Hispanic drivers in Maricopa County are still 40% more likely to be arrested than white drivers, stopped 30% longer, and searched 2.5 times as often, according to the ProPublica reporting on the ruling. The Maricopa County Sheriff’s Office complies with more than 90% of the reforms a federal court imposed on it.
Both things are true. That is the problem.
U.S. District Judge G. Murray Snow refused to end 13 years of federal oversight of the department, in a 72-page ruling that reads less like a legal opinion and more like an audit of what compliance actually buys a community. In his conclusion, Snow wrote that the record did not support finding that, absent federal monitoring, the department was unlikely to resume the conduct that necessitated the court’s orders in the first place.
Translation: the paint is fresh. The foundation is not.
On paper, drivers live in a jurisdiction whose sheriff’s office hits nine out of ten reform benchmarks. In practice, the traffic-stop data says Hispanic drivers are measurably more likely to be pulled over, held longer, and searched than white drivers. Those two facts are supposed to be incompatible. They are not.
The case that produced this oversight, Melendres v. Arpaio, began in 2007 as a class action over illegal traffic sweeps targeting Latino residents. It outlasted Joe Arpaio himself. His former chief deputy, Jerry Sheridan, now runs the department and spent much of the last year arguing that the oversight should end. Snow’s ruling describes a pattern of obstructing investigations and manipulating the internal misconduct process. Sheridan had already been held in civil contempt in 2016 for denying knowledge of a court order to stop making illegal immigration arrests.
That is the character portrait embedded in the ruling. It matters because the argument for ending oversight is essentially an argument about trust.
The department’s position, backed by Republican county supervisors, is that oversight has cost more than $200 million and produced diminishing returns. It is an appealing frame for taxpayers. It is also, according to a court audit surfaced by Arizona Luminaria and ProPublica, misleading. Auditors found that only about $63 million of the claimed $200 million had been appropriately charged to the settlement. Nearly 72% of the spending was misattributed or misappropriated. The sticker shock the sheriff’s office has been selling to voters is, in large part, its own accounting.
This is where the ordinary reader tends to check out, and where the analysis should slow down.
Those who have followed the case in local news for a decade can name Arpaio, remember the sweeps, and recite the contempt finding. What they cannot easily do is answer the question: if they’re complying with 90% of the rules, why is a federal judge still watching them?
The answer sits in the gap between process compliance and outcome compliance. A department can install every training module, file every report, and staff every review board the court asks for, and still produce racially skewed traffic stops. Snow’s ruling essentially says that the 10% the department is missing is the 10% that matters most: eliminating racial bias in stops, and investigating deputy misconduct with honesty. Everything else is scaffolding.
The disparity numbers themselves come from an investigation by Arizona Luminaria and ProPublica drawing on analysis by a University of California, Berkeley computer science professor. Hispanic motorists 40% more likely to be arrested. Stopped 30% longer. Searched 2.5 times as often. In a county with roughly 1.5 million Latino or Hispanic residents, that is not a rounding error. That is a policing pattern.
Raul Piña, a member of the court-mandated community advisory board, offered the cleanest summary in the reporting: you can paint the house, but if the foundation is broken you will have the same problem.
The metaphor is generous to the department. In some ways the foundation has been actively resisting repair. Snow’s finding that Sheridan, as chief deputy and now as sheriff, abused his authority over the internal affairs process is not a bureaucratic footnote. Internal affairs is how a department polices itself. If the person running it is manipulating misconduct investigations, then the reforms above that layer — the trainings, the data collection, the community boards — are running on a corrupted operating system.
Snow’s remedy is unusually blunt for a federal judge. He set a threshold: three continuous years of full compliance before oversight can be lifted. Not 90%. Full. The department has not yet cleared a single such year. The clock, in effect, resets whenever the monitor finds a shortfall.
This standard has quiet national implications. Maricopa County was an early testing ground for Immigration and Customs Enforcement’s 287(g) program, which deputizes local officers to enforce federal immigration law. Local police agencies have joined that deportation force in record numbers, often without the oversight infrastructure Maricopa was forced to build after being sued. Tricia McLaughlin, who served as a Department of Homeland Security spokesperson until she left the agency in February 2026, said at the time that agents do not racially profile or target Americans. The Maricopa data is the counterweight to that assurance: even a department under federal supervision, with a monitor, a community board, and a 90% compliance rate, has not managed to close the racial gap in stops.
For those who work on immigration cases in Phoenix, the ruling reads as vindication and warning at once. Vindication because the court is refusing to accept a clean-looking dashboard as evidence of a clean department. Warning because a case that started in 2007 is still running in 2026, and the sheriff is still fighting to end supervision rather than end the underlying disparities.
The KTAR News coverage of the federal monitor’s evaluations frames the story around cost and compliance percentages, which is how the department prefers to be measured. Snow’s ruling reframes it around the people the reforms were meant to protect. Those two frames produce very different verdicts on the same set of facts.
What the ruling cannot do is answer the harder question underneath it: how long is long enough. Thirteen years of oversight has produced a department that mostly follows the rules and still profiles the people the rules were written to protect. Three more years of full compliance might close the gap. It might not. The court is betting that continuous, uninterrupted adherence to the reforms, under the eye of a monitor who cannot be manipulated by internal affairs, will eventually change the behavior at the stop, not just the paperwork after it.
The uncomfortable read of the ruling is that reform, at least in this form, is not self-executing. It requires a judge willing to say no when a department asks to be released, and a monitor willing to fail a department on the metrics that matter even when the overall grade looks good.
A 90% compliance rate is not the same as a fair traffic stop. Maricopa County has spent 13 years learning the difference. Judge Snow’s ruling is a refusal to let anyone forget it.