The thing that annoys many people is the fact that in order to make cash from them, they use the excuse that there have been 'x' accidents in that location over the past 'y' years. Now what they never say is what the accident level was before that spate of accidents.
So if, for example, there were 3 accidents over a 2 year period on what should be a safe piece of road, they might put a camera in. There may have been no accidents there over the 2 (or even more) years before that period though. Then, after putting in the cameras there might be no accidents for the next 2 years. The claim will be that the cameras have reduced the accidents from 3 to zero over a 2-year period. What has happened in fact is that the accident rate has just returned to what it was 4 to 6 years ago. It is called 'regression to the mean'. They could have put a garden gnome beside the road instead of a camera and the accident rate would still have regressed to zero over the same period.
How many of you know that the Road Transport Research Laboratory, a Gov't funded professional investigation group, have established that only between 4% and 5% of all road traffic collisions (RTC's) involve (not necessarily are caused by) vehicles being driven in excess of the posted speed limit? Thus around 95% of RTC's do not have speed in excess of the posted limit as a factor in their causation. The main causes are inattention, inappropriate speed for the conditions, driving to close or carelessly and failure to judge the path and speed of other vehicles. In other words, simply poor driving. Driving will not be improved by simply 'driving by numbers', the numbers being those in black on the white circle with the red outline!
That is why cameras are a cash-raising scam.