I read/check Shinar’s book “Traffic Safety and Human Behavior” almost daily. Two figures caught my attention a few weeks ago. Figure 1-12 (page 28) and Figure 1-13 (page 29) were reproduced from SARTRE 4 report. If you compare the proportions in these two figures, you will see they do not match (see my Figure A below). So I ordered SARTRE 4 data and found out that the authors had very likely used a wrong variable in Figure 1-13. I got a perfect match when I used the variable Concern about the rate of crime.
The original SARTRE 4 figures can be found in the final report in Chapter 1.2 Car Drivers’ General Attitudes, Beliefs and Reported Behaviors. Figure 1 is on page 47 and Figure 2 is on page 49. They are also below this paragraph. Please note that for Figure 2 “The medians of the variables were used to establish four quadrants distinguishing low vs. high concern and low vs. high road safety.”
Figure A. A mismatch between two figures in SARTRE 4 report and Shinar’s book
Since I got the data, I made a figure using the correct variable. That is the left figure. The right figure contains road fatalities per 100,000 inhabitants in year 2021 according to OECD.
Figure B. A correct figure produced by using SARTRE 4 data (on the left) and a new figure that plots 2021 fatality rate data (on the right).
Given that the authors write in SARTRE 4 report (page 48): “Assuming that a particular amount of concern in the car driver population might be essential for the acceptance and effectiveness of road safety policies and measures…” I made a figure that shows how much traffic safety has improved in terms of fatalities per 100,000 inhabitants between 2009 and 2021 in the relation to reported personal concern about road safety in 2010. You can make your own conclusions whether the concern about road safety in 2010 was a good predictor of the actual road safety in 2021. I will only say Heja Sverige!
Figure C. Road fatalities per 100,000 inhabitants in 2009 vs. 2021 in the relation to reported personal concern about road safety in 2010
Two notes: (i) this data contain personal concern about road safety among drivers only; (ii) perhaps I should have used a three-year average for new data (2019-2021), but this was already too much work for me;-)
And yes, if you find a mistake in my figures please let me know. It seems easier to spot others’ mistakes than one’s own;-)
