San Diego has deployed cameras on more than 3,000 streetlights, covering roughly 5 percent of the city’s public rights-of-way, and a new investigation looks at whether they’re spread equally in neighborhoods.
(TNS) — A few years back, San Diego approved a plan to install energy-saving, and money-saving, LED street lights.
San Diego was deploying what was billed as the world's largest smart-city platform, powered by thousands of cutting-edge, data-gathering sensors. The plan was to track movements of cars and people, particularly in busy urban areas. Deeper understanding of mobility could follow. Parking and traffic might be eased. Apps could be developed.
All that data-gathering required the installation of cameras on what is so far a little more than 3,000 streetlights, covering about 5 percent of the city's public rights-of-way. For the public, that information came as a surprise — as did the revelation that police had access to the footage.
Many critics didn't trust the system. They feared the existence of a mass surveillance network, which some said could lead to targeting communities of color and potential civil rights abuses. And they've raised questions about where the cameras were installed, how those locations were selected and whose movements are most likely to be recorded.
Police officials have said that the cameras are not live monitored, that they are accessed in limited instances. The city said the locations were selected for mobility — traffic, pedestrian, parking, bicycle — and environmental factors, and that the selection had nothing to do with the makeup of the neighborhoods they were in.
The Union-Tribune analyzed demographics in areas where San Diego's smart streetlights are located, including nearly 230 that have been accessed by police since summer 2018.
While the streetlight cameras were well distributed among the city's racial and economic populations, the analysis showed a greater proportion of black residents and poor residents live in neighborhoods where smart streetlights have been accessed by police.
Streetlights across the city
In an effort to better understand the demographics of the neighborhoods containing smart streetlights, the Union-Tribune mapped each light and noted which census block group it fell within. Block groups, as opposed to much larger census tracts, are small geographic zones — generally with populations between 600 and 3,000 people — as defined by the U.S. Census Bureau.
Data show that more than 946,000 people — about two-thirds of San Diego's total population — live in a block group where the city has placed a smart streetlight camera.
Among the city residents who live near a camera, about 46 percent are white, 25 percent are Hispanic, 18 percent are Asian and 6 percent are black. The remaining 5 percent are Native American, Pacific Islander or another race.
San Diego's total population is about 42 percent white, 30 percent Latino, 17 percent Asian, 6 percent black, and 5 percent other races. Considering these totals, white residents are slightly over-represented and Hispanic residents are slightly under-represented in neighborhoods where the smart streetlights are located.
The Union-Tribune also found that the distribution of streetlight cameras matched the city's distribution of household incomes.
There's a much denser concentration of camera-equipped streetlights in downtown San Diego than anywhere else in the city. There is a reason for that, according to Erik Caldwell, deputy chief operations officer for the city's Smart and Sustainable Communities division.
The cameras are part of a system that uses artificial intelligence to track data. In downtown San Diego, the city wanted granular information about parking and mobility. That meant the cameras needed to be in close proximity to one another, in some cases less than 100 feet apart.
There are 12 census block groups in which the city placed more than 50 street lights. All 12 are downtown.
There are about 31,000 people living in those areas. Of that number, about 56 percent are white, 23 percent are Latino, 9 percent are Asian, 8 percent are black and 4 percent are other races.
Cameras accessed by police
San Diego police praise the cameras as a game-changer for investigating serious or violent crimes. In several cases when police accessed the cameras, arrests have followed.
Police have credited the cameras with helping them to quickly identify a suspect in a shooting last year that killed one employee and injured two others at an Otay Mesa Church's Chicken restaurant. The cameras helped clear a man under investigation for a downtown homicide, after the footage revealed the killing had happened in self-defense.
Through the middle of last month, investigators had accessed the cameras in 246 cases. Sometimes multiple cameras in an area are accessed as part of one case.
Based on information provided by police, the Union-Tribune mapped the approximate locations of all but 22 of those incidents. Police did not disclose information in 19 cases, citing ongoing investigations. Another three sites provided by police listed only a street name, so the location of the incident couldn't be placed in a census block.
About 262,000 people live in the areas where police have accessed the streetlight cameras. Of those residents, 9 percent are black, compared with the city's overall black population of 6 percent.
Hispanic or Latino residents are also somewhat over-represented in these areas. Thirty-three percent of residents who live in areas with streetlights that have been accessed by police are Latino, compared with the city's Latino population of 30 percent.
White people make up about 42 of San Diego's population, and Asians make up 17 percent. In areas where police accessed cameras, about 39 percent of the residents are white and 16 percent are Asian.
Other races account for 5 percent of the city's residents. In areas where police accessed cameras, those other races make up 4 percent.
Police have accessed cameras across the city, including many in downtown San Diego. In an area near the convention center, police have accessed cameras 12 times. In an area in downtown's Core-Columbia community, sandwiched between Ash Street and Broadway, investigators have accessed cameras there nine times.
These two neighborhoods also see some of the highest violent crime numbers in the city. More than 300 violent crimes, including two homicides, were reported in the Core-Columbia area between 2014 and 2018. In the area near the convention center, more than 600 violent crimes, including nearly 50 rapes and attempted rapes, were committed between 2014 and 2018.
Critics of the program say that they are interested in public safety, but argue that a move toward greater public safety shouldn't come at the expense of privacy and civil rights. Pitting those interests against one another other is unreasonable, they argue.
"No one has ever said that we don't want safer communities. It's the use of this intrusive surveillance technology without proper oversight (that is troubling)," Geneviéve Jones-Wright with TRUST SD, a vocal critic of the program, told the Union-Tribune late last year.
San Diego Police Chief David Nisleit has said he understands the concern that the cameras could be used to racially profile or otherwise discriminate against people. But he said those fears are unwarranted and stressed that there is no live monitoring of the camera footage.
"As long as there is not a crime, we are not looking at video," he told the Union-Tribune last year.
The chief said police used the cameras "very sparingly, for only the worst type of cases, very violent type of cases, or serious or fatal injury collisions." He said they are used as "a reactive tool, as an investigative tip to lead us in the path of who is responsible for the crime. It actually lets us narrow our focus."
Of the 224 cases about which police provided information, most involved serious crimes. The most common crimes to prompt streetlight access: fatal or serious injury collisions, robberies, assaults with deadly weapons and homicides.
Through mid-January, San Diego police have accessed cameras during 23 homicide investigations.
According to the department's smart streetlight policy, the decision to view the footage is driven by several considerations, including how serious the crime was and how significant the threat to the public.
"We are looking at incidents that have resulted in a life being lost or placed at significant risk, as well as property crimes with exceptional damages that threaten the safety of our communities," said San Diego police Capt. Jeffrey Jordon.
Data show that the types of cases that prompt police to access streetlight footage are similar across communities. In census block groups where people of color make up 51 percent or more of the population and census block groups that were predominately white, police most commonly used streetlights cameras to investigate serious or deadly crimes, from robbery to homicide.
Dave Maass, a senior researcher for the San Francisco-based Electronic Frontier Foundation, said the Union-Tribune's findings didn't surprise him and that he has seen trends where surveillance technology has a higher impact on communities of color.
He said there are several factors that could explain why police access cameras in some communities more than others. Maybe those communities experience more crime or residents are less likely to work with police. It could be related to the nexus between crime and poverty.
According to the Union-Tribune's analysis, of more than 100,000 households located in areas where police had accessed smart streetlights, about 23 percent of homes reported household incomes of less than $30,000 a year. About 18 percent of San Diego households bring in less than $30,000 a year.
With the streetlight cameras, Maass said, "the risks could be racial bias, change in service or crimes they solve."
"Suddenly it becomes very easy to prosecute jaywalking cases," Maass noted as an extreme example of risks associated with cameras. "If it makes it easier to do one thing than another, police may gravitate toward low hanging fruit."
Jordon stressed that the department accesses streetlight sensors usually after a violent crime has been committed. He added that according to the department's crime statistics, 36 percent of violent crime victims are Hispanic and 19 percent are black — a larger percentage than their population estimates. Given these statistics, Jordon said it's reasonable to expect a greater proportion of violent crimes investigated through streetlight sensors would involve Hispanic or black residents.
"Smart Streetlights enhance (the department's) ability to investigate violent crimes, and reduce their traumatic impact in our neighborhoods regardless of race or socio-economic status," Jordon said.
How were locations selected?
Cody Hooven, who heads up the city's Sustainability Department, which oversees the Smart Street Lights program, said in a statement to the Union-Tribune that the decision about where to place street light sensors came after a joint assessment between the city and GE Current, the company the city is partnering with on the project. (Until April 2019, GE owned Current. It is now owned by a private equity firm, American Industrial Partners.)
She said locations were analyzed by their ability to collect data on traffic, parking, bicycle and environmental factors. She said the locations "have a cross-section of activity and had nothing to do the makeup of the neighborhoods they currently serve."
Caldwell said one driver of where the sensors were put was the physical condition of the streetlights, and whether they had enough voltage to handle the sensors. That's why none of the smart streetlights are in pedestrian-heavy Balboa Park.
"That was a great example of very early on saying 'Hey, it would be fantastic to capture pedestrian data as people are moving through the park and getting some really more accurate information — more than we have today,'" Caldwell said. "But the voltages were wrong (in Balboa Park's street-light circuits), and that was just not possible."
Caldwell said the department "could have done better in the actual selection of locations by seeking input from outside interests."
He and others have said repeatedly that San Diego police had no input in the selection of the locations. Police officials were not even aware the cameras existed until April 2018. Caldwell said police officials have told him that had they been asked for input, they would have suggested different locations.
The Police Department has penned a policy outlining the situations in which investigators — inside the department and outside law enforcement agencies — can access the camera footage.
In addition to the more than 3,000 smart streetlights already installed around the city, another 1,200 had been slated to be put in later this year. That has since been put on hold, given the community push back.
City Council President Georgette Gómez said she has "on multiple occasions expressed serious reservations about the smart streetlights program — regarding both the community's concerns about how it's used and its increasing cost."
She said she wants the council to take up discussion on the streetlight cameras.
The city had proposed drafting a policy to govern the streetlights program. But in January, a City Council committee said it wanted to go further. It wanted an ordinance — one that would set rules not just for the smart streetlight cameras but other current and future forms of surveillance.
The city is working to draft that ordinance.
©2020 The San Diego Union-Tribune, Distributed by Tribune Content Agency, LLC.
Looking for the latest gov tech news as it happens? Subscribe to GT newsletters.