Figure 4. Car drivers killed and seriously injured in the UK, 1969-1984. From:Harvey and Durbin (1986)
Car passengers were further split into front seat passengers who were directly affected by the law and rear seat passengers. Applying statistical structural time-series models the authors were able to model the series for the before period and estimate the effect of the law by comparing the actual fatality and serious injury figures with their estimate of what those figures would have been without the introduction of the law. Their estimate of the effect was a significant reduction of 23 per cent in the numbers killed and seriously injuries for car drivers and 30 per cent for front seat passengers. For rear seat passengers, however, a small increase of 3 per cent was found. For the number of persons killed similar large reductions in the number of fatalities were found for car drivers and front seat passengers, but the number of rear seat passengers killed increased by 27 per cent. The authors could not find a satisfactory explanation for this increase. One explanation put forward was that there was the possibility that after the introduction of the law, which did not affect rear seat passengers, a certain number of passengers transferred from the front seat to the rear seat.
It is exactly this type of question and outcome that could have been answered in detail, if at the time a series of exposure measurements had been in place counting, over time, the number of passengers in front seats and in the rear of cars.