Bonilla v. City of Los Angeles
Bonilla v. City of Los Angeles
2016 WL 11507319 (C.D. Cal. 2016)
September 23, 2016
Segal, Suzanne H., United States Magistrate Judge
Summary
The court denied the plaintiff's motion to compel the production of the LAPD's K-9 and Use of Force databases in electronic format, finding that the production of the electronic databases would be disproportionate to the needs of the case. However, the court granted the motion to the extent that the plaintiff was entitled to hard copy documents for the two-year period from November 14, 2013 to December 15, 2015. The court also denied the proposed redactions of certain documents and denied the plaintiff's request for sanctions.
Vincent BONILLA, Plaintiff,
v.
CITY OF LOS ANGELES, et al., Defendants
v.
CITY OF LOS ANGELES, et al., Defendants
Case No. CV 15-3932 CMB (SS)
United States District Court, C.D. California
Signed September 23, 2016
Counsel
Donald Webster Cook, Donald W. Cook Attorney at Law, Los Angeles, CA, Peter Goldstein, Law Office of Peter Goldstein, Culver City, CA, for Plaintiff.Christian R. Bojorquez, Los Angeles City Attorney's Office, Los Angeles, CA, for Defendants.
Segal, Suzanne H., United States Magistrate Judge
ORDER: (1) GRANTING IN PART AND DENYING IN PART PLAINTIFF’S MOTION TO COMPEL FURTHER RESPONSES, ETC.; AND (2) DENYING REQUEST FOR SANCTIONS (Dkt. No. 66)
I.
INTRODUCTION
*1 On June 14, 2016, Plaintiff Vincent Bonilla filed a Motion to Compel Further Responses to Requests for Production and Request for Sanctions against Defendant City of Los Angeles. (“Motion” or “MTC,” Dkt. No. 66). The Parties filed a Joint Stipulation in connection with the Motion pursuant to Local Rule 37-2 (“Jt. Stip.”), including the declarations of counsel Donald W. Cook (“Donald Cook Decl.,” id. at 31-36) and applications programmer analyst Dwight W. Cook (“Cook Decl. I,” id. at 37-43)[1] in support of the Motion, and, in opposition to the Motion, the declarations of counsel Christian R. Bojorquez (“Bojorquez Decl.,” id. at 44-50), Los Angeles Police Department (“LAPD”) detective Diane Escalante (“Escalante Decl.,” id. at 51-52), LAPD programmer analyst Terry Soo Hoo (“Soo Hoo Decl.,” id. at 53-55), LAPD programmer analyst Gene Stolarov (“Stolarov Decl.,” id. at 59-61), and LAPD senior systems analyst Gerardo Limtao (“Limtao Decl.,” id. at 56-58). Plaintiff seeks production of certain sets of data or “items” from the LAPD and sanctions of “at least $24,245.00.”[2] (MTC at 2; Jt. Stip. at 4). On June 16, 2016, Plaintiff filed a “Reply Declaration of Dwight W. Cook,” apparently in lieu of a supplemental memorandum. (“Cook Decl. II,” Dkt. No. 68).
*2 Also on June 16, 2016, the Court held a telephonic hearing on an ex parte application related to the Motion, during which the Parties addressed certain substantive issues presented by the Motion. The Court concluded that an evidentiary hearing was necessary to resolve several factual disputes.
In anticipation of the Evidentiary Hearing, the Parties submitted declarations in lieu of direct testimony for witness Dwight W. Cook on behalf of Plaintiff (“Cook Hrg. Decl.”) and, on behalf of Defendant, for witnesses Diane Escalante, Terry Soo Hoo, Gene Stolarov, and Gerardo Limtao, all of whom had submitted declarations with the Joint Stipulation.[3] (Dkt. Nos. 76-77). The Court held the Evidentiary Hearing on August 18, 2016, at which all of the aforementioned witnesses testified except Escalante. (See Hearing Transcript (“Tr.”), Dkt. No. 83). In addition, LAPD Sergeant James McDonald, who did not submit a declaration, also testified on behalf of Defendant. (Id. at 133-43). At the end of the Evidentiary Hearing, the Court announced that it would hold a follow up telephonic hearing to provide the Parties an opportunity to make closing arguments and address any outstanding issues.
In anticipation of that telephonic hearing, on August 30, 2016, Defendant lodged for in camera review copies of several sample documents that it proposed would be less burdensome to produce than the electronic data sought by Plaintiff. (See Notice of Lodging, Dkt. No. 85). Similarly, on August 31, 2016, Plaintiff filed a brief captioned “Plaintiff’s Position re: Production of Use of Force and Canine Data.” (“P Position Br.,” Dkt. No. 86). The Court held the telephonic hearing on September 1, 2016.
For the reasons stated below and on the record, Plaintiff’s Motion to Compel is GRANTED IN PART and DENIED IN PART. Plaintiff’s Request for Sanctions is DENIED.
II.
ALLEGATIONS OF THE FIRST AMENDED COMPLAINT
In the operative First Amended Complaint (“FAC”), Plaintiff alleges that he was unlawfully attacked by an LAPD “K-9 unit”[4] (i.e., police dog) even though he was unarmed and was ultimately not charged with a crime. (SeeFAC, Dkt. No. 23, at 6-7). According to Plaintiff, between 11:00 and 11:30 p.m. on November 14, 2014, he was walking down an alley with “another individual” in an area of Los Angeles that had “a history of street crime” when he noticed that a car without its headlights on was following them. (Id.at 6). Not knowing that it was a police car, Plaintiff turned into another alley and went into the back yard of one of the residences. (Id.). Plaintiff eventually realized that “it was the police, but fearful of being arrested for evading a police officer, he hid in a shed in the backyard of the home where he was ultimately found.” (Id.).
*3 A K-9 unit entered the shed where Plaintiff was hiding, left without barking, and returned with four LAPD police officers. Upon the officers’ commands, the K-9 unit dog bit Plaintiff multiple times in his upper and lower arm, causing “profuse bleeding.” (Id.). The officers drew their guns and ordered Plaintiff to “crawl out of the shed on his hand and knees with his arms still bleeding.” (Id. at 6).
According to Plaintiff, after being dragged and handcuffed, he was taken to the Los Angeles County/USC Medical Center, where he was admitted and treated for his injuries until his discharge eight days later, on November 22, 2014. Because Plaintiff is a hemophiliac, he “lost half of his blood as a result of the injuries he sustained from the K-9 attack.” (Id. at 6). After Plaintiff was released from the hospital, he appeared in court and was informed that the District Attorney’s Office declined to file charges against him. (Id. at 7).
Plaintiff filed the instant action on May 25, 2015. Following Plaintiff’s dismissal of his claims against the individual officers on June 15, 2016, the only surviving Defendant is the City of Los Angeles. (See Dkt. No. 67). Plaintiff is asserting a Monell claim[5] against the City for its allegedly unconstitutional custom or policy of using excessive force through the improper deployment of its K-9 unit. (FAC at 10-13). Plaintiff seeks compensatory and punitive damages, but does not seek declaratory or injunctive relief.[6]
III.
THE DATA SETS AT ISSUE AND THE PARTIES’ CONTENTIONS
Plaintiff seeks the production of three outstanding sets of data generated and maintained by the LAPD. These items are:
1. LAPD K-9 database
2. LAPD Arrest Reports associated with K-9 deployments
3. LAPD Use of Force database
(Jt. Stip. at 4).[7] Plaintiff stated at the Evidentiary Hearing that he is not interested in a “paper document” production of the information in the K-9 and Use of Force databases because he wants “electronic data” that he “can analyze and [use to] do [his] own compilations.” (Tr. 167).
*4 The Joint Stipulation merely lists these data sets without limitation, which would, as Defendant asserts, seemingly require the production of “essentially every piece of data ever collected regarding every [u]se of [f]orce [that] any LAPD officers [have] ever been in and every canine encounter that has ever occurred.” (Jt. Stip. at 2). However, Defendant represents that after “significant discussion between counsel,” Plaintiff limited the production of these data sets to a “six-year period”[8] and, with respect to the Use of Force database (but not the K-9 database), agreed to permit redaction of personal identifying information of non-officers, such as the social security numbers of suspects. (Id.). Furthermore, Plaintiff is seeking production of only thirteen of the thirty-eight tables in the Use of Force database. (See Cook Decl. I ¶ 8).
Plaintiff contends that the information sought is “essential to [his] case” as it will supply the foundation for him to “show that the dog attack [against him] was pursuant to City custom, policy or practice, and/or ratified by the City of Los Angeles.” (Id. at 10) (emphasis in original). In particular, Plaintiff asserts that the K-9 database will enable him to “establish that LAPD police dog deployments result in a use of force rate far higher than other LAPD uses of force” and that “the frequency that LAPD K9 deployments cause injury, including serious injury ... is far higher than any use of force save the firearm.” (Id.) (emphasis in original). Plaintiff anticipates that the K-9 database “will also show that the K9 deployments are used in instances no more dangerous and probably less so ... than instances of other suspects where dogs are not used” and that “the use of force rates of dog handlers are much higher than the use of force rates of regular LAPD officers.” (Id.). Plaintiff further states that a comparison of the K-9 Deployment Reports and K-9 Arrest Reports will allow him to “determine more accurate (and probably higher) bite rates than reported by the K9 unit.” (Id. at 11).
Plaintiff states that information in the Use of Force database will permit him “to compare use of force and injury rates between LAPD officers who do not use police dogs, and those that do.” (Id. at 24). According to Plaintiff, the Use of Force database and the Force Investigation Division database, the latter of which has already been produced, “provide detailed data on the circumstances under which the various LAPD force options were used, thus enabling a more accurate ‘apples to apples’ comparison of comparable incidents where LAPD officers use force options other than dogs.” (Id.). Plaintiff states that the data will show that K-9 unit officers are “far more likely to use force, to use force that injures, and to use force that seriouslyinjures, when the officers deploy an LAPD dog.” (Id.) (emphasis in original).
Defendant argues that the data requested is irrelevant, contains privileged information protected from disclosure by the official information privilege and the right to privacy, and would be unduly burdensome and costly to produce. (Id. at 2). In sum, Defendant argues that “[t]his is a basic dog bite case and Plaintiff is turning [discovery] into a fishing expedition” to search for information well beyond the needs of this case. (Id. at 11).
IV.
SUMMARY OF TESTIMONY AND OTHER EVIDENCE
A. Defendant’s Evidence
1. K-9 Database
a. Diane Escalanate
*5 Escalante is a Detective II assigned to the LAPD Legal Affairs Division. (Escalante Decl. ¶ 1). A K-9 “Deployment Report” is prepared whenever a K-9 unit is called out to assist in locating a suspect. A K-9 “Contact Report” is prepared “when a K-9 bites during a K-9 deployment.” (Id. ¶ 2). The K-9 database contains 1,852 individual Deployment Reports for the period between November 15, 2009 and November 14, 2014, reflecting all deployments, whether with or without bites. (Id. ¶ 2). In addition to the “Deployment Report” information stored electronically in the K-9 database, K-9 platoon members create hard copy files that contain additional information such as photos; photo CDs; arrest reports; statements by victims, witnesses and arrestees; field interview cards; arrestee confidential medical information; and use of force information. (Id. ¶ 3). The aggregate number of physical pages in hard copy K-9 files for a five-year period “has a potential to exceed approximately 74,000 pages.” (Id.). Each hard copy file is maintained in a large file envelope that is boxed, stored, and archived by year. (Id. ¶ 4).
b. James McDonald
McDonald has been a supervisor in the LAPD K-9 unit for approximately one year and has been assigned to the K-9 unit since approximately 1996 or 1997. (Tr. 134). There are approximately 400 K-9 deployments per year, “minus SWAT searches and things of that nature.” (Tr. 135). Most deployments do not result in anyone getting bitten or even arrested. (Id.).
K-9 Deployment Reports are created from information entered into the database by the dog handler involved in the incident. (Id.; see also Jt. Stip., Exh. D). K-9 Contact Reports are created from information entered by a supervisor. (Tr. 135; see also Jt. Stip., Exh. E). Deployment Reports may identify a named suspect or witness, if any, by first and last name, date of birth, address, telephone number, driver’s license number, or similar identifiers. (Tr. 139). The information collected in the K-9 database for K-9 Deployment and Contact Reports, including a suspect or witness’s personal identifying information, is “more or less[ ] the same data that used to be recorded on a paper, like a Noncategorical Use of Force report.” (Tr. 137-38 & 140). K-9 officers do not input any other data into the K-9 database except for the information contained in Deployment Reports and Contact Reports. (Tr. 135).
Statistical reports are generated from the information collected in Deployment and Contact Reports. (Id. at 136). “The Statistical Reports depict everything from breakdown of race, time that we would have our busiest call load, to statistics specific to every dog that we have in terms of what their find ratio is, how many times they’ve been deployed, how many times they’ve had contact with a suspect.” (Tr. 141-42). Drop down menus on the system would permit the creation of a report that would show, for example, how many incidents in a given time period involving a dog resulted in a bite. (Tr. 142). The LAPD’s Information Technology Division (“ITD”) would know the range of information that could be generated in reports from the drop down menus. (Tr. 143).
c. Terry Soo Hoo
Soo Hoo is an ITD Programmer Analyst IV assigned to the K-9 database. (Soo Hoo Decl. ¶ 1). Soo Hoo prepared a list of all of the K-9 data fields, consisting of the K-9 deployment, K-9 team, and K-9 contact tables. (Tr. 126; see also Jt. Stip. Exh. C). Every time a dog is deployed in the field, a unique number is assigned to that deployment. (Tr. 128). The first two digits of the deployment number indicate the year, followed by sequential numbering of the deployments for that year. (Id.).
Soo Hoo migrated the K-9 database from the old Oracle (an object-relational database management system) database to an “incident tracking system” in a SQL[9] database sometime in 2014 or 2015. (Tr. 129). From the end-user perspective, entry of K-9 data on the computer screen looks the same in the new database as in the old database. (Id.). The K-9 reports generated by the new database also look “close” to the reports generated by the old database. (Id.). To generate K-9 reports in the new database, “programming code has to identify which fields concern the relevant K-9 data element to the exclusion of other fields.” (Id. at 130). In addition, the code must link to another table called “Ink Prop,” which is itself “linked to the main incident table by the incident number.” (Id.). Soo Hoo wrote the programming code to generate K-9 reports in the new database that “match up” with the K-9 reports generated in the old database. (Tr. 131). It took “months” for Soo Hoo to write that code, “almost a year.” (Tr. 132). However, the work to generate a “canned” report in Microsoft Access,[10] the program in which Plaintiff wants data produced, would not take that long because the work has already been done. (Id.). Nonetheless, Soo Hoo testified that “the coding is done for displays and the report. It’s not done for looking at data. That’s a different effort.” (Id.).
d. Gene Stolarov
*6 Stolarov is an ITD Programmer Analyst V. (Stolarov Decl. ¶ 1). Stolarov admitted that the K-9 database is not “of any significant size” -- perhaps no more than 200-300 megabytes -- and affirmed that “size is not the consideration” in the LAPD’s objection to producing it. (Tr. 104-05). Similarly, copying raw data from the K-9 database to another SQL server like the one on which it is housed is also neither difficult nor particularly time-consuming. However, the data would not be in an intelligible format without a “labor-intensive” field-by-field conversion process that, coupled with the time required to copy the data after the “field” programming is finished, could take up to a full work week to complete. (Tr. 97).
The K-9 database was created a “long time ago” using Oracle tools and an Oracle database, but the data was moved within the last two years to a larger, more comprehensive “incident tracking system, which is a collection of over 3,000 modules” containing information about all sorts of cases, not just those involving the K-9 unit. (Tr. 93). To incorporate K-9 data into the incident tracking system, the names of the various K-9 data fields and their “values” (i.e., the information in those fields) were translated into the common “language” of the incident tracking system “to accommodate this generic collection of multiple modules that it’s now a part of.” (Tr. 97). For example, the field for “victim ethnicity” in the old K-9 database might be designated simply as the number “3” in the new system, and the code or value for “Asian” in that field might appear as the number “1.” (Tr. 97). In other words, instead of producing data as “victim ethnicity -- Asian,” the raw data would simply read “field 3 -- code 1.”
As Stolarov explained, simply copying raw data from the K-9 database to a server “only makes sense if you’re willing to receive fields called Field 1. Field 2. Field 3.” (Tr. 117). He continued,
Nobody argues that it’s not possible. The argument is to provide the mapping from Field 1 to whatever the field is, deployment number or race of the victim. If you want to see fields -- instead of Field 1, if you want to see the field “victim ethnicity” instead of the value for victim ethnicity, you want to see, not “1,” but “Asian,” that takes time. [¶] If you’re willing to receive the data as it exists right now, “Field 1” with a value of “3,” then we don’t need an external file. We can basically truncate all the nonrelevant data and give you what is there right now. However, what you’ll get is the collection with illegible [sic] field names and field values.
(Tr. 117) (emphasis and some punctuation added).
A “data dictionary” must convert these “generic” codes into intelligible words. (Tr. 98). “Canned” reports already incorporate the data dictionary conversions for each field and value so that the resulting report is intelligible to the reader. (Tr. 117). However, any specialized report or request would require the LAPD to “build a new query” that “does all this aggregation and conversion” to “present data in the terms that [are] understandable to the requestor.” (Id.).
Stolarov emphasized that creating the conversion mechanism to translate “field” and “value” designation numbers is not difficult, but is “labor intensive.” Stolarov estimates that setting up the conversion would take “five to ten minutes for each field.” (Tr. 118). Because “136 fields” must be converted in the K-9 database, the conversion process could take one-and-a-half to three days,[11] “plus another day or two of integration validation, moving data back and forth, exporting, stuff like that.” (Tr. 100). Stolarov estimates that the total time required for the creation of the conversion mechanism and the copying and validation of the data for the K-9 database would therefore be “about one man week.” (Id.; see also Stolarov Decl. ¶¶ 7 (estimating that producing the K-9 database would require three working days of programmer time and approximately two additional days “for troubleshooting, integration and logistics”)). Stolarov further estimates that the total cost for the production would be approximately $3,300.00, “which equates to a rate of $55.00 per hour, plus the time and one half (1.5) for overtime.” (Id. ¶ 8).
2. Use Of Force Database
*7 Gerardo Limtao is a Senior Systems Analyst II with the LAPD’s Application & Database Services Section. (Limtao Decl. ¶ 1). Limtao was Defendant’s only witness to testify about the LAPD’s Use of Force database. (Id. ¶ 2).
The Use of Force database is part of the LAPD’s “Teams II” Oracle database.[12] The Teams II database, which includes “several subsystems [and] databases,” is a “transactional database” into which data is continuously being added; it is not a “warehouse.”[13] (Tr. 33). The Use of Force data takes up about 700 gigabytes; the Teams II database in its entirety is necessarily larger. (Tr. 34).
Ad hoc queries are not run against the Teams II database because they would slow it down. (Tr. 36). The LAPD maintains a “data warehouse” outside of the Teams II database that is optimized for running reports or ad hoc queries.[14] (Tr. 35). However, the warehouse does not contain a complete copy of all the Use of Force data in the Teams II database. (Tr. 35). The warehouse is updated nightly with “whatever Use of Force data ... that gets copied into [it],” but Limtao does not know exactly “to what extent all the Use of Force data that is in the transactional database also ends up in the data warehouse.” (Tr. 74). “Canned” reports are designed to pull up data from the data warehouse, not the transactional database. (Tr. 40). The Business Intelligence section has a list of all the “canned” reports that can be generated from the data warehouse. (Tr. 77). Although it took Limtao only two hours to generate a list of the Use of Force tables in the Teams II database, it would take “much longer” to generate a list of the Use of Force tables in the data warehouse because the warehouse is structured differently. (Tr. 76).
Limtao emphasized that LAPD policy generally prohibits IT from querying or copying the transactional database because doing so would affect its performance. (Tr. 42). However, if the LAPD were ordered to produce a complete set of Use of Force data, the “easiest way” to “make a copy of the Use of Force data out of the transactional database” would be “to make a straight copy of all the Oracle Use of Force data as Oracle data onto an external drive.” (Tr. 78). While it would take only approximately six hours to “export” the Use of Force data and “move” or copy it “to the regular disk connected to the [Use of Force] server,” that server does not have “a lot of extra storage space.” (Tr. 40, 44). Therefore, to make a copy of the full set of data, Limtao would likely need to “write it to a map network device.”[15](Tr. 40). This would involve an “export dump” into some kind of outside storage, such as a hard disk or a “USB-compliant device,” which is “very slow.” (Tr. 37-38, 41). The process of exporting and copying data to an external device could take approximately two days, perhaps even “a little bit more.”[16] (Tr. 78).
*8 There can be errors in exporting the data. (Id. 45). If an error occurs, the entire export process must be repeated. (Tr. 46). Once the export is completed, the next step would be to import the data into another server with Oracle database software already installed, a process that Limtao estimates would take “at least 24 hours.” (Tr. 47).
After the import is completed, the next step in the production would be to extract the thirteen tables of Use of Force data requested by Plaintiff.[17] (Tr. 48). Queries for the data parameters would need to be run for each table. (Id.). However, most of the tables do not have dates, so Limtao would first need to “join the other tables that don’t have dates with the main table” to be able to limit the extraction to the correct time period. (Tr. 48-49). The incident date is one column out of approximately 100 columns in the Use of Force main case table, so to extract data from other columns for a given time period, Limtao would need to match up each of those columns with the date column. (Tr. 49-50). Limtao estimates that the table extraction process for a five-year period of data, with redactions where necessary, would require approximately “three or four full days of work,” beyond his normal duties. (Tr. 51).
After the thirteen tables of data are extracted, they would have to be copied into a Microsoft Access file or database. (Tr. 52). Microsoft Access is not normally used in Limtao’s section because unlike Oracle or SQL, Microsoft Access is not an “enterprise” database capable of handling large amounts of data. (Tr. 37, 52-53). Because Limtao does not normally work with Microsoft Access, he can only estimate how long it would take to copy the data from Oracle into Microsoft Access, but estimates that it would take anywhere from a day or two to one week. (Tr. 64).
In his declaration, Limtao estimates that to collect the data Plaintiff requested would require a total of two weeks of work (80 hours),[18] at an estimated cost of $7,500.00 at a rate of $63.00 per hour plus time and a half. (Limtao Decl. ¶ 7). Limtao further declares that converting the data from Oracle to Microsoft Access would entail an additional week (40 hours) of work at the same hourly rates, for a combined final cost of $11,250.00. (Id.). These estimates “do not factor in the additional amount of time that would be required based upon the already existing workload and resource availability that [the LAPD] would be able to dedicate to such a project.”(Id.).
3. Sample Hardcopy Documents And “Canned” Reports
Defendant lodged (for in camera review) sample documents reflecting “what Defendant represented to the Court could be produced as hard copies in lieu of the requested databases which Counsel for Plaintiffs was seeking.” (Notice of Lodging at 1). Defendant contended that while the review, redaction and production of even these hard copy documents would “constitute an overly burdensome, costly and time-consuming process,” it would be less burdensome to produce them than the K-9 and Use of Force databases that Plaintiff seeks in the instant Motion. (Id. at 1-2). The sample documents submitted are: (1) a “hardcopy K-9 package,” which includes copies of a K-9 Contact Report, Deployment Report and Arrest Report in addition to other items related to a specific incident; (2) the LAPD publication “2015 Use of Force Year-End Review” and an accompanying “Executive Summary” booklet summarizing the same, both of which Defendant represents are publicly available; (3) printouts of statistical and other information “automatically gathered via canned reports”; and, as the last item on Defendant’s list of “canned reports,” (4) a “Use of Force Report.”[19] (See Notice of Lodging, Dkt. No. 85). The “canned” reports are:
*9 1. Categorical Investigation Recap
2. Individual Use of Force and Other Categorical Investigation Records Report
3. Less Lethal Device Incident Listing Report
4. Non-Categorical UOF 14-Day Compliance Report
5. Parameters for Pending Cases Report
6. Police Commission Weekly Report
7. Internal Due Date Report
8. Force Investigation Division Case Status Report
9. Case Workflow Status Report
10. Categorical Case Investigations by Day
11. Categorical Case Investigations for Date Range
12. Non-Categorical Use of Force Cases by Suspect Descent Report
13. Non-Categorical Use of Force Cases by Day
14. Non-Categorical Use of Force Cases for Date Range
15. Non-OIS [“Officer-Involved Shooting”] Categorical Case Statistics
16. OIS Case Statistics
17. Non-Cat Use of Force Statistical Summary
18. Use of Force Status Report by Investigator
19. Use of Force Case Summary Report/UOFS Review Report (“Use of Force Report”)
Defendant submitted unredacted copies of the sample K-9 package, the Use of Force Report, and one of the other “canned” reports (“Individual Use of Force and Other Categorical Investigation Records Report”). However, it highlighted proposed redactions of purportedly “sensitive” information in these documents in the event that the Court orders production of such materials. As noted earlier, Plaintiff has agreed to permit redaction in Use of Force Reports of personal identifying information of non-officers, such as the social security numbers of suspects, but has not agreed to any redaction of any information in K-9 Reports.
B. Plaintiff’s Evidence
1. Donald Cook
Plaintiff’s counsel Donald Cook states in his declaration that he was counsel for plaintiffs in Lawson v. Gates, Los Angeles Superior Court Case No. 0312323, which was a civil action for damages and injunctive relief concerning the LAPD K-9 unit that settled in the 1990s. (Donald Cook Decl. ¶ 2A). In that case, the LAPD produced hard copies of over 12,000 LAPD K-9 unit reports and 400 Arrest Reports, as well as “the entire LAPD Use of Force and Officer-Involved-Shooting (“OIS”) databases” without restrictions or redactions. (Id. ¶ 2B-C).
This data was critical because “by matching use of force and OIS data to incidents comparable to those in which LAPD deployed dogs,” the Lawsonplaintiffs were able to determine that the use of force rates for dog handlers was up to fifty times greater than the use of force rates for officers not using dogs. (Id. ¶ 3). Furthermore, the data showed that “[t]he only time the dog handlers’ use of force rate dropped significantly (practically to zero) is when handlers believed the suspect had, in his possession, a gun or knife. That is, when handlers believed that the suspects really were dangerous, the dogs were rarely allowed to bite.” (Id.). The data also showed that the frequency with which dog handlers opt to use force and the likelihood of serious injury when LAPD dogs bite is much greater than for other uses of force. (Id. ¶¶ 4-5). Finally, a comparison of Deployment Reports and Arrest Reports permitted plaintiffs to determine that K-9 officers exaggerated the number of times dogs were used for non-violent purposes. (Id. ¶ 6).
2. Dwight W. Cook
*10 Dwight W. Cook is an applications programmer analyst for the state of Minnesota.[20] (Cook Hrg. Decl. ¶ 1). As an outside expert, Cook admitted at the Evidentiary Hearing that he has not personally extracted any information from the LAPD’s K-9 or Use of Force databases. (Tr. 155).
a. K-9 Database
Even though K-9 data was migrated from an Oracle database to a SQL server in the last year or two, Cook maintains that “it is still easy (almost absurdly so) to isolate, select, copy and export only K9 related data” from the SQL server where the K-9 database is housed. (Id.). Far from the two to three working days that Stolarov estimated it would take to create a “conversion” program for all of the K-9 database fields, Cook states that it would take him “a minute or two to create queries for [the fields in] each table,” (Tr. 145), which would require a total of “no more than a few hours, at most.”[21] (Cook Hrg. Decl ¶ 12).
According to Cook, to produce the K-9 database, the programmer would first establish an ODBC (open database connectivity) connection between the local computer and the K-9 database SQL server, then “simply” build a table in Microsoft Access “to include all K9 deployment records in the [requisite] date range,” which would be saved as a Microsoft Access file. (Id.¶ 12). Cook argues that “anyone managing the program ... can easilyidentify which fields correspond to K9 data fields ....” (Id. ¶ 13) (emphasis in original). The K-9 data entry screens generate the “same” K-9 Deployment and Contact Reports as the old Oracle system. Therefore, the input screens “necessarily cross-reference to the K9 data field that matches to the labeled box the end user sees on the screen when he or she inputs data.” (Id. ¶ 13A). Because the cross-referencing has already been done, the programmer would simply use “the existing coding to identify, extract and copy the desired data.” (Id. ¶ 19).
To see the coding for the K-9 related data fields, the programmer “simply opens a copy of the K9 input screen in design view,” a process that “takes but seconds.” (Id.) (emphasis in original). Doing so would show the source data for any selected field, which would in turn take only a “few seconds” to copy. (Id. ¶ 13C). Cook estimates that “it should not take more than a couple of hours” to copy “the existing queries for all 136 K9 data fields.” (Id.¶ 13D).
Once that information is copied, it would take only a “few minutes” to “construct the queries so that when executed, they select, copy and then export data from the K9 data fields to an ODBC compliant data file (such as Access).” (Id.). Cook states that it would take him only a half day at most to prepare the program for copying and exporting K-9 data into a Microsoft Access file, and that it should take Soo Hoo even less time given his familiarity with the database program. (Id ¶ 14). Cook argues that “the basic problem” with Soo Hoo’s and Stolarov’s testimony is that they ignore that the cross-reference coding for each of the K-9 fields has already been created, and need only be copied from the system’s design view. (Id. ¶ 16).
*11 Cook did not directly challenge Stolarov’s contention that once the programming was created, the “actual extracting, copying” would take approximately twenty-four hours. However, Cook stated that instead of taking “three working days,” the “actual copying ... can run continuously.” (Tr. 146). Cook further claimed that errors in the copying process are rare -- he states that they occur in “less than one percent of [his] copying operations” -- and when they occur, they do not significantly affect the amount of time required to complete the task overall. (Cook Hrg. Decl. ¶ 14).
b. Use of Force Database
The majority of Cook’s testimony with respect to the Use of Force database, both in his declarations and in his live testimony at the Evidentiary Hearing, was based on his original belief that the Use of Force data was just a “very small subset” of the TEAMS II data -- “probably less than a gigabyte and maybe only a few hundred megabytes.” (Id. ¶ 7). Cook repeatedly insisted that the “700 gigabytes” estimate that Limtao had given at his deposition encompassed the entire Teams II database, not the subset of Use of Force data within that database.[22] (Id. ¶ 9A; see also Tr. 8). However, after listening to Limtao’s testimony at the Evidentiary Hearing, Cook admitted that his assumption was wrong and that the Use of Force database is 700 gigabytes. (Tr. 156). Nonetheless, he also testified that even if the Use of Force database were larger than he originally believed, the “copy process is the same” and would take about the same “[l]ength of time.” (Tr. 144).
The Use of Force database consists of 38 tables, “with each table consisting of a collection of data fields.” (Cook Hrg. Decl. ¶ 8). Cook asserted that, similar to the K-9 database, the first step to copying the Use of Force database would be to create a link between the Teams II database and a Microsoft Access file via an ODBC connection, which he estimates would take approximately 15 minutes. (Id. ¶ 8A & n. 2). Then, using Microsoft Access, the programmer would select the thirteen tables requested by Plaintiff and import them into Microsoft Access, which Cook estimates would take an hour. (Id. ¶ 8B). The program would automatically save the data as a Microsoft Access data file, which could then be moved to a USB drive for production. (Id.). Finally, Cook states that it would take just a few additional minutes to limit the data by date range, which could be executed with just “a few commands.” (Id. ¶ 8C). Cook estimates that the total time for the entire copying and transfer process would take “no more than an hour or so, if that.” (Id. ¶ 9) (emphasis in original).
The primary flaw in Limtao’s testimony, according to Cook, is that Limtao fails or refuses to “acknowledge that one obtains a copy of the [Use of Force] data via an ODBC link” between the Teams II database and Microsoft Access. (Id. ¶ 10). Cook insists that “[t]he whole purpose behind ODBC is to permit access to data regardless of the [database management system] in which the data was originally created and stored, without the need for any specialized programming or effort to convert the data into a readable format by another [database management system].” (Id. ¶ 10A).
*12 According to Cook, it is no more difficult to “pull” data from a transactional database than from a warehouse. (Tr. 15). Nonetheless, Plaintiff states that even if the Use of Force data from the database warehouse does not include all of the data in the transactional database, Plaintiff “will accept a production of U/F data obtained from the database warehouse and not the transactional database.” (P Position Br. at 4).
V.
STANDARDS
A. Scope Of Permissible Discovery
Federal Rule of Civil Procedure 26(b)(1), as amended on December 1, 2015, provides:
Parties may obtain discovery regarding any nonprivileged matter that is relevant to any party’s claim or defense and proportional to the needs of the case, considering the importance of the issues at stake in the action, the amount in controversy, the parties’ relative access to relevant information, the parties’ resources, the importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit. Information within this scope of discovery need not be admissible in evidence to be discoverable.
Fed. R. Civ. P. 26(b)(1). Evidence is relevant if “it has any tendency to make a fact more or less probable than it would be without the evidence, and the fact is of consequence in determining the action.” Fed. R. Evid. 401. “Proportionality focuses on the marginal utility of the discovery sought.” In re Methyl Tertiary Butyl Ether (“MTBE”) Prod. Liab. Litig., ––– F. Supp. 3d ––––, 2016 WL 1367226, at *4 n.43 (S.D. N.Y. Apr. 4, 2016) (internal quotation marks and citation omitted).
The revisions to Rule 26 make clear that the right to discovery, even plainly relevant discovery, is not limitless. As one court explained,
The 2015 amendments to Rule 26(b)(1) emphasize the need to impose “reasonable limits on discovery through increased reliance on the common-sense concept of proportionality.” The fundamental principle of amended Rule 26(b)(1) is “that lawyers must size and shape their discovery requests to the requisites of a case.” The pretrial process must provide parties with efficient access to what is needed to prove a claim or defense, but eliminate unnecessary or wasteful discovery. This requires active involvement of federal judges to make decisions regarding the scope of discovery.
Roberts v. Clark Cnty. Sch. Dist., 312 F.R.D. 594, 603 (D. Nev. Jan. 11, 2016) (quoting John Roberts, 2015 Year-End Report on the Federal Judiciary 6 (Dec. 31, 2015) (internal citations omitted)); see also Fed. R. Civ. P. 26 advisory committee notes (2015 amendments) (“The parties and the court have a collective responsibility to consider the proportionality of all discovery and consider it in resolving discovery disputes.”).
The December 1, 2015 amendments to Rule 26(b)(1) “were designed to protect against over-discovery and to emphasize judicial management of the discovery process, especially for those cases in which the parties do not themselves effectively manage discovery.” Noble Roman’s, Inc. v. Hattenhauer Distrib. Co., 314 F.R.D. 304, 308 (S.D. Ind. 2016). An additional purpose was “to contribute to the overall goal of regulating the time and expense of litigation.” Davita HealthCare Partners, Inc. v. United States, 125 Fed. Cl. 394, 398 (2016).
On a motion to compel, it is the burden of the moving party “to show that the material it requests is relevant to its claims.” First Niagara Risk Mgmt., Inc. v. Folino, ––– F.R.D. ––––, 2016 WL 4247654, at *4 (E.D. Pa. Aug. 11, 2016). If that burden is met, “the opposing party must show that the [proportionality] factors in Rule 26 weigh in favor of [the court] denying [the moving party’s] request for otherwise relevant information.” Id.
B. Official Information Privilege
*13 “Federal common law recognizes a qualified privilege for official information” including, but not limited to, government personnel files and other data. Sanchez v. City of Santa Ana, 936 F.2d 1027, 1033 (9th Cir. 1990); Rogers v. Giurbino, 288 F.R.D. 469, 481 (S.D. Cal. 2012) (same). To determine whether the information sought is privileged, “courts conduct a case by case balancing analysis, in which the interests of the party seeking discovery are weighed against the interests of the governmental entity asserting the privilege.” Soto v. City of Concord, 162 F.R.D. 603, 612 (N.D. Cal. 1995) (citing Sanchez, 936 F.2d at 1033-34). Several courts have noted that this balancing analysis raises special considerations in civil rights suits against police departments. See e.g., Soto, 162 F.R.D. at 612; Kelly v. City of San Jose, 114 F.R.D. 653, 660 (N.D. Cal. 1987). Specifically, in weighing the competing interests of the parties, courts should consider that “the policies that inform federal civil rights laws are profoundly important” and that “confidence in our system of justice ... is threatened when relevant evidence is not made available ....” Id. Furthermore, “in civil rights cases against police departments in general, it is not likely that plaintiffs can obtain information of comparable quality from any other source.” Soto, 162 F.R.D. at 616. Accordingly, in such cases, the balancing is “moderately pre-weighted in favor of disclosure.” Kelly, 114 F.R.D. at 661; Rogers, 288 F.R.D. at 481 (quoting same).
Before the court engages in the balancing analysis, however, the party asserting the official information privilege must make a “substantial threshold showing.” Kelly, 114 F.R.D. at 669. To make the threshold showing, a defendant must submit a declaration to the court from the head of the department having control over the records that includes each of the following:
(1) an affirmation that the agency generated or collected the material in issue and has maintained its confidentiality; (2) a statement that the official has personally reviewed the material in question; (3) a specific identification of the governmental or privacy interests that would be threatened by disclosure of the material to plaintiff and/or his lawyer; (4) a description of how disclosure subject to a carefully crafted protective order would create a substantial risk of harm to significant governmental or privacy interests, and (5) a projection of how much harm would be done to the threatened interests if disclosure were made.
Chism v. County of San Bernardino, 159 F.R.D. 531, 533 (C.D. Cal. 1994). In addition, “[a] strong affidavit would also describe how the plaintiff could acquire information of equivalent value from other sources without undue economic burden.” Soto, 162 F.R.D. at 613.
C. Privacy
“Federal Courts ordinarily recognize a constitutionally-based right of privacy that can be raised in response to discovery requests.” Id. at 616 (citing, inter alia, Breed v. United States Dist. Ct. for Northern District, 542 F.2d 1114, 1116 (9th Cir. 1976)). “The right to privacy is not absolute and can be outweighed; courts generally balance the need for the information against the severity of the invasion of privacy.” Rogers, 288 F.R.D. at 484 (citing Ragge v. MCA/Universal Studios, 165 F.R.D. 601, 604–05 (C.D. Cal. 1995)); see also E.E.O.C. v. California Psychiatric Transitions, 258 F.R.D. 391, 395 (E.D. Cal. 2009) (“[T]he right to privacy is not a recognized privilege or absolute bar to discovery, but instead is subject to the balancing of needs.”); Soto, 162 F.R.D. at 616 (“Resolution of a privacy objection or request for a protective order requires a balancing of the need for the information sought against the privacy right asserted.”). While courts have recognized that privacy concerns raised by the disclosure of police files are “not inconsequential,” these privacy interests “must be balanced against the great weight afforded to federal law in civil rights cases against police departments.” Id. (quoting Kelly, 114 F.R.D. at 660).
VI.
DISCUSSION
A. The Electronic Discovery Requested Is Disproportionate To The Needs Of This Case
*14 The three surviving data sets at issue are (1) the K-9 database (including, but not limited to, the “narrative portions” of K-9 Deployment Reports), (2) Arrest Reports associated with K-9 Deployment Reports, and (3) the Use of Force database. The exact temporal scope of the production requests is presently unclear, as the Parties have represented that it ranges from a six-year period (from November 14, 2009 through December 15, 2015), to a shorter period of two to five years, so long as it is “large enough ... to show that this is what typically happens” when K-9 units are deployed. (Tr. 162). The Court concludes that the difficulty and cost of producing the K-9 database and the Use of Force database would be disproportionate to the needs of this case because Plaintiff is able test his allegations with comparable discovery that is less cumbersome and expensive to produce. However, as explained in subpart VI.B below, the Court also concludes that Plaintiff is entitled to hard copy productions of the reports and other statistical information for which Defendant lodged samples for in camerareview with only limited redaction of personal identifying information. Production of these materials shall be limited to the two-year period beginning on November 14, 2013 and ending on December 15, 2015.
With respect to the evidence presented by the Parties through declarations and live testimony, the Court finds that Defendant’s witnesses, who have a comprehensive understanding of the complexities and limitations of the LAPD’s systems, credibly explained why production of the data requested by Plaintiff is more challenging and time-consuming than suggested by Plaintiff’s expert. Defendant’s witnesses’ descriptions of the steps required to produce the requested data in the format demanded by Plaintiff were informed, trustworthy, and persuasive.
In contrast, Plaintiff’s expert, Cook, lacked the degree of familiarity with the data and relevant systems that Defendants’ witnesses possessed. In addition, Cook appeared to over-simplify the challenges to preparing the LAPD’s data for production and minimized the procedures, time and cost that the production would entail. Cook’s inconsistent statements on the size of the Use of Force database, as well as his last-minute attempt to dismiss the importance of database’s size once he was forced to admit his error, did not enhance his credibility as to the ease of using a “non-enterprise” platform, Microsoft Access, to capture data from the Use of Force database, or of the time required to cross-reference codes in the K-9 database. Accordingly, the Court affords greater weight to the declarations and testimony of Defendant’s witnesses.
As noted earlier, Rule 26(b)(1), as revised, provides that a party may obtain discovery that is not privileged, relevant, and “proportional to the needs of the case.” The Rule sets forth seven factors a court should consider in evaluating the proportionality of the discovery sought, which include considerations of the scope of the materials to be produced as well as the cost and relative burden the production would impose. In the context of the needs of this particular case -- a suit brought by a single plaintiff with estimated damages of no more than $100,000 -- production of the LAPD’s electronic databases, as demanded by Plaintiff, would provide Plaintiff with discovery in excess of what he needs to support this particular case and place a disproportionate financial and staffing burden on Defendant. Principles of proportionality heavily favor the use of alternative -- less cumbersome and less expensive -- means of providing Plaintiff with the information he needs.
1. Importance Of The Issues At Stake In The Action
Plaintiff alleges that he was the victim of an alleged LAPD policy that promotes the deployment of its K-9 unit in unwarranted circumstances, even though the Department knows that serious harm is likely to result. Allegations that the LAPD’s K-9 unit shows “dangerous propensities for abusing [the City’s] authority and for mistreating citizens” raise important issues of civic concern. (FAC ¶ 40). The harm inflicted on Plaintiff was far from trivial, and, if caused by an unconstitutional policy or practice, renders this a significant case. Accordingly, the issues at stake in this action are important, both to Plaintiff and to the public, and support discovery extensive enough to enable the Parties to compare the use of the K-9 unit in Plaintiff’s case to other K-9 deployments and other uses of force.
2. The Amount In Controversy
*15 The FAC does not specify the amount of monetary damages at issue. However, Plaintiff’s counsel has acknowledged that the upper range of damages that he would anticipate if this matter goes to trial would be in the “high five figures or low six figures.” While an upper range of $100,000.00 does not qualify as nominal damages, it does not justify any discovery plan regardless of the cost of production. Although Plaintiff disputes the amount of time and effort that would be required to produce the K-9 and Use of Force databases, Defendant’s witnesses credibly explained that the process of collecting and preparing the data would be quite burdensome and cost approximately $15,000.00. This amount is based only on the employees’ hourly rates and does not take into account the cost of disrupting the LAPD’s normal operations or the opportunity cost of diverting skilled manpower and resources from other assignments. This action involves one plaintiff -- it is not a class action -- and does not seek declaratory or injunctive relief. As such, the damages are largely confined to Plaintiff’s injuries, and Plaintiff admits that he has not suffered any permanent damages. The relatively limited amount in controversy therefore suggests that discovery should be confined to the most cost-effective means of proving the existence (or non-existence) of an unconstitutional policy involving K-9 units.
3. The Parties’ Relative Access To Relevant Information
The information that Plaintiff seeks is solely within Defendant’s custody, and it does not appear that Plaintiff could obtain comparable information from any other source. Accordingly, the Parties’ relative access to relevant information weighs strongly in favor of disclosure. However, this factor does not necessarily require that Defendant provide all possible relevant information in the specific form requested by Plaintiff.
4. The Parties’ Resources
Plaintiff has not submitted evidence of his finances with this Motion. However, even assuming that Plaintiff’s resources are limited, Defendant is a governmental entity that must allocate its limited resources among many competing responsibilities and priorities. Accordingly, while the Court will assume that Defendant’s resources are greater than Plaintiff’s, Plaintiff’s relative lack of resources does not require that Defendant incur unreasonable costs in providing the discovery necessary to litigate this case.
5. The Importance Of The Discovery In Resolving The Issues
This factor requires the Court to weigh the importance of the specific discovery at issue in resolving the Parties’ claims and defenses. Plaintiff seeks this information to support his allegations, including raw data and statistical compilations that would permit Plaintiff to quantify his assertions about K-9 deployments and the alleged policy underlying them. However, Plaintiff has not convinced the Court that the electronic data he seeks is necessary, or even the most efficient way, for him to prove up his case. Plaintiff might prefer data to be produced in a certain form, i.e., electronic data. However, the form in which information is produced is a separate consideration from the importance of its substance, unless the form severely impairs the requesting party’s ability to understand the substance.
That is not the case here. The K-9 “packages” and other “canned” reports that Defendant has stated a willingness to produce (if ordered) will provide Plaintiff with an abundance of highly relevant information that he can use, if he chooses, to conduct statistical analyses. Plaintiff has not shown that electronic data is critical to the prosecution of his case. Accordingly, this factor, i.e., the importance of the discovery sought in resolving the issues, does not weigh in favor of requiring an electronic production.
6. Whether The Burden Or Expense Of The Proposed Discovery Outweighs Its Likely Benefit
Defendant has clearly established that the production of the data Plaintiff seeks, in the format requested by Plaintiff, would involve a considerable outlay of money and time. Furthermore, it would require the diversion of highly skilled employees from other tasks critical to the LAPD’s public safety mission. Those challenges would not, by themselves or collectively, justify a categorical bar on the production of the LAPD’s electronic data in all cases. The government cannot shield itself from disclosing information in the proper case simply because it houses its data in a system that makes retrieval of the data in a particular format costly and difficult to produce.
*16 However, the Rule 26 factors, as applied to the requested discovery in this specific case, lead the Court to conclude that an electronic production of the information requested is not required. A paper production of K-9 “packages” and “canned” reports will provide Plaintiff with the substantive information he seeks while permitting Defendant both to avoid the technical difficulty and expense of producing data in an electronic format and to assign less specialized (and less expensive) employees to collect and redact the information to be produced.
In sum, the electronic data Plaintiff seeks is disproportionate to the needs of this case, both in its scope and in the burden and expense such a production would impose. Plaintiff’s injuries, while serious, were not permanent and even counsel admits that damages are unlikely to exceed the low six figures, if that. Plaintiff does not even seek an injunction that would restrict Defendant’s actions in the future. These considerations lead to the conclusion that while Plaintiff should be permitted to obtain substantial information about K-9 deployments and other uses of force, Plaintiff should not be allowed to compel Defendant to create an exhaustive and expensive production of electronic data. Comparable, in-depth information exists that is less cumbersome and less expensive to produce, and that appears to be sufficient for the needs of this case. Accordingly, to the extent that Plaintiff’s Motion seeks production of the LAPD’s K-9 and Use of Force databases, the Motion is DENIED.
B. Plaintiff Is Entitled To The Requested Documents With Limited Redactions And Timeframe Restrictions
While the production of the electronic databases sought in this case would be disproportionate to the needs of the case, Plaintiff is entitled to comparable information in other formats. Specifically, the Court finds that Plaintiff is entitled to the hard copy documents for the two-year period from November 14, 2013 to December 15, 2015 of “K-9 packages” (including Contact Reports, Deployment Reports, and Arrest Reports, and other materials maintained in the hard copy files, except as exempted below), annual “Use of Force Year-End Review” and related “Executive Summary” publications (for the same time period), and all of the “canned” reports for which samples were submitted for in camera review, including Use of Force Reports (for the same time period). Accordingly, to the extent that Plaintiff’s Motion seeks production of information relating to K-9 deployments and the Use of Force, the Motion is GRANTED, with the time limitations described above. To the extent Plaintiff’s Motion seeks production of this information in a particular electronic format, the Motion is DENIED.
The Court further concludes that Defendant’s proposed redactions of certain of these documents are excessive and unwarranted. Defendant has not made the required “threshold showing” that the information it proposes to redact is protected from disclosure by the official information privilege. There is no declaration from the relevant LAPD department head showing that the information is confidential, describing the specific harm that would result from its disclosure, and explaining why a carefully crafted protective order would be insufficient to protect any governmental or privacy interests at stake.[23] See Chism, 159 F.R.D. at 533.
The Court finds, however, that Plaintiff’s need for certain personalidentifying information is outweighed by the invasion of privacy that would result from the disclosure of this information. Plaintiff has not shown that the specific identities of the officers, supervisors, suspects or witnesses reflected in Defendant’s data is at all relevant, much less useful or critical, to the case he is trying to make. Plaintiff’s ability to establish the existence of an LAPD policy to misuse K-9 deployments does not depend on the specific identity of supervisors or officers, so long as their titles or positions are clear, or on the identities of third parties such as suspects, witnesses, or health care providers. Similarly, Plaintiff has not shown the relevance of the addresses, telephone numbers, vehicle identification or license plate numbers, etc. of these individuals. Accordingly, Defendant may redact personal identifying information as proposed to the extent that the documents contain personal identifying information such as first and last names of LAPD employees and third parties (but not, with respect to LAPD employees, their job titles, the K-9 unit to which they are assigned, or the division or department they work for, and, with respect to third parties, their job titles or honorifics such as Ms., Mr. Mrs., etc.); their birthdays or birth places; their residential or business street addresses (but not the city or state in which the street is located); and their telephone, social security, driver’s license, vehicle identification, booking or badge numbers.
*17 In contrast, a suspect’s race, age, height and weight, physical “oddities” or marks (such as tattoos), and the location where an incident occurred, are not, by themselves or even in combination, particularly revealing or sensitive, and may inform Plaintiff’s analysis of the circumstances under which the LAPD deploys its K-9 units. Similarly, photographs of a suspect showing the extent of the suspect’s injury may be relevant to Plaintiff’s contentions. Accordingly, those fields may not be redacted, except that a suspect’s face in any photographs may be blurred or blacked out.
Additionally, narratives describing the circumstances and events of a K-9 deployment or other use of force are relevant to Plaintiff’s claims, particularly to the extent that they reveal discrepancies between a handler and a supervisor’s descriptions of the events, or a supervisor’s ratification of a particular deployment. Similarly, information revealing whether the LAPD considered the deployment of a K-9 unit in a particular case to have been “in policy” or granted “administrative approval” for a use of force is relevant to Plaintiff’s attempt to establish the existence of a policy. Accordingly, except to the extent that the following report sections contain personal identifying information as described above, Defendant’s proposed redactions to the following report sections are DENIED and Defendant must produce these sections in unredacted form:
Contact Reports: narratives captioned “supervisory insight,” “chief trainer insight,” and “officer in charge insight”;
Individual Use of Force and Other Categorical Investigation Records Reports: columns for “drawings,” “tactics,” and “use of force”;
Use of Force Reports: narratives and sections identified as “investigative narrative comments,” “medical treatment descriptions,” “watch commanders insight,” “actions taken” (including “work history review comments” and “finding review comments”), “chrono log,” “routing notes,” and “suspect oral statement.”
Defendant shall produce the documents required by this Order no later than fourteen days from the date of this Order. If additional time is necessary, the Court encourages the parties to submit a Stipulation and Proposed Order, extending the time for production.
C. Sanctions Are Not Warranted
Rule 37 provides in relevant part:
If the [discovery motion] is granted -- or if the disclosure or requested discovery is provided after the motion was filed -- the court must, after giving an opportunity to be heard, require the party ... whose conduct necessitated the motion, the party or attorney advising the conduct, or both to pay the movant’s reasonable expenses incurred in making the motion, including attorney’s fees.
Fed. R. Civ. P. 37(a)(5)(A). Conversely, if the discovery motion is denied, the court must require the movant, the attorney filing the motion, or both to pay the party who opposed the motion its reasonable expenses, including attorney’s fees, incurred in opposing the motion. Id. 37(a)(5)(B). Finally, if the motion is granted in part and denied in part, the court “may, after giving an opportunity to be heard, apportion the reasonable expenses for the motion.” Id. 37(a)(5)(B). However, if the non-prevailing party can demonstrate “substantial justification” for its motion, nondisclosure, or opposition, Rule 37 provides that the court must deny sanctions. Fed. R. Civ. P. 37(a)(5)(A)(ii).
Sanctions are not warranted here. Resolution of this dispute required extensive investigation into the LAPD’s data storage operations. The Parties’ respective positions were well supported and argued in good faith. Furthermore, the Court did not grant Plaintiff’s Motion in full. Accordingly, Plaintiff’s request for sanctions is DENIED.
VII.
CONCLUSION
*18 For the foregoing reasons, Plaintiff’s Motion to Compel is GRANTED IN PART and DENIED IN PART. Plaintiff’s request for sanctions is DENIED. Defendant shall produce hard copies of the documents identified above, with the redaction of only limited information as authorized by this Order, within fourteen days of the date of this Order. If additional time is required for the production, the Court encourages the parties to submit a Stipulation and Proposed Order for an extension of time to respond.
The Court will refer to the declaration submitted by Plaintiff’s counsel Donald Cook by the declarant’s first and last names. The Court will refer to the declarations and testimony provided by Plaintiff’s computer expert, Dwight W. Cook, by his last name only.
While the Joint Stipulation attempts to tie the production of the requested data sets to specific Requests for Production (“RFP”), the attempt is inconsistent and confusing. For example, the Notice of Motion states that twelve RFPs are in dispute. However, three of the “disputed” RFPs listed in the Notice -- Nos. 36, 37, and 38 -- are not mentioned anywhere in the Joint Stipulation. An additional four RFPs listed in the Notice -- Nos. 48, 50, 51, and 52 -- are not transcribed in the Joint Stipulation, as required by Local Rule 37-2, and are referenced only by number in a single introductory paragraph listing the data sets at issue, without any discussion of the substance of these RFPs or their connection to the data sets. (Jt. Stip. at 4). Only five of the “disputed” RFPs listed in the Notice of Motion -- Nos. 30, 31, 32, 49 and 52 -- are transcribed and specifically associated with specific data sets in the discussion section of the Joint Stipulation. (Jt. Stip. at 4-8; 21-24).
This confusion is ultimately irrelevant to the resolution of the instant Motion. While Defendant argues that the data sets contain information that is irrelevant to the claims and defenses in this action and that the production of the electronic databases in particular would be unduly burdensome, it does not argue that the requested data sets would not be responsive to one or more properly propounded RFPs. Accordingly, this Order will address the discoverability of the data sets directly, without attempting to link them to any particular RFP.
The declarations submitted by Escalante, Soo Hoo, Stolarov and Limtao in connection with the Evidentiary Hearing are identical to their declarations submitted in opposition to the Motion with the Joint Stipulation. Accordingly, the Court will not distinguish between the Joint Stipulation and Evidentiary Hearing declarations for these four witnesses. However, the Evidentiary Hearing declaration submitted by Cook is considerably more detailed than his declarations submitted in support of the Motion with the Joint Stipulation. The Court will therefore distinguish between Cook’s Joint Stipulation declarations (“Cook Decl. I & II”) and his Evidentiary Hearing declaration (“Cook Hrg. Decl.”).
Each of the Parties uses the spellings “K-9” and “K9” interchangeably. (See, e.g., Jt. Stip. at 4 (Plaintiff) & Exh. D at 88 (Defendant)). For the sake of consistency, the Court will use “K-9” except when directly quoting a source that uses “K9.”
The Ninth Circuit has explained:
In Monell v. Department of Social Services, 436 U.S. 658, 98 S.Ct. 2018, 56 L.Ed. 2d 611 (1978), the Supreme Court held that a municipality may not be held liable for a § 1983 violation under a theory of respondeat superior for the actions of its subordinates. In order to establish municipal liability, a plaintiff must show that a “policy or custom” led to the plaintiff’s injury. Id. at 694, 98 S.Ct. 2018. The Court has further required that the plaintiff demonstrate that the policy or custom of a municipality “reflects deliberate indifference to the constitutional rights of its inhabitants.” City of Canton v. Harris, 489 U.S. 378, 392, 109 S.Ct. 1197, 103 L.Ed. 2d 412 (1989).
Castro v. Cnty. of Los Angeles, ––– F.3d ––––, 2016 WL 4268955, at *9 (9th Cir. Aug. 15, 2016).
The copy of the FAC on the Court’s docket is missing the last page. (See Dkt. No. 23). The FAC ends abruptly with the first enumerated item in the prayer for relief and does not include a signature block at the end of the pleading. (Id. at 20). However, the original Complaint seeks only monetary (i.e., “special, general, and punitive”) damages, and Plaintiff has not indicated that the FAC altered the kinds of relief sought in any significant way. (See Complaint, Dkt. No. 1, at 17). Counsel for Plaintiff confirmed at the September 1, 2016 telephonic hearing that Plaintiff is not seeking injunctive relief.
Plaintiff’s Motion originally placed at issue five data sets. In addition to the three items above, the Joint Stipulation also requested production of the LAPD’s K-9 Deployment Reports and its Force Investigation Division (“FID”) database. (Jt. Stip. at 4).
Plaintiff admitted at the Evidentiary Hearing that because the information contained in K-9 Deployment Reports is stored in the K-9 database, the separate request for K-9 Deployment Reports was redundant and did not need to be addressed separately. (Tr. 160). At the same time, Plaintiff emphasized that the “most important field of information” in the Deployment Reports, however produced, is the “narrative portion” of the report that tells “the details of what happened.” (Tr. 167). Plaintiff further stated that he is still separately seeking “copies of the Arrest Reports.” (Tr. 161).
Plaintiff also informed the Court that Defendant has produced the entire FID database in response to a public records request. (Tr. 160). Therefore, the FID database is no longer at issue and will not be addressed here.
According to Defendant’s counsel, at an informal meeting of counsel on April 22, 2016, Plaintiff agreed to limit the relevant period to five years prior to the November 14, 2014 incident (i.e.November 14, 2009), through December 15, 2015. (Bojorquez Decl. ¶ 6). At the Evidentiary Hearing, Plaintiff stated that he was looking for a “large enough time period to show that this is what typically happens. This is the trend. [¶] From what I see in cases, it’s anywhere from two to three years to five years. I’m not seeing anything larger than five years.” (Tr. 162).
“SQL” stands for “Structured Query Language.” (Tr. 58).
Cook explained that Microsoft Access is “a popular off-the-shelf relational database program” that is “widely and commonly used by systems professionals.” (Cook Hrg. Decl. ¶ 8A).
136 fields x 5 minutes per field = 680 minutes, or 11.33 hours. If the conversion process took ten minutes per field, the time required would be 22.66 hours.
Cook confirmed in his Evidentiary Hearing declaration that unlike the K-9 database, which was migrated from an Oracle database to a SQL server, the LAPD’s Use of Force database is “recorded and managed in an Oracle dataset” on the Teams II database. (Cook Hrg. Decl. ¶ 6).
Cook explained at the Evidentiary Hearing that businesses generally maintain a “warehouse” to store electronic data separately from the “transactional” or “working” database in which the business enters, updates and deletes records on a day-to-day basis. (Tr. 10). New data, consisting only of activity in the transactional database since the last transfer, is typically transferred nightly to the warehouse. (Id.). The use of a warehouse permits the entity to generate reports from warehouse data without impairing the operation of the transactional database. (Id.). According to Cook, usually a warehouse will use a different server than the transactional database, but a single server could “serve multiple Oracle databases.” (Tr. 11). The same data may be stored simultaneously on a “transactional” database and in a “warehouse.” (Tr. 14-15).
Limtao stated that the LAPD also maintains a back up of the Teams II database for recovery purposes. (Tr. 65-66). However, queries cannot be run against the recovery back up. (Tr. 66). It is possible to extract data from the back up database only if it is restored to another server. (Id.).
Limtao testified that he does not know if there is a third-party application that would export Oracle data onto a SQL server. (Tr. 79).
It is not always clear whether Limtao’s references to “days” mean 8-hour working days or 24-hour calendar days. Elsewhere in his testimony, Limtao appeared to suggest that copying data to an external device would take an additional two to eight hours over the six hours required to export it, for a total of eight to fourteen hours to export and copy the data before importing it to another server. (Tr. 61). Because fourteen hours is approximately two working days, Limtao’s two-day estimate above appears to refer to working days.
Limtao testified that he does not know the amount of data contained in the thirteen tables and that it would take approximately two hours to calculate the amount. (Tr. 84).
This eighty-hour global estimate appears to be slightly longer than the more detailed estimate Limtao provided at the Evidentiary Hearing. As summarized above, Limtao testified that it would take approximately two days to export and copy Use of Force data from the Teams II database to an external device, at least twenty-four hours to import the data from the external device to another server, and three to four days to extract and redact the specific Use of Force tables requested by Plaintiff. (See generally Tr. 61-64).
The “Use of Force Report” is entitled “Use of Force Case Summary Report/UOFS Review Report.” The Use of Force Report is #19 on the Table of Contents (copied here) that Defendant submitted with the sample “canned” Reports.
Cook testified that he is Plaintiff’s counsel’s brother and that he was not paid for his expert testimony. (Tr. 153).
Cook explained elsewhere that the K-9 database consists of three tables (“K9_deply,” “K9_team,” and “K9_cntct”) encompassing 136 data fields. (Cook Hrg. Decl. ¶ 11).
Cook’s “reply” declaration submitted in response to the Joint Stipulation appears to concede that the Use of Force database contains 700 gigabytes of data. (See Cook Decl. II ¶ 3 (“If one is exporting the entire U/F [Use of Force] database, that is 700 gigabytes of U/F data, Mr. Limtao’s two day estimate is not unreasonable.”)). Cook appeared to retract that concession at the Evidentiary Hearing, as he initially testified that it was his understanding that the entire Teams II database was 700 gigabytes. (Tr. 8). However, Cook also admitted that he does not have evidence to show whether his understanding “is correct or not.” (Tr. 9).
A stipulated protective order was entered by the Court on April 26, 2016, but it does not contain the required showing. (Dkt. No. 60).