Cameras, cops, and red light-running: Can Oakland solve its collision problem?

When 69-year-old Pua Hernandez and her husband moved to Oakland from Hawaii nine years ago, they didn’t expect their lives to change much. Maybe they’d have less-scenic walks in their new Fruitvale neighborhood. But they never thought they’d feel unsafe. Now, Hernandez said the couple’s daily activities have slowed to a crawl, in large part due to their fear of walking on Foothill Boulevard and International Boulevard, where drivers routinely speed and run red lights.

“It’s all hours of the day,” said Hernandez. “If I’m coming down 27th [Avenue] to turn left onto International, I never pull out right away on the green light because people just go right through that red light. [The city] recently added a four-way stop at 27th, but not everybody stops.”

A self-described social and independent person by nature, Hernandez said she stays home most days because she’s afraid to go outside. She used to enjoy walking to her niece’s nearby home, but can’t now because her niece moved to Alameda. When her husband goes out for a walk, Hernandez makes sure to accompany him because of his mobility issues. The couple was almost hit by a red-light runner two weeks ago as they stepped onto a crosswalk. “I feel pretty isolated now,” Hernandez said. 

For years, residents like Hernandez have complained to Oakland officials about cars running red lights. Crossing a street by foot, bike, or car, after all, is an exercise in trust. In Oakland, that trust can often feel broken.

In September, The Oaklandside launched a reader survey to gather information about people’s experiences on our city’s dangerous roadways. More than a third of the 100-plus people who responded told us that red-light running is a major concern. Respondents were from neighborhoods across the city and included Black, Latinx, and Asian residents in the East Oakland flatlands, where collisions occur more frequently. One person told us that she saw a car burst through a red light on Grand Avenue and Mandela Parkway in West Oakland, colliding with another vehicle and throwing it in the air. It landed two feet from her car. Another Oaklander said he was hit by a car running a red light at an intersection downtown while biking to work but avoided serious injury by swerving. 

Others have not been as lucky. Eight percent of crashes that result in death or injury in our city are caused by people “disobeying traffic signals and signs,” according to the Oakland Department of Transportation. In the last 10 years, according to an Oakland Police Department statement, there have been 24 emergency calls for fatalities caused by red-light violations and 75% of road fatalities in the city occur at signalized intersections.

Oakland is not alone in these problems, and collisions involving red-light running appear to be increasing nationally. According to the American Automobile Association, 939 people were killed by drivers running red stoplights in 2017, at the time a 10-year high. Sacramento sees about the same percentage of its traffic fatalities or serious injuries at signalized intersections as Oakland. In a Houston Chronicle analysis from 2001 through 2016, the metro areas with the most intersection fatalities per capita were Miami, Phoenix, and Los Angeles. (It’s unclear how Oakland would have ranked in the list because it was included as part of the larger San Francisco metro region, which ranked 10th among the 12 largest metro areas.) 

Regardless of where Oakland stands in comparison to other cities, many residents will tell you (and we heard from a lot of them) that red-light running is a big problem in Oakland. And they have a lot of theories about why that is. Some, like Hernandez, believe there are not enough cops to enforce traffic laws, and nearly everyone who responded to our survey said they have never seen anyone pulled over in Oakland for a traffic violation. Others think that some drivers have adopted a reckless and careless attitude when behind the wheel. One resident, Josh Frank, expressed a sentiment held by many when he told The Oaklandside there should be speed and red-light cameras to catch scofflaws. 

“A camera is going to take a picture of your license plate and send you a ticket. That’s gonna probably cause people to think differently,” said Frank, who works in the marketing department of a local software company and has lived in Oakland for almost ten years. Frank frequently bicycles along Park Boulevard and is concerned about the speed of cars on the road. “We all have a right to privacy, but if you’re in a motor vehicle registered with the state and you’re breaking the law running through red lights, that’s infringing on the right to a safe environment.”

Automated red-light cameras monitor intersections 24 hours a day, taking high-resolution photos and videos of cars and drivers at different angles when they cross. They record the vehicle’s time, date, speed, and the elapsed time of both the yellow and red lights. In theory, the cameras are supposed to trigger a capture when a violation occurs. Law enforcement officials then review each digital image for accuracy before sending out a citation by mail. Supporters of red-light cameras also note that they eliminate the need for a police officer to pull a motorist over, a means of enforcement that has been proven to be racially biased, and which can result in violence.

But do red-light cameras actually make roads safer? What are the pros and cons? Are there more effective ways to bring down collisions? In this article, we set out to explore these questions as they apply to Oakland. 

Buckle up, because the answers are complicated. 

A brief history of red-light cameras and privacy concerns

The city today has no working red-light cameras. Oakland had 13 of them set up to monitor 11 intersections between 2008 and 2014, but let its camera-system contract lapse, following a national trend: Oakland was one of nearly 160 cities that stopped using the cameras between 2011 and 2016. The programs ended despite some studies, including one that Oakland commissioned in the early 2010s, finding that the average number of collisions per year at intersections with red-light cameras went down over the six years they were in use, from 7.2 to 3.2 per intersection. Some nearby towns like Fremont still operate the cameras, and their police departments have said cameras can decrease collisions.

Above: An interactive map showing the location of red-light cameras used in Oakland between 2008-2014, and current high-injury intersections and high-injury corridors. Data provided by the city of Oakland. Credit: Jose Fermoso

So why did Oakland shut its cameras down? There are several likely reasons: pressure from local privacy advocates, conflicting studies about the cameras’ effectiveness, and the fact that the system was a huge money loser. All of this contributed to the cameras not being very popular with some residents. 

Oakland’s Red Light Photo Enforcement Program, launched in 2008, was supposed to pay for itself. But by 2014, despite numerous red-light citations at intersections like Northgate Avenue and 27th Street (leading to over $2 million in fines from that crossing alone), city officials like District 5 Councilmember Noel Gallo, who chaired the city’s Public Safety Committee at the time, felt Oakland was barely getting its money’s worth. And for most of the program’s duration, it was indeed in the red, with monthly revenues not covering the big overall investment. Although the red-light cameras eventually turned a profit, the city didn’t feel the returns on safety or revenue were enough to overcome the early financial losses. Los Angeles had reached the same conclusion three years earlier in 2011 when its city council voted unanimously to end that city’s program.

Cost aside, a debate has raged about whether or not the cameras truly work to reduce collisions. Despite some positive numbers in Oakland, the data is largely inconsistent. At several Oakland intersections, close to half of all infractions that were caught by cameras ended up being unenforceable due to the poor image quality of license plates and vehicles. A camera at 66th Avenue and San Leandro Street, for example, captured more than 7,500 red light violations between September 2008 and July 2011. Yet only 3,308 of them resulted in citations. In addition, several major Oakland intersections with red-light cameras, such as High Street and Brookdale, and Macarthur Boulevard and 82nd Avenue, saw no significant decrease in collisions. 

Tracy Rosenberg, the advocacy director at Oakland Privacy, a nonprofit that monitors surveillance tech, said any small gains the cameras provide are overshadowed by their harmful impacts.

“Technology as the magic bullet is always tempting, but it greatly increases criminal justice involvement,” said Rosenberg. “Any time we capture a car crossing an intersection, we end up with a whole pile of data. And the question of how that all gets handled, who has access to it, what the protections are, is something that, as a society, we are not good at. Sooner or later, you find examples of misuse or inappropriate access.”

Privacy advocates have been pushing back against requests by the city, especially OPD, for new surveillance technologies for more than a decade. In 2007, the police department asked the city’s lobbyist, Townsend Public Affairs, to see if the state legislature would change the California Vehicle Code to allow them to use red-light cameras for “any” law enforcement purpose, not just to write traffic tickets, but the request went nowhere. In 2012, during debates over the city’s proposal to create a “Domain Awareness Center”—a sweeping city-wide surveillance network that would feed information from cameras and microphones throughout the city into a command center—City Council members struggled to get accurate information about whether a new software update could add facial-recognition tech or other controversial tools to the system. Ultimately, after pressure from community organizers and organizations like the ACLU of Northern California, the City Council in 2016 decided to limit the DAC to just the Port of Oakland and created a Privacy Advisory Commission to advise the city on all future surveillance technology proposals. Only two months ago, the chair of Oakland’s privacy commission sued the Oakland Police Department for failing to adequately disclose their use of license plate data culled from automated city cameras, which are different from red-light cameras. 

Cameras and other types of surveillance tech have also been criticized for being wielded unfairly against people of color and other vulnerable groups, and for causing financial harm by increasing the debt burden on people who are already struggling. 

Camera advocates reject claims of inequity by saying automated cameras are, by their nature, objective. But critics say enforcement is often meted out unequally and have called for more human judgment when it comes to making determinations between hazardous and non-hazardous traffic violations. For example, cameras today punish people for driving through a red light, even for a fraction of a second, whether they are driving within the speed limit or bursting through at a much faster speed. 

“If we’re giving $100 tickets to everyone who’s going 42 miles per hour in a 40 zone, we are treating that person the same as if they’re going 80,” Rosenberg said. “On a common-sense level, that’s not the same thing. The level of danger is not the same. So why hurt their pocketbooks the same?”

Do red light cameras reduce collisions?

The widespread removal of red-light cameras from cities in recent years has allowed researchers to compare their effectiveness before, during, and after adoption, and three very different schools of thought have emerged. First, red-light cameras are less effective at changing driver behavior than road design. Second, red-light cameras can actually lead to more collisions and injuries. And third, cameras can save lives by reducing collisions at intersections if cities just put them in the right places. 

The Insurance Institute for Highway Safety, a nonprofit created by the auto insurance industry, holds the latter view. In a 2017 study, IIHS researchers Wen Hu and Jessica Cicchino published a study that found fatal crashes caused by red-light running rose 30% in some areas after cameras were taken down. “Oakland was one of the cities in our study. We saw fatalities go back up when the cameras were turned off,” said Cicchino. “It’s something that concerns us since the number of cities with red-light cameras peaked 10 years ago. From 2011 to 2014, there were about 500 cities using red-light cameras. Now there’s about 345.”

Other researchers have found that red-light cameras slightly reduced t-bone collisions—when the front of one vehicle hits the side of another—but increased sideswipe and rear-end collisions. In a 2013 study of 32 intersections in Charlotte, North Carolina, UNC Charlotte’s Srinivas S. Pulugurtha found red-light cameras increased sideswipe and rear-end collisions at more than 50% of the intersections, indicating that drivers were breaking harder to avoid citations.

Not surprisingly, another factor found to increase collisions on intersections, whether there is a red-light camera or not, is how fast people drive. In the North Carolina study, one-third of rear-end crashes happened because the driver failed to reduce speed.

In Oakland, Redflex Traffic Systems, the private camera operator contracted by the city, conducted two studies to help determine camera locations. But when Oakland installed the cameras in 2008, some people criticized the city for choosing the wrong intersections. The camera at Northgate and 27th Street, for example, picked up a lot of red-light violations in part because of its location near a freeway offramp. But there were other intersections in the city with a higher number of serious accidents that didn’t get a camera.

The city’s transportation manager at the time, Wladimir Wlassosky, told The Oaklandside during a recent interview that his department did look at Redflex’s studies, but ultimately settled on camera locations based on a separate review of collision data compiled by the Statewide Integrated Traffic Records System (SWITRS).

Ultimately, it doesn’t matter how you choose the locations of the red-light cameras, said Paul Fisher, an economist at the University of Arizona who has studied the cameras. According to Fisher, cameras tend to cover only 5% of city streets, on average. With about 700 light-controlled intersections at the time it was operating cameras, Oakland only had about 1.5% of its intersections covered, meaning that numerous intersections with a history of deadly collisions weren’t being monitored.

Furthermore, Fisher and other researchers say intersections are made dangerous by the built and natural environment, things that cameras can do little to change. “Cities tend to attribute the natural ebb and flow of traffic accidents to the cameras instead of recognizing collisions would have happened anyway,” he said. Visibility fixes like cutting down trees that obscured drivers’ views of traffic and signals, Fisher told us, can have more of an impact on lowering collisions than the presence of cameras. 

“If you have a lot of red-light running consistently at a particular location, then you have an engineering problem,” said Jay Beeber, a Research Fellow at the National Motorists Association, a nonprofit that advocates on behalf of automobile owners. If collisions are happening year after year, the city should study the environment to figure it out, he said. 

High rates of red-light running have also been observed in cities with similar street infrastructure issues as Oakland, including Tucson, Houston, and Cleveland.

What about the yellow light?

Vehicle speed is the primary cause of death in 25% of fatal collisions at Oakland intersections, according to the city. International studies have found that at 40 mph, 80% of people die from a collision. The number dips to 10% when cars are driving 20 mph. One obvious solution that could reduce collisions and save lives then, is reducing car speeds. 

But getting drivers to slow down isn’t easy. California Governor Gavin Newsom recently passed a law that will allow city civil engineers to set lower speed limits. But some experts say this law won’t make a difference as long as Oakland roads continue to invite speeding on their wide avenues and long, swooping driving lanes; that the way a street is designed causes people to often drive much faster than the posted speed limit.

Some believe that more enforcement of speeding through the issuance of citations could deter speeding. Oakland officials have said they would like to add automated cameras equipped with radar to catch people speeding and reduce collisions, but unlike red light cameras, California does not allow speed limit enforcement cameras. Instead, speeding citations today are given out entirely by police officers, and city officials say there aren’t currently enough police in Oakland to enforce speeding laws. Police accountability activists have also been pushing over the past year and a half to dramatically reduce the role of police in responding to nonviolent service calls and enforcing things like traffic violations. 

Earlier this year, State Assemblymember David Chiu sponsored AB 550, a speed camera bill with privacy and equity guard rails to satisfy surveillance fears, that would have given cities the authority to install speed limit cameras—and allowed people cited by them to work off fines through community service. The bill never made it to Newsom’s desk. 

Oakland does currently have speed cameras set up at numerous intersections, which are used for traffic management. Placed at the top of the lights or other poles nearby, the cameras can measure vehicle speeds and help city engineers determine the timing of red, yellow, and green lights. Some say adjusting these traffic signal times—specifically, making yellow lights longer—might be one way to reduce collisions. 

“An appropriately long yellow light has always been important to the reduction of collisions that lead to injuries,” University of Arizona researcher Fisher told us. 

According to several traffic experts we spoke to, the duration of the “dilemma zone” is key. The dilemma zone is the period when the light turns yellow and drivers must decide to either run a red light or stop so hard it might cause a dangerous fender bender or worse. The shorter the yellow-light time, the faster a driver has to make this crucial decision. Less decision time can lead to bad decisions that cause accidents, experts say.

Yellow-light durations, and by extension, dilemma zones, are calculated by cities using an equation that takes into account the average vehicle speed on a particular street, the average person’s reaction time, and the time it takes to safely bring a car to a stop. While it’s very complex, traffic engineers at the peak of the car revolution in the mid-20th century appeared for a long time to have figured it out. In 1963, a trio of scientists came up with a kinematic formula—a math equation used to describe the motion of physical objects—to measure all of the data consistently. The equation was used to set federal traffic engineering standards that later on informed California’s method of setting yellow light durations. 

Here’s how it works in Oakland: For a flat road with a posted speed limit of 25 mph, such as International Boulevard, a traffic engineer would use 32 mph for the approach speed of an average car heading towards an intersection. Calculated with the two other factors, each intersection’s yellow light duration would, at the very minimum, have to be 3.3 seconds long.

But is 3.3 seconds really the right amount of time for a yellow light on International Boulevard? Would it help if the yellow lights were actually a little longer?

When cities increase yellow times, they usually do lead to positive results. In a 2004 Texas study, intersections with yellows set one second longer than the minimum time suggested using the state’s engineering guidelines based on the kinematic equation decreased collisions by 40% and red-light violations by 53%. The Institute of Transportation Engineers, a professional association that sets many of the standards that end up being adopted by states and cities, also conducted studies showing that increasing yellow light times leads to fewer collisions. 

Wlassosky, Oakland’s former transportation director, said that Oakland actually took ITE’s advice and in December 2009 added one second to most of its traffic signals where red-light cameras had also been installed. According to Jay Beeber, executive director at Safer Streets L.A. and a research fellow with the National Motorists Association, the one-second change led to a significant difference in Oakland’s red-light running for four months. There was an 87% reduction in the number of violations within the first second of the red phase, Beeber wrote. Fewer people were rushing through intersections because of a short dilemma zone. 

People who run a red light the first second or two after it turns red are committing what traffic engineers call “early-into-red” violations. These violations do not cause many collisions, according to Beeber. “Late-into-red” violations occur long after a light has turned red and are considered the most dangerous. During the four-month period that Oakland had longer yellow lights, Beeber found only one red-light running collision at all 10 photo-enforced intersections combined. 

“In contrast, before and after this four-month period, when the yellow timing was one-second shorter, the photo-enforced intersections experienced an average of about two crashes per month,” he wrote. 

But the Public Works department, which managed transportation infrastructure prior to OakDOT’s creation in 2016, decided to change the yellows back to their original timing lengths. Emails from that time suggest the police department wanted the durations shortened because fewer citations were reducing the red-light camera program’s revenue. Ultimately, OPD’s position ended up winning out. 

Wlassosky said that four months of data was not enough to determine whether longer yellow light times actually made the intersections safer. “Typically, you need a minimum of three to five-year studies and even 10-year studies comparing before and after,” he said. 

The retired former director also said that when they changed yellow-light times back to the shorter duration, they did so to follow the state and federal rules. “We had a meeting with OPD. At the end of the day, we agreed to implement the book standard that gave us allowances in the yellow light. They were a fair bit longer than they had been pre-red-light cameras,” Wlassosky said. 

Beeber believes Oakland should have kept the longer yellow lights and that the city should be looking at any and all solutions that can improve safety. “Not just the ones that make a profit for the city or pays for police officers’ salaries,” he said.

Could a new yellow light equation be the solution?

In 2014, an electrical engineer from Oregon named Mats Jarlstrom got angry when his wife got a pricey ticket for running a red light. Jarlstrom went before the local court and argued she did not have enough time to make a decision to either stop or continue through the intersection. She had been unwittingly dragged into the dilemma zone. The court upheld the ticket but Jarlstrom threw himself into the field of traffic engineering and became convinced that the kinematic equation being used was flawed because it assumes all drivers pass through an intersection at the same speed, regardless of whether they are going straight or making a left or right turn. When in fact, those who are turning tend to slow down much more, and earlier, than those going straight.

Jarlstrom decided to update the equation, taking into account the changing speed cars make on a turn. His calculations led to yellow light times that were two to three seconds longer than the equations in use across the United States. 

“Any extra time can help people not run the red light,” he said in an interview. 

For the next several years, Jarlstrom tried to convince Oregon’s transportation department to increase the length of time for yellow lights for left turns. His formula would add not just one second to some lights, as engineers did in Oakland in 2009, but depending on the length of the block or width of the street, sometimes more. 

Instead of listening to him, Oregon fined Jarlstrom for civil engineering without a license. He sued the state, eventually won on the grounds of free speech, and continued asking traffic engineers to use his method. In March 2020, he achieved his goal. In an official journal, ITE accepted the premise of his “extended kinematic” formula and recommended that cities across the U.S. consider the new equation. Even one of the original creators of the old kinematic equation, now reaching 90 years old, approved of his work. According to Jeff Lindley, the Chief Technical Officer of ITE, some transportation agencies in the U.S. have started trying the new equation to change their jurisdiction’s light timing. Additionally, the U.S. Department of Transportation has created a Traffic Signal Change and Clearance Interval Pooled Fund to study the equation and its effects over the next five years. States like Connecticut, Georgia, and Illinois have signed up to be a part of that study.

What would the extended kinematic equation mean for Oakland roads if it was used here? Many yellow lights might need to be extended at least an extra second, to five on average, and maybe more on dangerous roads like Foothill or International. And according to the U.S. Department of Transportation, Oakland does have the legal right to extend the length of the yellow lights to match the new equation, granted that they have a solid engineering reason to do so. 

Dozens of jurisdictions and academic researchers are now studying the real-world implication of the new equation. One of them is University of Michigan Professor Henry Liu. Liu has attached GPS trackers to cars to test the accuracy of Jarsltrom’s new equation. He said in an interview that it’s good on balance, but might be too aggressive in that it adds too much yellow time. 

“In our study of four intersections of vehicle trajectories, left-turn cars slow down earlier and also speed up halfway through an intersection, and so a better equation that calculates a yellow light is somewhere in between the old equation and the new one,” Liu said. 

Liu is also finding that adding time to yellow lights might be a double-edged sword. By increasing the yellow light, you also reduce the green light’s duration, leading to more delays and more congestion. Liu also noted the worst accidents at intersections come from the most irrational people, or the physically or psychologically impaired, for whom a longer yellow light, a red-light camera, a stop sign, or even a red light won’t make them stop. ITE, the professor concluded, should not have issued the new guideline because it might lead to adverse effects. 

OakDOT officials haven’t said whether or not they’re planning to make any changes to the duration of yellow lights, based on Jarlstrom’s new equation or for any other reason. And the city doesn’t have plans to reinstall red light cameras any time soon. 

As for residents like Pua Hernandez, all they want is for safer streets to walk, bicycle, and drive. 

When I mentioned to Hernandez that longer yellow lights tended to reduce red-light running and could make the streets safer for pedestrians and other drivers, she was excited. In addition to her daily walks in Fruitvale, one of the things she used to enjoy the most was getting in her car for hours to explore the coast. That’s partly why, despite all the road dangers, she doesn’t want to leave Oakland. 

“We did a road trip to Oregon and the cost of living there was a lot easier,” she said. “But deep down, we love this area. We really do.”

Jose Fermoso is a University of Michigan Knight-Wallace Reporting Fellow