ACLU of Florida Analysis of Tampa CCTV Issue Jan 2014
Content
Greater Tampa Chapter ACLU of Florida Analysis of Tampa Police Department Surveillance Cameras
On Wednesday, August 29, 2013, City Council held a workshop on the retention of surveillance cameras procured for the Republican National Convention. Council members were briefed by TPD Captain Mike Baumaister who stated that the cameras were ―just another tool‖ but that they had contributed to reducing crime in downtown Tampa. He also provided Council a 13 page document with written answers to ―City Council Questions.‖ At that time, the Council did not elect to hear from individuals and groups that oppose the retention and continued funding of TPD’s surveillance camera system. Retaining the cameras and requisite software licenses required an appropriation of approximately $186,000, presumably a recurring annual expenditure, for maintenance and software. We have since had an opportunity to critically review TPD’s response to Council questions and its justification for retention. The ACLU of Florida continues to oppose retention of the RNC CCTV cameras because they are an unnecessary intrusion on the privacy of our citizens with minimal contributions to public safety. We believe that the Information provided by TPD does not justify retention of the cameras and the continued violation of our privacy. TPD’s answers do not constitute a detailed program evaluation of the cameras’ contributions to law enforcement and public safety. Do the Camera’s Make Downtown Tampa Safer? Captain Baumaister’s comments to the Council and the TPD’s response to Council questions asserted that the cameras contributed to the reduction of crime in the downtown area. To support that conclusion, TPD provided extracts from studies indicating significant reductions in crime in other cities (Baltimore, Chicago, and Washington D.C.) where CCTV cameras were deployed, evidence that the cameras contributed to a reduction in crime in Zone
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014 Page 1
F3 Downtown Area of Tampa, and cited a number of ―testimonials‖ of successes attributable to the CCTV cameras. Each of these conclusions is questionable. The Urban Institute Justice Policy Center’s study is one of many that address the effects of surveillance cameras on reducing crime. There is extensive academic literature on the subject—studies carried out over many years. That research strongly indicates that video surveillance has no statistically significant effect on crime rates. (1) IPVM, one of the world’s leading resources on video surveillance, recommends that proponents “stop claiming that CCTV can generally reduce crime.” They note that studies raise serious doubt about “CCTV’s ability to reduce crime generally, but they note that a strong consensus exists about CCTV’s ability to reduce premeditative/property crime.‖ This is consistent with the ACLU’s recognition that cameras are most effective when assisting in securing fixed facilities like government buildings/facilities, parking garages, and similar locations as opposed to general surveillance of public areas. (2) Crime rates are down in Zone F3 of Downtown Tampa, but it is not clear that the newly installed CCTV cameras played a role in that reduction. Since 2002, the TPD has enjoyed extraordinary success in reducing both the overall and the Big Four Crime Rates in the city. TPD’s ―Focus on Four Plan & 2013 Departmental Annual Report‖ states that crime in Tampa has been reduced 65% since 2002. (3) The report attributes success to proactive, preventative, and intelligence led policing. Of particular interest is the SAFECOP system that provides planners and officers on the street with computer displays of crime information that is used to allocate and deploy resources in response to emerging crime patterns. The report’s only mention of the role of surveillance cameras was to note that officers can use screen shots to quickly create ―crime bulletins.‖ However, it does not indicate if, how frequently, or how successfully this has been done. The numerically small and somewhat inconsistent crime reductions identified in Zone F3 since the installation of the cameras are most logically attributed to a continuation of effective policing methods that have, over a decade, led to significant city-wide annual crime reductions achieved without broad CCTV coverage. Section 2 of the TPD report on surveillance cameras which cites twenty-four “testimonials” to their contributions is also problematic. With minor exceptions,
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014 Page 2
cameras do not appear to have had a serious impact on resolving major crimes. In general, the language of TPD’s ―testimonials‖ does not indicate that cameras played a significant role. Most entries claimed that the cameras ―supported‖, ―assisted‖, ―helped investigate,‖ ―documented‖, ―freed-up manpower‖, ―manage‖ and ―monitor.‖ In only eight of the 24 testimonials do cameras appear to have played some role in the identification and/or arrest of criminal suspects? In most instances, the specific contribution of the cameras is not clear. In three instances, the presence of mobile and/or covert cameras is claimed to have deterred criminal activity – something police presence would have done just as well. Two claimed successes were attributed to FDOT traffic cameras - not the CCTVs TPD is attempting to justify. The testimonials clearly indicate that, in most instances, the general surveillance CCTV cameras were nice to have but were mostly used in a supporting role. The claimed impact on crime does not justify the continued intrusion on the privacy of our citizens or the continued expenditure of funds to maintain the system. Council might ask whether hiring three additional TPD officers would have had a greater impact on reducing crime in the city. TPD’s “testimonials” indicate that “covert cameras made greater contributions to public safety than the general surveillance CCTV cameras. The city has at least 129 cameras of various types, but eight of the 24 ―testimonials‖ addressed contributions by the ten ―covert‖ cameras. Unfortunately, the use of covert cameras has even greater potential for the violation of individual privacy and the abuse of civil liberties. Covert surveillance is particularly susceptible to abuse through racial and ethnic profiling. Council should, for example, inquire into the use of covert cameras to ―manage‖ crowds and ―trespassers‖ in public parks. Parks are set aside for the use and enjoyment of the community. From a civil liberties perspective, an overt police presence is preferable to covert surveillance cameras. An overt police presence deters bad behavior. Using covert cameras would appear to be focused on generating arrests – not deterrence. The use of covert cameras should be limited to the resolution of specific crime problems – not for surveillance of parks and other public areas.
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014
Page 3
If the Council is to approve additional funding (possibly $186K annually) to retain CCTV cameras, it should require that TPD provide rigorous annual program evaluations to answer several critical questions. How many times did police access stored imagery for investigative leads? How many times did cameras identify otherwise unknown suspects or contribute unique information? If cameras contributed to the resolution of a case, what specific roles did they play? Information provided by the TPD last year is not sufficient to justify continued invasions of our privacy and the additional costs involved. Comments on the Tampa Police Department’s CCTV Standard Operating Procedure (SOP) TPD’s CCTV SOP raises several issues that the Council should address. Para. V,C,1, as written, confuses the Criminal Intelligence Bureau’s issuance of a username and password with ―specific authorization‖ to access stored imagery. It appears that, following issuance of username/password, users may access stored imagery at will. The SOP does not define criteria for accessing stored imagery. IT logs users and presumably creates some sort of computerized record of who accessed the system; however, there does not appear to be a requirement for CIB supervisors to specifically approve accessing stored imagery. The SOP should prescribe criteria for accessing stored imagery and require specific approval to access the system. TPD should maintain logs reflecting each time stored images are accessed and the purpose, the specific probable cause or reasonable suspicion for doing so, and the result. Logs should be retained. Logs are critical to evaluating the system’s actual contributions and preventing abuses. SOP Para. V,D, addresses the use of AISight Video Analytics. From a civil liberties perspective, this is one of the system’s most troublesome capabilities. A1 Sight independently analyzes the activities, behaviors, and the scenes that it sees and identifies and alerts users to
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014 Page 4
―suspicious‖ activity (abnormality detection). The system supposedly ―learns‖ what is normal. The system identifies ―unusual‖ activities in the field of vision, categorizes them as suspicious, and the issues an alert. A man walking on a roadway late at night in an area where there is normally no pedestrian traffic could trigger in an alert. A Slate article identifies the problem: ―AISight and similar technologies transform citizens into suspects. Because AISight is used to monitor and detect not just acts of crime but potential acts of crime, based purely on a set of algorithms, it is considered part of the push towards pre-emptive—or ―pre-crime‖—policing, which treats everyone as a potential criminal and targets people for crimes they have not yet committed (and may never commit).‖ (4) The TPD SOP is correct to ensure that using the system requires consent of the CIB, but it is not totally clear who in the CIB has that authority. Council should review TPD’s use (frequency, why, when and for what purpose) of the A1Sight capability since it has been installed and actions taken based on the system’s alerts (if any). The Deployment of Mobile Trailers and Covert Cameras The list of ―testimonials‖ indicates that TPD has deployed both mobile trailer and covert cameras. Claims for successes of the mobile trailer cameras are quite modest. TPD’s response to Council indicates that they ―helped to support‖ crime control at city malls during the holiday season. Interestingly, the department’s annual report for 2012 discussed ―Operation Safe Shopper‖ in some detail without mentioning the use of cameras. The report did note that ―officers on horseback, motorcycles, bikes and unmarked vehicles saturated mall parking lots.‖ (2) The TPD report does not address the use of mobile cameras in residential areas ; however, this capability has exceptional potential to disproportionally target poor and minority neighborhoods. Council should request information concerning the location/purpose of all mobile camera deployments and results (if any) to identify any patterns in their deployment. .
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014
Page 5
TPD’s SOP does not establish criteria for using covert cameras. Council should inquire about the use of covert cameras to ―manage‖ crowds and trespassers in public parks and insist that there should be clear policy outlining when the use of covert cameras is appropriate/necessary. Covert capabilities create opportunities for racial and ethnic profiling. Parks are areas set aside for the use and enjoyment of the community. From a civil liberties perspective, a deterrent overt police presence is much to be preferred. We recognize that TPD may be reluctant to provide outsiders with information on their use of covert cameras, but if Council decides to fund retention of the system, it should exercise effective oversight of their use and ensure that the system is not abused. CCTV Surveillance Images are Public Records In deciding whether to continue funding the system, Council should recognize that CCTV images are public records. Council should request a periodic report on requests for stored images to determine how often, who and for what purpose CCTV images collected by the TPD are being used. We recommend that Council defund the CCTV surveillance system; however, if Council continues to fund system maintenance, it should insist on a rigorous annual program evaluation to determine if the cameras contribution to public safety is sufficient to justify the intrusions on our privacy. (1) (2) (3) (4) http://www.slate.com/blogs/future_tense/2012/06/11/aisight_from_brs_labs_and_other_tec hnologies_to_detect_suspicious_behavior_.html https://www.aclu.org/files/images/asset_upload_file708_35775.pdf http://ipvm.com/report/is_public_cctv_effective http://www.tampagov.net/dept_Police/Files/publications/2012_FOF_Web.pdf
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014
Page 6
On Wednesday, August 29, 2013, City Council held a workshop on the retention of surveillance cameras procured for the Republican National Convention. Council members were briefed by TPD Captain Mike Baumaister who stated that the cameras were ―just another tool‖ but that they had contributed to reducing crime in downtown Tampa. He also provided Council a 13 page document with written answers to ―City Council Questions.‖ At that time, the Council did not elect to hear from individuals and groups that oppose the retention and continued funding of TPD’s surveillance camera system. Retaining the cameras and requisite software licenses required an appropriation of approximately $186,000, presumably a recurring annual expenditure, for maintenance and software. We have since had an opportunity to critically review TPD’s response to Council questions and its justification for retention. The ACLU of Florida continues to oppose retention of the RNC CCTV cameras because they are an unnecessary intrusion on the privacy of our citizens with minimal contributions to public safety. We believe that the Information provided by TPD does not justify retention of the cameras and the continued violation of our privacy. TPD’s answers do not constitute a detailed program evaluation of the cameras’ contributions to law enforcement and public safety. Do the Camera’s Make Downtown Tampa Safer? Captain Baumaister’s comments to the Council and the TPD’s response to Council questions asserted that the cameras contributed to the reduction of crime in the downtown area. To support that conclusion, TPD provided extracts from studies indicating significant reductions in crime in other cities (Baltimore, Chicago, and Washington D.C.) where CCTV cameras were deployed, evidence that the cameras contributed to a reduction in crime in Zone
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014 Page 1
F3 Downtown Area of Tampa, and cited a number of ―testimonials‖ of successes attributable to the CCTV cameras. Each of these conclusions is questionable. The Urban Institute Justice Policy Center’s study is one of many that address the effects of surveillance cameras on reducing crime. There is extensive academic literature on the subject—studies carried out over many years. That research strongly indicates that video surveillance has no statistically significant effect on crime rates. (1) IPVM, one of the world’s leading resources on video surveillance, recommends that proponents “stop claiming that CCTV can generally reduce crime.” They note that studies raise serious doubt about “CCTV’s ability to reduce crime generally, but they note that a strong consensus exists about CCTV’s ability to reduce premeditative/property crime.‖ This is consistent with the ACLU’s recognition that cameras are most effective when assisting in securing fixed facilities like government buildings/facilities, parking garages, and similar locations as opposed to general surveillance of public areas. (2) Crime rates are down in Zone F3 of Downtown Tampa, but it is not clear that the newly installed CCTV cameras played a role in that reduction. Since 2002, the TPD has enjoyed extraordinary success in reducing both the overall and the Big Four Crime Rates in the city. TPD’s ―Focus on Four Plan & 2013 Departmental Annual Report‖ states that crime in Tampa has been reduced 65% since 2002. (3) The report attributes success to proactive, preventative, and intelligence led policing. Of particular interest is the SAFECOP system that provides planners and officers on the street with computer displays of crime information that is used to allocate and deploy resources in response to emerging crime patterns. The report’s only mention of the role of surveillance cameras was to note that officers can use screen shots to quickly create ―crime bulletins.‖ However, it does not indicate if, how frequently, or how successfully this has been done. The numerically small and somewhat inconsistent crime reductions identified in Zone F3 since the installation of the cameras are most logically attributed to a continuation of effective policing methods that have, over a decade, led to significant city-wide annual crime reductions achieved without broad CCTV coverage. Section 2 of the TPD report on surveillance cameras which cites twenty-four “testimonials” to their contributions is also problematic. With minor exceptions,
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014 Page 2
cameras do not appear to have had a serious impact on resolving major crimes. In general, the language of TPD’s ―testimonials‖ does not indicate that cameras played a significant role. Most entries claimed that the cameras ―supported‖, ―assisted‖, ―helped investigate,‖ ―documented‖, ―freed-up manpower‖, ―manage‖ and ―monitor.‖ In only eight of the 24 testimonials do cameras appear to have played some role in the identification and/or arrest of criminal suspects? In most instances, the specific contribution of the cameras is not clear. In three instances, the presence of mobile and/or covert cameras is claimed to have deterred criminal activity – something police presence would have done just as well. Two claimed successes were attributed to FDOT traffic cameras - not the CCTVs TPD is attempting to justify. The testimonials clearly indicate that, in most instances, the general surveillance CCTV cameras were nice to have but were mostly used in a supporting role. The claimed impact on crime does not justify the continued intrusion on the privacy of our citizens or the continued expenditure of funds to maintain the system. Council might ask whether hiring three additional TPD officers would have had a greater impact on reducing crime in the city. TPD’s “testimonials” indicate that “covert cameras made greater contributions to public safety than the general surveillance CCTV cameras. The city has at least 129 cameras of various types, but eight of the 24 ―testimonials‖ addressed contributions by the ten ―covert‖ cameras. Unfortunately, the use of covert cameras has even greater potential for the violation of individual privacy and the abuse of civil liberties. Covert surveillance is particularly susceptible to abuse through racial and ethnic profiling. Council should, for example, inquire into the use of covert cameras to ―manage‖ crowds and ―trespassers‖ in public parks. Parks are set aside for the use and enjoyment of the community. From a civil liberties perspective, an overt police presence is preferable to covert surveillance cameras. An overt police presence deters bad behavior. Using covert cameras would appear to be focused on generating arrests – not deterrence. The use of covert cameras should be limited to the resolution of specific crime problems – not for surveillance of parks and other public areas.
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014
Page 3
If the Council is to approve additional funding (possibly $186K annually) to retain CCTV cameras, it should require that TPD provide rigorous annual program evaluations to answer several critical questions. How many times did police access stored imagery for investigative leads? How many times did cameras identify otherwise unknown suspects or contribute unique information? If cameras contributed to the resolution of a case, what specific roles did they play? Information provided by the TPD last year is not sufficient to justify continued invasions of our privacy and the additional costs involved. Comments on the Tampa Police Department’s CCTV Standard Operating Procedure (SOP) TPD’s CCTV SOP raises several issues that the Council should address. Para. V,C,1, as written, confuses the Criminal Intelligence Bureau’s issuance of a username and password with ―specific authorization‖ to access stored imagery. It appears that, following issuance of username/password, users may access stored imagery at will. The SOP does not define criteria for accessing stored imagery. IT logs users and presumably creates some sort of computerized record of who accessed the system; however, there does not appear to be a requirement for CIB supervisors to specifically approve accessing stored imagery. The SOP should prescribe criteria for accessing stored imagery and require specific approval to access the system. TPD should maintain logs reflecting each time stored images are accessed and the purpose, the specific probable cause or reasonable suspicion for doing so, and the result. Logs should be retained. Logs are critical to evaluating the system’s actual contributions and preventing abuses. SOP Para. V,D, addresses the use of AISight Video Analytics. From a civil liberties perspective, this is one of the system’s most troublesome capabilities. A1 Sight independently analyzes the activities, behaviors, and the scenes that it sees and identifies and alerts users to
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014 Page 4
―suspicious‖ activity (abnormality detection). The system supposedly ―learns‖ what is normal. The system identifies ―unusual‖ activities in the field of vision, categorizes them as suspicious, and the issues an alert. A man walking on a roadway late at night in an area where there is normally no pedestrian traffic could trigger in an alert. A Slate article identifies the problem: ―AISight and similar technologies transform citizens into suspects. Because AISight is used to monitor and detect not just acts of crime but potential acts of crime, based purely on a set of algorithms, it is considered part of the push towards pre-emptive—or ―pre-crime‖—policing, which treats everyone as a potential criminal and targets people for crimes they have not yet committed (and may never commit).‖ (4) The TPD SOP is correct to ensure that using the system requires consent of the CIB, but it is not totally clear who in the CIB has that authority. Council should review TPD’s use (frequency, why, when and for what purpose) of the A1Sight capability since it has been installed and actions taken based on the system’s alerts (if any). The Deployment of Mobile Trailers and Covert Cameras The list of ―testimonials‖ indicates that TPD has deployed both mobile trailer and covert cameras. Claims for successes of the mobile trailer cameras are quite modest. TPD’s response to Council indicates that they ―helped to support‖ crime control at city malls during the holiday season. Interestingly, the department’s annual report for 2012 discussed ―Operation Safe Shopper‖ in some detail without mentioning the use of cameras. The report did note that ―officers on horseback, motorcycles, bikes and unmarked vehicles saturated mall parking lots.‖ (2) The TPD report does not address the use of mobile cameras in residential areas ; however, this capability has exceptional potential to disproportionally target poor and minority neighborhoods. Council should request information concerning the location/purpose of all mobile camera deployments and results (if any) to identify any patterns in their deployment. .
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014
Page 5
TPD’s SOP does not establish criteria for using covert cameras. Council should inquire about the use of covert cameras to ―manage‖ crowds and trespassers in public parks and insist that there should be clear policy outlining when the use of covert cameras is appropriate/necessary. Covert capabilities create opportunities for racial and ethnic profiling. Parks are areas set aside for the use and enjoyment of the community. From a civil liberties perspective, a deterrent overt police presence is much to be preferred. We recognize that TPD may be reluctant to provide outsiders with information on their use of covert cameras, but if Council decides to fund retention of the system, it should exercise effective oversight of their use and ensure that the system is not abused. CCTV Surveillance Images are Public Records In deciding whether to continue funding the system, Council should recognize that CCTV images are public records. Council should request a periodic report on requests for stored images to determine how often, who and for what purpose CCTV images collected by the TPD are being used. We recommend that Council defund the CCTV surveillance system; however, if Council continues to fund system maintenance, it should insist on a rigorous annual program evaluation to determine if the cameras contribution to public safety is sufficient to justify the intrusions on our privacy. (1) (2) (3) (4) http://www.slate.com/blogs/future_tense/2012/06/11/aisight_from_brs_labs_and_other_tec hnologies_to_detect_suspicious_behavior_.html https://www.aclu.org/files/images/asset_upload_file708_35775.pdf http://ipvm.com/report/is_public_cctv_effective http://www.tampagov.net/dept_Police/Files/publications/2012_FOF_Web.pdf
ACLU Analysis of TPD Surveillance Cameras – Feb. 2014
Page 6
