Seattle’s public-safety debate often begins with a convenient lie. We pretend not to know what to do, then spend years studying familiar problems as they grow worse.
We commission another study. We assemble another advisory group. We argue over semantics and wait for evidence that has already arrived. When conditions deteriorate, we declare the problem too complicated to solve and leave it for the next administration.
San Francisco convinced me that Seattle is not stuck because we lack solutions to make people safer. We are stuck because we are unwilling to put those solutions into practice.
The people responsible for solving our hardest problems often know exactly what is wrong. They have read the research, studied the alternatives and watched successful approaches work elsewhere. If pressed privately, many could describe what should happen next.
The barrier is not knowledge. It is whether our leaders are willing to act when the evidence challenges their political identity.
We resist solutions because accepting them would require us to revise the story we tell about ourselves. If cameras prevent crime, we may have to reconsider whether every camera is an instrument of oppression. If focused deterrence reduces gang violence, we may have to accept that targeted enforcement can protect communities of color rather than threaten them. If legal intervention helps move people toward treatment, we must acknowledge that compassion and consequences can coexist.
Those conclusions can feel like concessions to people who have spent years opposing them. So we continue diagnosing problems we already understand. Diagnosis allows us to express concern without testing our beliefs against results.
People are not always afraid of change itself. Sometimes we are afraid of the identity we might have to surrender to change.
Malcolm Gladwell offers a useful frame in his new book, The American Way of Killing. He argues that America’s gun-violence debate has become dishonest because both sides remain invested in familiar arguments while practical interventions receive less attention. Political identity determines what evidence we are willing to see.
That dynamic came to mind repeatedly during our visit to San Francisco. Like Seattle, San Francisco has a deeply progressive political culture. It has strong civil-liberties advocates and a long history of skepticism toward police surveillance. The difference is that the city has begun treating these concerns as requirements for how technology should be used, rather than reasons it cannot be used at all.
We saw the result in Union Square, the heart of downtown San Fransico.
The district is supported by a continuous public-safety presence and a network of more than 500 privately operated cameras. The cameras support a command center that helps businesses, police and safety ambassadors understand what is happening in real time.
When a business calls for help, responders do not have to begin with a vague description and hours of uncertainty. They can see what occurred, determine where the person went and give officers useful information before they arrive.
This helps prevent one incident from becoming a pattern. It can also help officers respond more safely. An officer who is more well informed about an incident is less dependent on a hurried eyewitness description. A drone that follows a fleeing vehicle can reduce the need for a dangerous pursuit. Technology does not replace police. It can reduce the number of encounters that depend entirely on a hunch or split-second judgment.
A camera can be reviewed and audited. A hunch cannot.
Research spanning 40 years found that cameras produce a modest but significant reduction in crime. Across 76 studies, overall crime fell by an average of 14 percent in areas with cameras. Property and vehicle crime each declined by 16 percent. The strongest results came when cameras were targeted, actively monitored and connected to a broader public-safety response.
San Francisco’s recent progress cannot be attributed to cameras alone. They are one part of a coordinated strategy that combines technology with enforcement, outreach and rapid response. Still, the results are difficult to dismiss. Reported crime fell 22 percent during the first half of 2026 compared with the same period in 2025. The previous year, San Francisco recorded 28 homicides, its fewest since 1954.
The progress extends beyond conventional crime statistics. A preliminary count found fewer than 100 people living in tents in 2026, down from nearly 650 in 2024. That is an 85 percent decline. Preliminary overdose deaths also fell 27 percent during the first half of 2026.
Residents appear to recognize the change. A recent San Francisco Chronicle poll found that 74 percent of registered voters approve of Mayor Lurie’s performance.
No single statistic proves that one policy produced the turnaround. Together, however, these results show a city moving in the right direction. San Francisco paired technology with clear priorities and an expectation that government would respond. The cameras did not replace leadership. Leadership made the cameras useful.
Crime is also highly concentrated. Decades of research show that a relatively small number of repeat offenders can account for a disproportionate share of offenses. Cameras help identify those patterns and provide evidence tied to conduct rather than intuition or appearance.
That is one reason technology can advance progressive goals. It can make policing more precise. Instead of stopping more people in the hope of finding the right one, police can use evidence to identify the person connected to a specific act.
The alternative to precise intervention is not neutrality. Too often, it is neglect.
San Francisco has not abandoned its concerns about government power. Its Union Square camera network is largely privately operated. Camera owners control what they share. Police access is governed by written rules. SFPD may not use facial recognition on footage obtained through the non-city network. Officers are prohibited from using the system solely to assist federal civil immigration enforcement. Lawful warrants and court orders still apply. There is no standing permission for government to search everything for any purpose.
San Francisco did not resolve the tension by pretending technology was harmless. It built rules around the risk and proceeded.
That is governing.
I was in China around this time last year and the contrast was striking. Crime seemed almost absent from ordinary conversation and from the public spaces we visited. Someone told me the country was so safe that I could leave my wallet on the sidewalk and expect no one to touch it.
I did not test the theory.
Crime certainly exists in China, but its extensive surveillance system is part of daily life. The government possesses a level of power to monitor people that I would never want to import. Citizens have few meaningful ways to know how that information is used or to challenge the government when it crosses a line.
That is a surveillance state.
A limited camera network governed by retention rules, public policies, independent audits, warrants and constitutional protections is not the same thing. Treating them as equivalent weakens our ability to distinguish accountable use from actual government abuse.
The choice is not between China and blindness. We can use modern public-safety technology within the boundaries of a democratic society.
Washington has already shown that balance is possible. When I worked on our state’s facial-recognition law, we did not choose between unrestricted use and prohibition. The law requires notice, testing and meaningful human review. Persistent surveillance and real-time identification generally require a warrant. Technology can assist an investigation, but it cannot replace evidence or human judgment.
Those protections reflect progressive values in practice. They recognize both the usefulness of the tool and the danger of unchecked power.
I have had to confront this same conflict between evidence and identity in my own decisions.
After the Washington Supreme Court struck down the state’s drug-possession law, the Legislature repeatedly considered proposals to create a new criminal penalty. I believed then, as I do now, that addiction is fundamentally a public-health crisis. I worried that criminalization would reproduce the harms of the failed war on drugs without creating meaningful access to treatment.
I voted against final-passage versions that criminalized possession three times. I voted no twice in 2021 and again in March 2023.
After the regular session ended without a final agreement, the Legislature returned for a one-day special session in May. By then, the consequences of having no workable intervention had become impossible for me to ignore. Public drug use was increasing. People were dying. The voluntary system we had hoped would replace criminal enforcement was not reaching enough people.
On the fourth vote, I voted yes.
That decision did not mean I had stopped believing in public health or embraced incarceration as treatment. It meant reality had forced me to acknowledge that my preferred framework was incomplete. A public-health system without a reliable point of intervention was leaving people to die in public.
Changing my vote required confronting part of my political identity.
I thought opposition of criminalization was evidence of compassion. Eventually, I had to ask whether maintaining that position was producing a compassionate result. The answer was clearly no.
Changing your mind is not a betrayal of your values when the evidence shows that your policy is failing. Sometimes it is the only way to honor them.
At some point, refusing to change stops being caution. It becomes an ethical decision.
When leaders know the current approach is failing, understand which alternatives have worked elsewhere and still choose not to act, the status quo is no longer morally neutral. The costs are transferred to the people government was created to serve.
The business owner pays for another broken window. The student changes the route they take home. The family loses someone to violence. Communities already carrying the greatest burden are asked to keep carrying it, so those in power do not have to reconsider their assumptions.
We rigorously examine the possible harm of every new tool. We spend far less time examining the harm caused by relying on systems built for another era.
Police are still expected to depend heavily on patrol cars, radio calls, and eyewitness descriptions. Cameras can verify what occurred. Drones can follow a dangerous vehicle without initiating a pursuit. Real-time information can help responders reach someone before an incident becomes a tragedy.
New technology brings new risks. Those risks require rules, oversight and accountability. Old tools have risks too. They are slower, less precise, and more dependent on human memory and judgment. Refusing modern tools does not eliminate harm. It preserves the harm we have grown accustomed to.
Seattle’s Real Time Crime Center already demonstrates what is possible. Its original system included 62 camera locations in three defined areas. The cameras are not continuously monitored. Footage is generally deleted after five days unless it becomes evidence. The city reports that the center has helped solve hundreds of violent-crime cases.
Recent oversight found that police generally follow the city’s surveillance rules and uncovered no improper sharing of information. It also identified weaknesses in documentation and policies governing private cameras. Those deficiencies should be corrected. Their discovery is evidence that oversight can improve the system, not proof that it should be dismantled.
We should take significant civil-liberties concerns seriously. We should also be discerning about who claims to speak for the community.
Some critics have spent years working on privacy, constitutional rights, and police accountability. Their expertise deserves respect. Other organizations appear almost overnight around a single controversy and create the impression of widespread opposition through press releases and social media.
They have every right to participate. But advocacy is not constituency, and volume is not representation.
It is easy to reject a public-safety tool when you will not bear the consequences of that decision. It is different when you own the business replacing another shattered window. It is different when you are a student avoiding a bus stop or a parent worried about violence outside a school. It is different when disorder has become part of your daily life and the people explaining why nothing can be done live somewhere else.
Distance makes ideological purity inexpensive.
The business owners, students and communities of color we have met want public safety. They also want fairness, privacy and accountability. They do not experience those values as mutually exclusive. They want police to respond when called. They want protection from bias and misuse of authority. They also want the person repeatedly victimizing their neighborhood to be stopped.
I choose to listen most closely to the people who must live with the consequences of the policy.
This is the conclusion connecting all four parts of this series. Seattle understands its problems better than we admit. We have studied them, funded responses, and built many of the necessary tools. What we have avoided is confronting the political identity that prevents those tools from being used.
San Francisco is making progress because its leaders have become less interested in defending an identity than in changing conditions on the ground. They turned meetings into decisions. They built RESET when the existing choices were failing. They use technology to extend a depleted police force. They focus resources on the people and places where harm is concentrated.
Seattle does not need another philosophical debate about whether cameras can ever be consistent with progressive values. We have laws, retention limits, access controls, audits, and independent oversight. We can strengthen those protections as experience reveals weaknesses.
We should turn on the cameras.
Keep the existing system operating. Activate the cameras that have been approved. Expand their use in places where violence and persistent crime are concentrated. Publish the results and correct problems when they emerge.
The possibility of misuse is not an argument for refusing all use. Imperfection is not failure, and inaction is not harmless.
Progressive government cannot define itself only through the powers it refuses to exercise. It must prove that it can use power responsibly to protect people.
That is what Seattle is getting wrong. We know more than we act upon. We have evidence we would rather debate than follow. Too often, we allow our ideology to describe the city we wish existed while asking residents to endure the one that actually does.
The final lesson from San Francisco is the same as the first.
It comes down to leadership.
A leader must be willing to confront the problem, the opposition, and their own assumptions. They must be secure enough in their values to change a policy that is failing. Most of all, they must be willing to act and accept responsibility for the result.
Turning on cameras will not make Seattle less progressive. It will show that we are confident enough in our values to use modern tools responsibly.
The evidence is in front of us. The guardrails are within our control.
It is time to turn on the cameras.