San Francisco did not need another coordinating body to coordinate the existing coordinating bodies. It needed the people already at the table to make decisions.
Seattle and this region have rarely encountered a hard decision that could not be handed to a workgroup.
When the answer is politically difficult, we study the question. When a study confirms what we already know, we appoint a task force to recommend how its findings might be implemented. The process creates the appearance of movement while postponing the moment when someone must choose a direction.
Workgroups have a legitimate purpose when the problem is new or the answer is unclear. Too often, however, they are used to distributing the risk of decisions that government already knows how to make. The issue is no longer a lack of information. It is a lack of willingness to act on it.
Seattle does not need another study to tell us that open-air drug use is dangerous, that fragmented responsibility weakens our response, or that too many people are cycling between the streets, emergency rooms, and jail without receiving sustained help. We have studied these problems from every angle. Another report may describe them more precisely, but it will not make the decision for us.
San Francisco has not abolished workgroups. Under Mayor Daniel Lurie, it has changed what they are expected to do.
The workgroups are for working.
Their purpose is not to admire the complexity of a problem or produce another set of recommendations. They exist to make decisions, identify who owns the next action and determine whether conditions changed as a result. A meeting is not evidence of progress. It is where responsibility should become impossible to avoid.
Part One of this series examined the leadership required to choose a direction. This article is about what happens next. Once leaders decide that failure is no longer acceptable, every coordinating body must be converted from a place where government talks about the problem into a place where government acts on it.
The clearest example from our trip was San Francisco’s Drug Market Agency Coordination Center, better known as DMACC.
Former Mayor London Breed deserves credit for creating DMACC in 2023. It brought together police, the sheriff, prosecutors, public health officials, homelessness agencies and emergency management. State and federal partners also participated. San Francisco even published weekly arrest and drug-seizure data.
San Francisco did not lack a coordinating body. It already had one.
Yet people we met described DMACC’s earlier culture as closer to an anemic workgroup than a command center. Agencies exchanged information and reported on their individual activities. The meeting did not consistently produce decisions about what should happen next.
Mayor Lurie did not implement the typical playbook and create another coordinating body to coordinate the coordinating body. He made DMACC permanent and changed what the city expected from it.
The mayor’s office began treating it as a command center. Meetings followed a regular cadence and worked from a shared picture of street conditions. The agenda, set by the Mayor, focused on the next action. When the group met again, the question was not whether everyone had remained busy. It was whether anything had changed.
A workgroup observes the handoffs. A command center owns them.
That distinction changes what government measures. Public institutions naturally gravitate toward activity because activity is easier to count and defend. An agency can report the meetings it held and the emails it sent. It can tally referrals, outreach contacts and hours worked. Every number may be accurate while the person remains in crisis, and the block continues to deteriorate.
Activity tells us what government did. Outcomes tell us whether it worked.
Throughout our visit, every meeting connected to the mayor’s priorities had clear goals and specific measures. That remained true even when no one from the mayor’s office was present. People across agencies described the work with the same sense of purpose. They understood what the administration wanted to accomplish and which conditions would show whether it was working.
That consistency was a measure of leadership in itself. The direction from the top had traveled through the organization.
DMACC also benefits from a deliberately focused mission. It is not charged with “solving homelessness,” a challenge that encompasses everything from housing costs and mental illness to addiction, poverty and decades of public policy. Its mission is focused on coordinating enforcement, outreach, and treatment around open-air drug markets while creating pathways to recovery.
That narrower focus does not deny the complexity of homelessness. It prevents complexity from becoming an excuse for doing nothing at a particular place on a particular day.
Homelessness itself is not a crime. But when someone declines housing or treatment, the city cannot simply walk away. If that person remains in danger or their conduct prevents a public space from functioning as intended, the city still has a responsibility to act.
DMACC asks governable questions. What conditions are we trying to change this week? Who owns the next action? What will the evidence show when we meet again?
Our conversation with District Attorney Brooke Jenkins revealed how difficult that becomes once the work extends beyond City Hall. However, Jenkins began with a clear premise.
Everyone deserves a safe and clean community.
She also described a justice system that is not really one system at all. An arrest begins a chain that includes police officers, prosecutors, public defenders, judges, treatment providers and correctional agencies. Each institution has its own authority and culture. Each has its own definition of success. A failure anywhere along that chain can undo the work everywhere else.
Jenkins’s comments about judges were especially interesting. Judges rarely appear in public debates about street conditions, yet they make some of the most consequential decisions in the process. Their rulings influence whether an intervention interrupts a destructive cycle or merely pauses it.
Judges came up in several other conversations as well. One of the strongest criticisms we heard came from someone who had previously been incarcerated. In his view, judges were among the greatest obstacles to meaningful change.
That observation does not resolve the question, nor does it mean every judge is responsible for the system’s failures. Judicial independence and due process are fundamental constitutional principles, and every case turns on its own facts. But his comments revealed something important.
The most influential parts of a system are often the least visible to the public.
Police can count arrests. Prosecutors can count cases filed. Courts can count matters resolved. Providers can count services offered. Every institution can meet its own target while the same person continues cycling through the system, and the community continues living with the same conditions.
If the public outcome does not improve, institutional activity is not enough.
A mayor cannot direct an independent judge or order another elected official to act. Leadership across institutions does not mean pretending those boundaries do not exist. It means refusing to allow every institution to retreat behind its own boundary while the collective result remains unacceptable.
This requires a broader conversation about what the justice system is designed to produce. The goal cannot simply be more arrests, prosecutions or referrals. It should include fewer overdose deaths and fewer repeat encounters. More people should move towards recovery. Public spaces should function for everyone who depend on them.
Once those outcomes are clear, every institution should be able to explain how its work contributes to them.
This is where a traditional workgroup often fails. It gathers representatives from every institution and gives each of them time to describe their constraints. The meeting ends with a better collective understanding of why the problem is difficult. Everyone agrees to continue coordinating.
A command center does something different. It identifies the exact point where the response is failing. If the barrier is legal, it names the law. If the barrier is money, it calculates the cost. If capacity is missing, it defines what is needed. If the problem is that someone has not made a decision, it makes that visible too.
The objective is not to eliminate disagreement. It is to stop using disagreement as a substitute for a decision.
The Caltrans agreement and Embarcadero cleanup I described in Part One are not isolated anecdotes. They reflected the same operating rules. Jurisdictions may determine who has authority, but they cannot become permission to walk away. Boundaries may divide responsibility, but they cannot be allowed to dissolve it.
Mayor Lurie’s policy chiefs are important for more than their titles. They keep problems from disappearing between departments, follow issues across government, identify the next decision, and remain accountable until it is made.
Too often, we mistake leadership for a box on an organizational chart. Real leadership is the person who refuses to leave the room without an answer.
Seattle has no shortage of task forces, advisory groups, or reports. Many have produced thoughtful analysis. Some have helped government reach better decisions. But the existence of a workgroup is not evidence that the work is getting done.
We should expect every public safety meeting to begin with the condition government is trying to change. It should end with an action, an owner, and a deadline. When the group meets again, the first question should be whether the action occurred and what changed as a result.
If another jurisdiction controls the next step, its leaders should be brought into the conversation. If an independent institution is producing an outcome inconsistent with the broader public goal, that tension should be discussed openly. Coordination should make responsibility more visible, not less.
Not every meeting can produce an immediate decision. Some questions require public engagement, legal analysis, or additional evidence. But every workgroup should be moving towards a decision that someone has the authority to make. If no one can identify that decision, the group may exist only to absorb pressure that should be directed towards a leader.
San Francisco has not solved every breakdown in its system. Its agencies still disagree. Courts remain independent. State and local responsibilities still overlap. The difference is that the city is treating those boundaries as the beginning of the work, not the end of it.
That approach also revealed a critical gap. Officers confronting public drug use still faced two inadequate choices. Walk away or spend hours processing a jail booking that often did little to change the outcome.
DMACC gave the city a place to identify that failure. The RESET Center became one of its first major responses and opened within ten months of the decision to proceed. It gave officers another option and gave people in crisis somewhere more appropriate to go.
That is the subject of Part Three of this series.
Government is not a workgroup. If coordination does not change the outcome, it is only a more organized form of failure.