The substitution claim carries the argument, and it is separable from the price claim. The officer's account of imprecision is concrete: he pulled over a car because it matched the suspect's color, then spent time checking names and licenses while the person he wanted was somewhere else [12]. Against that he sets a hit and run he worked from a model, a color and an approximate time [13]. If those two are genuinely alternatives for the same case, the camera is the narrower intrusion; if they are additive, a department has bought a new capability and kept the old habit.
Precision describes the query, not the collection. A read is checked against the FBI's national file of stolen cars and wanted persons, and a hit pushes an alert to officers already in the area [3]. In the 2004 Project Laser trial, cameras read more than five million British plates and produced 3,071 arrests [8], which is roughly one arrest per 1,600 plates read [15]. Everyone else in that five million was scanned in order to find those people, so the trade on offer swaps a small number of in-person encounters for a very large number of logged reads the motorist never notices.
Price is the strongest part of the case, because it is checkable. A camera runs $3,000 a year against a typical department budget of $1 million [4], so one camera is 0.3 percent of that budget and the whole budget would cover about 333 camera-years [14]. That arithmetic matters most for the agencies that could never fund fixed or vehicle-mounted installations: almost half of US police departments employ fewer than ten full-time officers, spread across four million miles of public road [6]. With almost 30 percent of American agencies already subscribing [5], the binding constraint moved from procurement to policy some time ago.
A skeptic would say the pretextual stop does not go away because the camera arrives, and that the two accumulate. The honest answer is that this account does not settle it, because it reports one officer's practice and no measurement of stop rates before and after deployment [16]. The control it does describe is a case number and a stated reason for every search, compiled monthly and validated by department administrators [11]. That is a real record, and it is an internal one; nobody outside the department appears in the validation loop.
The sequencing deserves attention apart from the merits. A lookup can be filtered by vehicle description and by timeframe, run from the scene [10], which means much of the value of a subscription signed this quarter sits in the passes it records before anyone has a case to open. A council approving a $3,000 line item is therefore approving a retrospective search capability that its own future investigators, and anyone who can compel the data, will run against residents who were never suspected of anything.
