Red light running, according
Red light running, according to the FHWA.gov website, killed 762 people in 2008. Out of 7,770 fatalities in intersections that same year - barely 9% of the total. Other studies have also shown that these cameras increase rear-end collisions at lights, including one by the New Jersey DOT which showed they were up 20% and total crashes went up by nearly a percentage point at intersections where the cameras are situated.
Further, the installation and cost of these cameras is not cost-effective. NJ estimates that they spent $1 million per year per intersection with four red-light cameras (one each direction) in maintenance, upkeep, and ticketing costs and did not recover that in fines.
So while I feel bad for some of these people who've suffered because of others running red lights, I'm sorry if that emotion doesn't extend to wasting taxpayer money on something that isn't proven to work. At the same time, it's been proven in studies in New York that by leaving a 1.5 second delay between one direction going red and the opposing going green, wrecks in intersections are reduced substantially. A simple fix that requires a one-time reprogramming of the lights and nothing more.