Figures don't lie, but liars figure

[unixronin]

Interesting, if brief, reading on highway speed limits from the WSJ:

We now have 10 years of evidence proving that the only "assault" was on the sanctity of the truth.  The nearby table shows that the death, injury and crash rates have fallen sharply since 1995.  Per mile traveled, there were about 5,000 fewer deaths and almost one million fewer injuries in 2005 than in the mid-1990s.  This is all the more remarkable given that a dozen years ago Americans lacked today's distraction of driving while also talking on their cell phones.

(Emphasis mine.)

Comments

[beckyzoole.livejournal.com]

The news of lower death rates is very, very good. The article rightly attributes it to safer cars, safer roads, and the anti-drunk-driving campaign.

However, there is nothing in the facts and figures presented that correlates fatality rates and speed limits. In other words, from what I see in these charts, it is possible that the decline in fatalities would have been even lower if speed limits were lower. And it is certainly possible that speed limits have no significant effect on fatality rates -- it's just that these figures don't address that. I could just as well say that carrying a concealed weapon in one's car has no effect on motor vehicle fatalities. It probably doesn't, but these figures don't address that variable at all.

What I would like to see is how speed affects motor vehicle accident rates, and how it affects fatality rates when there is an accident. I'd like to see that one variable isolated from air-bags, drunkenness, and weather... and then, perhaps, combined with those factors.

[unixronin.livejournal.com]

I saw an interesting chart some years ago in Car & Driver. It was in an article talking about exactly this. The chart was a plot of annual road fatalities in the US, year by year, with the beginning and end of the 55 limit plotted on it.

It was a very interesting chart, because the trend was so visible. It was basically a straight line, tending downward year by year. Except between the boundaries of the 55 limit, where there was a flat-topped hump. There was a jump in the rate when the 55 limit went in, then it resumed its monotonic decline, parallel to the old line, but a little higher. Then when the 55 limit ended, it fell pretty much back to the old line.

Plotted out that way, that chart all by itself presented a pretty clear argument that far from saving lives, the 55 limit actually cost lives. The bump has been attributed to two main causes. A lesser factor in it is that, with no speed advantage to interstates, people were abandoning the interstates for secondary roads to avoid traffic congestion. But the major factor, as reported by state highway patrols, was "single vehicle accidents due to driver inattention." In other words, out in the big Western states, it was now taking as much as several hours longer to drive from city to city, and drivers were quite simply nodding off at the wheel from sheer boredom and fatigue.

"55. It's not even a good idea. It's just the law."

[ithildae.livejournal.com]

I also saw an article, many years back, about how much safer it is when most cars are traveling close to the same speed. The 55 speed limit is the most disobeyed federal law since prohibition (my opinion, as a driver) producing larger differences in speeds. With the faster speeds, I see dramatically fewer idiots weaving in and out of lanes trying to get through traffic "clumps". (I realize that this is anecdotal, but there it is.)

unixronin has already mentioned single car accidents, especially out west.

The funky thing about statistics is that, if we don't like what they are showing us, there is always something that could be added, so we don't have to accept the conclusions. (I find myself doing that more often that I want to admit.) Statistics is what it is. In this case, I think you are demanding more than statistics can provide.