cars were impounded in 1995 after the law was in effect. There are three problems with this conclusion: 1. With 20% of drivers’ licenses expiring every year, there would have been a large increase in unlicensed immigrant drivers no longer able to renew their licenses; in addition, it has been estimated that hundreds of thousands of new undocumented immigrants arrived in the state during this period and these persons were ineligible for driver’s licenses regardless of their driving ability. Hence, if unlicensed immigrant drivers were relatively good drivers, the new, larger 1995 unlicensed driver pool would be expected to have a lower accident and infraction rate than the 1994 pool that only consisted of those who lacked licenses for reasons unrelated to immigration status (suspended or revoked drivers and “bad 12500s”).

32 2. It is quite possible that the impoundment of vehicles driven by suspended or revoked drivers actually did deter drinking while driving; if so, the law accomplished its goal of reducing accidents and convictions through its effect on potential drunk drivers—even if it had no effect on unlicensed undocumented immigrant drivers. The effect on the driving habits of potential suspended and revoked (14501) offenders alone can explain the improvement in accident and conviction rates from 1994 to 1995. 3. Finally, the study fails to control for other variables which may have changed from 1994 to 1995; a single correlation is never sufficient to indicate a causal relation. There is no attempt in DeYoung’s 33-page report to separate the effects of impoundment on 12500 offenders from those on 14601 offenders even though it is clear from the discussion that these statistics were readily available. The conflation of these two very different groups is apparent from DeYoung’s use of the misleading label “disqualified drivers” for both types of drivers. DeYoung and Michael Gebers conducted another study for the DMV in 2002 (“An Examination of the Characteristics and Traffic Risk of Drivers Suspended/Revoked for Different Reasons”). The new study concluded that the revoked and suspended drivers in themselves form a heterogeneous group with very different accident rates. They state: “This law [14602.6] and others, should be modified to better reflect the nature and risk of the suspended/revoked offenders to whom they apply…. However, there are even more serious problems with the current suspension/revocation laws, as the findings on the risks of drivers suspended/revoked for failing to pay child support point out. This group, which is suspended/revoked for reasons completely unrelated to their driving, has the lowest crash risk of any suspended/revoked group, and poses little more risk on the highways than the validly licensed drivers.” They go on to recommend, “The current vehicle impoundment law, CVC 14602.6, should be rewritten to more rationally reflect the risks of the suspended/revoked drivers it includes and excludes from its provisions.” It is strange that the authors fail to apply exactly the same reasoning to unlicensed drivers. DeYoung and Gebers never consider or even mention statistics for 12500 (unlicensed) drivers. Again, there is absolutely no evidence that undocumented immigrants are dangerous drivers. Because unlicensed drivers fear being stopped by police patrols and then having their vehicle impounded, they are generally cautious drivers who strive not to provide any reason for a stop. In contrast, there is substantial evidence that drivers over age 80 or under age 18 are potentially dangerous but no one suggests that they should be automatically barred from obtaining licenses. If, as the DMV states in its 2008 Strategic Plan, its goal is “licensing drivers to protect consumers and promote traffic safety” then all persons driving vehicles in the state should not only be required to have a driver’s license but should be eligible to obtain one. As long as that is not the case, it is irresponsible for authorities to speak about the danger of “the unlicensed driver” based on a stereotype supported by statistics dating from a period when undocumented immigrants were eligible for licenses. 30-day impoundment

33 should be reserved (as noted by DeYoung and Gebers) for truly dangerous drivers such as those whose licenses were suspended or revoked for drunk driving.