The programme
How the Michigan police vehicle test works
What the Michigan State Police actually measure, how they measure it, and the three changes that make most published police car records wrong.
Every autumn the Michigan police vehicle test takes the coming year’s patrol cars to a proving ground and measures them. It is run by the Michigan State Police Precision Driving Unit. The results go into a book, and departments across the United States buy from it.
This site holds the results of 48 of those tests, transcribed. Before quoting any figure from them — from here or anywhere else — it is worth knowing what was actually measured, and when the measuring changed.
What gets measured
| Measure | What it is |
|---|---|
| Acceleration | 0–60 and 0–100 (and, in some books, the quarter mile), each the average of four runs rather than a best |
| Top speed | The highest speed reached on the track. From 1992 on it can be a governed or limited maximum rather than a capability — how that happened |
| Braking | Repeated 60–0 stops: two measured stops before the 1999 book and twelve from 1999, each sequence opened by heat-up stops; twenty from 2012, in two sets of ten with a cooling lap between. The scored figure is the average deceleration rate (in the 2002 book, of the best ten of the twelve stops); in the later books the distance is computed from it, while the earlier ones print the measured stops |
| Dynamics | Timed laps of a road course. The score is an average of the fastest laps across several drivers, not a single hot lap |
| Ergonomics and communications | Attributes graded by officers who sit in the vehicle and work in it — reach, visibility, where the equipment goes. Twenty-nine on the 1990 sheet and twenty-eight in the 1994 book, so the count varies from book to book; the 1978 ergonomics form alone has fifty-one |
| Fuel economy | Recorded in some years, absent in others |
The programme is a purchasing exercise, not a magazine test. Every score feeds a weighted total, and that total adjusts each manufacturer’s bid by a fixed amount per point, in the books that print the method. A vehicle that scores well can lose on price, and books from 1978 to 1987 name six vehicles eliminated or disqualified for missing a minimum standard. The 2015 book prints no scoring method, and the books from 2018 on print purchasing criteria but no category weights, weighted total or bid adjustment.
The arithmetic has changed too. The books held from 1978 to 1984 that print their method sum weighted Z-scores, each car measured against the field’s average, and turn the total into dollars at a share of the median bid (5% in the 1978, 1980 and 1984 books). From the 1985 scoring sheet on, the documents that print their method score each car instead by how far it falls short of the best car in each category, and those from 1985 to 2013 set the bid adjustment at 1% of the lowest bid. The weights have moved as well, so a total score from one era cannot be set against one from another.
The four things that break comparison
These are the reason most published police car records are not comparing what they think they are.
| Year | What changed | What it means |
|---|---|---|
| 2012 | Braking went from twelve measured 60–0 stops to twenty, in two phases of ten separated by a cool-down lap | A stopping distance from 2012 on comes from a different procedure, so compare it with an earlier figure only with that in mind. Neither the 2009 nor the 2012 book says which procedure is the more demanding |
| 2001 | The road course moved to Grattan Raceway after more than two decades at Michigan International Speedway | A pre-2001 lap is not a faster lap. It is a lap of a different circuit |
| 1999 | Braking went from two measured 60–0 stops, one after each set of heat-up stops from 90 mph (four in most books, three in the 1978 book; the 1980 mid-size class made its four from 80), to twelve, in two phases of six | A pre-1999 braking rate is the average of two stops, not twelve |
| 1992–2003 | Governors and electronic speed limiters spread from two vehicles to the entire field | A top speed may record what a manufacturer decided rather than what the vehicle could do — the full sequence is here |
Acceleration and top speed have also been recorded at more than one proving ground over the years — Chrysler’s for most of the programme’s history, Ford’s Michigan Proving Ground for the 1990 test, while a construction project closed Chrysler’s high-speed oval, and later FCA’s. Where a year page knows the venue, it says so. The road course at Michigan International Speedway was not one fixed length either: the 1978 book and the 1979 file give it as 1.82 miles, the year pages from 1981 on as 1.635.
Any police car record that spans 2012 without saying so is comparing twelve stops with twenty. Any lap record that spans 2001 is comparing two different circuits.
What the books do not record
Worth knowing before you go looking for it.
Horsepower. Where a book prints it, it is the manufacturer’s rating, labelled SAE net in some books, on the hardware pages; the bulletins and summaries usually list engines by displacement and fuel system only. It is a rating, not something Michigan measured.
Which standard a disqualified vehicle missed. The six vehicles the books name as out, in 1978, 1980 and 1987, all went out on acceleration, but only the 1987 results say which time was missed. Where a book does not say, this archive does not guess; it sets the printed averages beside the printed bars.
What departments actually bought. Michigan publishes what it tested and what it would pay. Purchasing decisions by other agencies are not in scope.
Anything about the vehicle in service. Durability, running cost and what breaks first are the things a fleet manager cares about most, and the test measures none of them. That is not a criticism — it is a two-day evaluation, not a hundred thousand miles.
What is in this archive
| Page | |
|---|---|
| Test results | Every vehicle ever tested, in one sortable and filterable table |
| Records | Fastest, quickest and shortest — split either side of the changes above |
| What gets tested | The field year by year, counted by body type. Six sedans in 1978, none in 2026 |
| Disqualified | The vehicles thrown out for failing a minimum standard |
| Speed limiters | When the top speed stopped being a measurement |
| Vehicles and manufacturers | The same data cut by subject |
| The year pages | Each book transcribed in full, with its own conditions and caveats |
Where the books come from, and what is still missing
The published reports are federal and state documents. This archive holds results for 48 tests, every one from 1978 to 2026 except 1979.
Not every year is held as a full book. Of the 1979 file only the title page, the method and the test schedules survive: no results. 1985 is held only as the scoring sheet, with the specification and the bid record, and several years only as the federal bulletin that summarised the results; each year page says which. A record set in a figure those documents do not print would not appear here.
Two of those gaps have been traced far enough to be worth recording. The 1985 test book is catalogued as NCJ 96486 in the National Criminal Justice Reference Service, which lists it but has not digitised it — so it exists in a known collection and is not simply lost. And a file circulating as the 1979 report is not one: it is NCJ 78761, the 1980 book, with the same title page and the same cars. Nine pages of the real 1979 file are now held — its title page, for tests on 28 and 31 October 1978, its method and its schedules — but its results are still the ones nobody has produced.
Before 1978 the record is thinner still. Michigan’s comparative programme began in 1976. The earliest book held here, the evaluation of the 1978 models written in November 1977, was already meant to be shared: its preface hopes the data “will be of use both to other law enforcement agencies in making decisions on their patrol car purchases, and to the vehicle manufacturers”. The department has indicated it does not hold complete reports from the 1970s — what survives of that decade is here.
If you hold one of these books, or know where one is, please get in touch. The 1979 results, or the full book for a year held here only as a bulletin or a scoring sheet, would close a gap this archive currently has to hedge.