The numbers behind Katie Wilson’s favorite public-safety program don’t add up

For nearly 15 years, Seattle has been experimenting with a different way to deal with people repeatedly cycling through drug use, homelessness and the criminal justice system.

Law Enforcement Assisted Diversion (LEAD) launched in Seattle in 2011. Instead of arresting and prosecuting some people accused of low-level crimes, police can divert them to case managers who help connect them with housing, treatment, benefits and other services. It began with low-level drug and prostitution offenses and has since grown into a much broader response to people whose criminal behavior is connected to addiction, mental illness or poverty.

There is an understandable appeal to the idea. If someone is arrested, spends a short time in jail and returns to the street with the same addiction and nowhere to live, very little has been accomplished.

But LEAD takes that philosophy further than many traditional diversion programs. It does not require participants to stop using drugs, enter treatment or meet mandatory conditions to continue receiving help.

That has made LEAD increasingly controversial as Seattle has leaned on it more heavily to address public drug use and disorder.

Mayor Katie Wilson has made clear which side of that debate she favors. During her campaign, Wilson called LEAD an “evidence-based” framework, included it under a plan to scale up “proven programs” and proposed restoring its ability to accept referrals citywide.

Then, just after Wilson took office, Seattle Police Chief Shon Barnes told officers that drug possession and public-use charges would be diverted from prosecution to LEAD. The announcement caused an immediate backlash. Wilson subsequently said police could still enforce drug laws and denied that she announced a sweeping policy change – but she continued to say she wanted the LEAD framework implemented.

Wilson’s support rests on an important premise: LEAD works.

A new report from the Washington Policy Center raises serious questions about how convincingly that has ever been established — and, more troublingly, whether Seattle residents have been given a much rosier picture of LEAD’s results than the research actually supports.

One of the most repeated claims was that LEAD participants were 58% less likely to be arrested.

To an ordinary person, that means something pretty clear: Put 100 people into LEAD, and you should expect dramatically fewer of them to get arrested again.

But that is not what the research showed.

The 58% number came from a complicated calculation about the odds of arrest. It was then repeatedly translated into much broader claims that LEAD reduced arrests by 58%, reduced recidivism by 58% or made participants 58% less likely to commit another crime. Washington Policy Center found versions of those claims repeated from 2015 through 2026.

Those claims sound similar. They are not.

When you simply look at the people in the study, the rough difference in arrest rates between the LEAD group and the comparison group was an estimated 22 percentage points, not 58. And even that cannot simply be credited to LEAD because the two groups were treated differently by police, entered the study in different ways and were not clean apples-to-apples comparisons.

In other words, an uncertain finding became a remarkably confident talking point.

The employment claim is even harder to defend.

LEAD supporters touted a 46% increase in participants’ chances of moving along an “employment continuum.” That sounds as though the program was helping large numbers of people get back to work.

Look at the actual people, however, and the story changes completely.

Of 176 LEAD participants, researchers estimated about 13 had jobs when they entered the study. During the following 18 months, about 16 were employed in an average month.

Thirteen became 16.

About 91% were still unemployed on average.

So where did the impressive 46% come from? Researchers were not measuring only whether someone got a job. They created an “employment continuum” that also counted things such as vocational training and even retirement. That broader category was then presented with a large percentage that sounded far more impressive than the real-world employment results.

Whatever the original intent, repeatedly presenting the results this way had the effect of hiding modest or uncertain outcomes behind much more dramatic numbers.

And that is why this report matters.

This is not an argument over a rounding error. Those big percentages helped build LEAD’s reputation as a proven success. Washington Policy Center found the claims repeated for more than a decade, including in government materials and, most recently, a Seattle City Council agenda packet supporting LEAD expansion.

That reputation matters when elected officials decide where to spend money and how police should respond to people using drugs in public. It matters even more when a mayor describes a program as “evidence-based” while seeking to expand its role in the city.

Success should mean fewer people returning to crime, more people overcoming addiction, more people moving into stable housing and more people becoming healthy and independent. Those results should be reported in plain English, without relying on impressive percentages that tell a very different story once you look underneath them.

Wilson promised a public-safety approach built around programs that work. If LEAD is one of them, Seattle should be able to prove it.

And if it isn’t providing real, lasting help to the people who need it, the city needs to be willing to try something that does.