I had a spectacular time last night at Springfield’s quietly famous annual Flannel Shirt Party. Despite some controversy over the authenticity of my flannel I was allowed entry, and the next three hours were filled with drinking, dancing, merry-making, and great conversation.
The interaction which served as inspiration for this current essay was a conversation with a local reporter. We had met before and he knew me by name, though I had to be reminded of his. The most important takeaway, on my part, was his commitment to the city and his very accurate understanding of the primary issues therein. We did, not surprisingly, disagree a on few very significant points. One was his assertion that people are aware of the risks inherent in automotive travel, and another, somewhat corollary to it, that precisely because people are already acutely and accurately aware of said dangers, it is not incumbent upon the media to make them aware. Beyond that he was unaware that, statistically speaking, a city which has a raw number of homicides in or around the mid teens is not unlikely to experience two unrelated homicides on “one day” every few years given the fluctuations in the yearly number and, well, how numbers work.
To the first point, people are horrible at understanding risk. Rather than reinvent the wheel here, I am going to link to my previous posts on these topics, and add a little air in the tires as I see fit.
Our understanding of risk is primed by our perception of availability. Unfortunately, our perception tends to be wrong. Parents often misunderstand risk and make poor decisions. The most famous being miscalculating stranger danger, and helmet use, but, in relationship to urbanism, it is simply not knowing that two of the most significant dangers to teens, car crashes and suicide, are positively correlated to suburbia and negatively correlated to density. If your teen isn’t in a gang and doesn’t buy drugs, then they are safer in Six Corners than Suffield.
Where I think anyone would have to admit a flaw in logic in last night’s conversation was our disagreement over the ordinal number always attached to murders in the city, but never attached to car related deaths in the region. One never sees an article on a Springfield murder without the crime being categorized as the 1st, or 8th, or 15th of the year. When I asked why such numbers were never attached to automobile related deaths I was told it was because the public is totally and completely AWARE of the dangers inherent in car travel. While I continue to dispute that, let’s grant it for sake of this part of the argument; that would mean that assigning an ordinal number to homicides is done because the public is NOT aware of the dangers of gang activity and drug sales in an urban environment or of the dangers of city living in general. Even the reporter doesn’t believe that, and by the transitive property, it simply makes no sense. If “A” need not be assigned an ordinal number BECAUSE “the public is aware of the danger”, then “B” need not be assigned an ordinal number IF “the public is aware of the danger”.
The reporter went on to claim that it isn’t the media’s responsibility to educate people on the dangers of car travel. My response now as last night is simply that it is precisely the media who must do that and, I would hasten to add that it is the media who are responsible for the lack of understanding in the first place precisely because, for example, homicides are overly contextualized to place and not behavior, and car crashes are overly contextualized to behavior and not place!
On the final statistical piece regarding homicides, or any randomly occurring events, when those events take place in a year they become likelier than not to make a pair occurring on the same day when the number of events reaches 23. At 16 occurrences per year (Springfield’s average number of homicides over the last decade) there is nearly a 30% percent chance that two of them will fall on the same day in an average year. If you expand the definition of day to mean NOT “the same date” but any 24 hour period selected from the starting point of a randomly occurring event only after the event has taken place(as was done by the reporter in the case in question), then the percentages fly upwards beyond 50%. It is meaningless.