Articles

Research

Findings from the test books — things that only show up when you read forty years of the same evaluation in order.

Michigan has tested police vehicles every autumn since 1976, and published the results every year since 1979. Read one book and you learn what won. Read thirty-six of them and you start to see the shape of things nobody set out to record: measurements that quietly stopped measuring anything, a phrase the industry uses that the department rejects, a pattern in one vehicle’s brake sheets that took four years to become undeniable.

These pieces come out of that reading. Each one is built on figures transcribed from the original reports, links to the test years it draws on, and states plainly where a claim is interpretation rather than source.

Testing and the record

What the Michigan books actually show — and the four places where a number everybody quotes turns out not to mean what it looks like.

Owning one

The two pieces here change a decision a reader might make this month.

  • Michigan blind-tested police brake pads

    Twenty-eight makers were asked for their severe-duty police pads. Some failed the federal standard in the laboratory. One of the survivors lost fifty feet of stopping distance in ten stops.

  • The most dangerous thing in the trunk

    Patrol cars caught fire often enough that Tucson’s crime lab and fire investigators ran experiments to find out why. The answer was the road flares — and every alternative tested was less visible.

Lighting and visibility

Five studies across forty-two years, all reaching versions of the same uncomfortable answer. They cross-link and are meant to be read in order.

The police lighting sequence

1978: a federal survey calls the field chaotic. 1982: Illinois measures what a light bar costs and takes sixty off. 1984: those cars come back with a fifth the accident rate. 2020: Virginia Tech finds that brighter lights hide the officer standing beside the car. And underneath all of it, a problem no other emergency vehicle has — a police car has to be conspicuous and inconspicuous in the same shift.

Read the sequence →

Pursuit

Two federal studies twenty years apart: one that could not find the records, and one that finally counted.

  • Sixty-eight thousand police pursuits a year

    American police chased 68,000 fleeing vehicles in 2012. Over twenty years, pursuit crashes killed more than 7,000 people — and a third of them had nothing to do with the chase.

  • The chases nobody counted

    When federal researchers asked 737 American police agencies for their pursuit records in 1994, more than a third could not supply them — because they had never collected any.

Money and fleet

The least glamorous documents here, and the ones that explain why the cars you can still buy are in the condition they are.

  • What a police car was actually worth

    A federal study priced every part of owning a patrol car. The police package cost $400, the big engine cost $67, and a big-city cruiser lost 87% of its value in two years.

  • When is a police car finished?

    An Idaho State Police lieutenant wrote the fleet manager’s side for the FBI in 2002: when to replace a patrol car, whether to lease it, and what actually breaks first — which is never the engine.

  • 343,758 police cars

    The federal government counted every vehicle operated by every local police department in America, and asked what was inside them. Four in ten departments serve fewer than 2,500 people.

  • How many police cars should there be?

    Patrol consumes more than half of most police budgets. In 1975 RAND wrote a program to work out how many cars each command needed — and it ran in 240 kilobytes.

Looking for the longer narratives? They are under History.

Sources. Every figure in these articles is transcribed from a Michigan State Police, National Institute of Justice or NLECTC publication and linked to the relevant year in the test archive. See how these videos are researched and corrections.