Speed kills saves

July 21, 2006 11:00 AM

Safe at Any Speed With higher speed limits, our highways have been getting safer.
posted by caddis (79 comments total)

message to self: remember, html tags do not work in Post Titles.

That should be "kills"
posted by caddis at 11:02 AM on July 21, 2006

Damn, the number of pedestrian deaths is huge!
posted by Ironmouth at 11:06 AM on July 21, 2006

But pedestrians aren't, you know, real people.
posted by hattifattener at 11:11 AM on July 21, 2006 [1 favorite]

God, I remember when they rased the speedlimit here from 65 to 70. My state senator was quoted as saying "This change will mean more blood on Iowa highways" Gag.

That said, deaths per year is a terrible metric for trying to figure out if these speed limit changes have fixed things. Car safety has become much better over the past ten years as well, so a better measure would be the number of accidents. Have accidents gone up or down?
posted by delmoi at 11:12 AM on July 21, 2006

Yes, I reckon it's car safety that's changed considerably. C'mon, is that all they got?
posted by basicchannel at 11:14 AM on July 21, 2006

Well, they've also got a comparison intended to make Iraq combat deaths look insignificant.
posted by rxrfrx at 11:15 AM on July 21, 2006

here-here, delmoi. I was wondering why the article didn't mention car safety technologies and requirements, nor did it mention raw accident numbers.
"Sorry, sir, we know you're not dead — "just" a quadrapalegic — but that doesn't register on our statistical radar."
posted by I, Credulous at 11:16 AM on July 21, 2006

yeah, and the inclusion of the Iraqi combat deaths was, um, odd.
posted by I, Credulous at 11:16 AM on July 21, 2006

Someone needs to be forced to write "correlation does not equal causation" 100 times on the blackboard.

Remember, folks, ice cream causes murder!
posted by Johnny Assay at 11:17 AM on July 21, 2006 [1 favorite]

The number of accidents per year have gone up--some 38,000 in 2004 versus 36,000 in 1994. But the WSJ is correct in asserting that fatalities per 100 million miles have declined.
posted by Iridic at 11:22 AM on July 21, 2006

I found it interesting that the United States is one of the most tolerant for driving drunk. Most of the Western world allows 0.02-0.05% BAC, while the U.S. and Canada's typical 0.08% seems to be among the highest. If anyone seriously cares about saving lives, seems fixing that limit would be one place to start. Who the hell needs to go drive after slamming six beers?
posted by zek at 11:24 AM on July 21, 2006

You have to be a pretty big boy to slam six beers and not blow a .08.
posted by caddis at 11:27 AM on July 21, 2006

Anti-lock brakes, electronic slip protection, multiple airbags, better crumple zones on cars, increasing popularity of all-wheel drive, better fog lights, more light reflective decals on the highways...

And the list goes on. The Journal needs to stick to what it seemingly does best: report financial news.
posted by SeizeTheDay at 11:28 AM on July 21, 2006

The chart in the article does include the number of crashes, which, according the the WSJ's source, declined 33% between 1995 and 2005.
posted by monju_bosatsu at 11:29 AM on July 21, 2006

Somebody should crank up some Sammy Hagar, I guess.
posted by jonmc at 11:29 AM on July 21, 2006

You have to be a pretty big boy to slam six beers and not blow a .08.

Depends on what you slam them into.
posted by oaf at 11:30 AM on July 21, 2006

Isn't the Autobahn the safest stretch of road in the world? I thought speed was all relative, especially considering modern cars have no problem accelerating to high speeds. Besides I've seen more accidents in construction zones and at 40MPH bumper to bumper traffic than I do on long stretches of highway.
posted by geoff. at 11:34 AM on July 21, 2006

Tthe article _does_ show that "crashes" have decreased overall. The chart on the article shows the number of crashes per 100 million vehicle miles traveled at 560 in 1995 and 375 in 2005.

So that doesn't mean the highway speeds had anything to do with it, just pointing it out for those who say the article only relied on number of deaths.
posted by rsanheim at 11:34 AM on July 21, 2006

It looks like it is just comparable to other stretches of highway that do have limits. Personally I find it inane on some stretches of road that I'm forced to go 70MPH when I don't see a single car in either for direction for several miles.
posted by geoff. at 11:37 AM on July 21, 2006 [1 favorite]

I found it interesting that the United States is one of the most tolerant for driving drunk. Most of the Western world allows 0.02-0.05% BAC, while the U.S. and Canada's typical 0.08% seems to be among the highest. If anyone seriously cares about saving lives, seems fixing that limit would be one place to start. Who the hell needs to go drive after slamming six beers?

It seems that way? Seems how? Do you have any evidence that lowering the BAC limit would actually have any lives? How many deaths per mile are caused by drivers with a BAC of 0.08% or less, compared to people who have not drank at all. Lowering the BAC level would basically make it illegal for some people (namely women) to have even a single drink and drive home.
posted by delmoi at 11:41 AM on July 21, 2006

Isn't the Autobahn the safest stretch of road in the world?

If so, it's an even greater victory for the refrain "correlation does not equal causation". If the Autobahn *is* so safe, I'd be inclined to credit extremely tough standards for both drivers and cars. A German driver's license is expensive and relatively difficult to get, and the roads in Spain and Eastern Europe are full of German cars that can no longer pass the demanding TÜV tests.
posted by Slothrup at 11:50 AM on July 21, 2006

It is hard to see how increasing the speed limit could decrease the number of accidents per mile. I could understand it decreasing the number of accidents per minute, since driving faster allows you to spend less time on the road, but a per mile decrease seems like it has to be attributable to something else.
posted by gsteff at 11:56 AM on July 21, 2006

Lowering the BAC level would basically make it illegal for some people (namely women) to have even a single drink and drive home.

Trying to frame a lowered legal BAC level as sexual discrimination is ludicrous.
posted by oaf at 11:57 AM on July 21, 2006

You have to be a pretty big boy to slam six beers and not blow a .08.

I guess that I'm a pretty big boy as I can drink six beers in two hours and still only be at 0.0726 acording to this calculator. Mind you, I haven't had that much to drink in fifteen years but if I did, I'd still be legal.
posted by octothorpe at 11:57 AM on July 21, 2006

But to Westerners with open spaces and low traffic density, the [55 MPH national speed limit] became a symbol of the heavy hand of the federal nanny state.

I notice no one is complaining about the "heavy hand of the federal nanny state" when it comes to regulations concerning the availability of seat belts, passive restraints, side-impact requirements, bumper rules, government crash tests, catalytic converters, etc., or for that matter when it comes to the design and construction of new roads and the maintenance of old ones.

The auto industry itself has a long track record of resisting regulations, including safety regulations of almost every sort. Modern cars are relatively clean and safe because government mandate requires them to be. The automobile industry is pretty much a perfect poster child for the value of government regulation. This being the WSJ, that aspect of the story is completely ignored.

Face it, WSJ: Without the government's involvement, cars would not be a workable system of transport for the masses at all, and without extensive "nanny state" rules, they would be very much dirtier and more dangerous than they are.
posted by Western Infidels at 11:58 AM on July 21, 2006 [2 favorites]

re: the Autobahn, it's also obsessively well-maintained. not a good comparison.
posted by paul_smatatoes at 12:01 PM on July 21, 2006

It is hard to see how increasing the speed limit could decrease the number of accidents per mile.

I think the only thing it refutes is the dire predictions of increased traffic fatalities which were going to follow the repeal of the 55 mph speed limit. They didn't happen. Why? This article is not going to answer that. Cars are better, but not that much, and the difference in speed is not that significant on an open highway (although it is significant on crowded ones, but most of those never went to 70 mph). Probably both are factors.
posted by caddis at 12:03 PM on July 21, 2006

Sorry, rsanheim, I guess I'm guilty of some kind of web-based visual fatigue, I think: I tend to assume (wrongly in this case!) that things stuck like that into the middle of the article are ads and hence ignore them completely. Thus I missed the chart with the stats in it.

See what the web ad has done to several decades of magazine design vocabulary?
posted by I, Credulous at 12:04 PM on July 21, 2006

They don't actually say that the higher speed limit is safer. The real thesis is just that "Lots of people said this would drastically inrease deaths. It didn't" However, they do go back and forth between carefully wording things to avoid false implications:

Two studies, by the National Motorists Association and by the Cato Institute, have compared crash data in states that raised their speed limits with those that didn't and found no increase in deaths in the higher speed states.

And making flagrantly misleading statements:

Of the 31 states that have raised their speed limits to more than 70 mph, 29 saw a decline in the death and injury rate

They also imply that the 1994 opponents of the higher speed limits were deliberately lying (We now have 10 years of evidence proving that the only "assault" was on the sanctity of the truth.) for some kind of weird political motive which is really hard to support; it's very likely that they just wrong about what the result of the law would be.

I'm glad to read this article for the facts it presents, but am sort of upset with the quality of their science reporting.
posted by aubilenon at 12:09 PM on July 21, 2006

High average speeds aren't dangerous - high standard deviations are.

Can't find a citation, but iirc, in the first couple years that 55mph was rolled back, the NHTSA changed the definition of fatality from DOA, to death within 48hrs, to death within a week, to death within 10 days, etc.
posted by klarck at 12:17 PM on July 21, 2006

Um, that "argument" is a complete piece of shit. What an embarrassment the Wall Street Journal editorial page is. Nothing but shit, all shit, every damn day!

Let me start by saying I was happy about the raising of the speed limit. And I would support keeping it raised.

But that guy didn't prove a damn thing about the "safety" of driving above 55 mph.

Since 1995: cars have gotten much safer. Anti-lock breaks, 4 wheel disk breaks, air bags, crumple zones, etc. are much more common now that they were in 1995.

Since 1995, ten more years of "don't drive drunk" have had their way. Since 1995, mandatory seat-belt laws, and wear your seat-belt campaigns, have become much more common.

In short, even the most cursory of examinations of the past 10 years indicates there are several factors would could play a big role in making highways safer today.

This author make zero attempt to control for that. He just finds a statistic that might support his viewpoint, and blithely assumes the causation behind that correlation is the causation he wants. Blech.

What really annoys me is that the folks at the Journal aren't complete idiots. The author most likely knows he's being a disingenuous shit. He just hopes you're too stupid to notice. That's standard operating procedure for the Wall Street Journal's editorial page. It's a joke and embarrassment to an otherwise decent paper.
posted by teece at 12:20 PM on July 21, 2006

Yeah, because being able to make big car go fast is central to the definition of political liberty.
posted by slatternus at 12:23 PM on July 21, 2006

I, Credulous writes "I was wondering why the article didn't mention car safety technologies"

They did mention the wild safety benefits of power steering and brakes.
posted by Mitheral at 12:25 PM on July 21, 2006

What an embarrassment the Wall Street Journal editorial page is. Nothing but shit, all shit, every damn day!

Nailed it.
posted by nofundy at 12:34 PM on July 21, 2006

Actually in parts of the authobahn they are not imposing speed limits because of numerous accidents and deaths--only in parts though. No one notes that speeds over 60 sends gas consumption way up. Fortunately, gas is so cheap it doesn't bother us.
posted by Postroad at 12:34 PM on July 21, 2006

Fortunately, gas is so cheap it doesn't bother us.

Soon to be followed by a WSJ oped showing that "When you factor in inflation, you'll clearly see that we're paying much LESS for oil today"
posted by slatternus at 12:37 PM on July 21, 2006

Exactly what teece said.
posted by rxrfrx at 12:39 PM on July 21, 2006

"It is hard to see how increasing the speed limit could decrease the number of accidents per mile."

By reducing the time spent on long trips. People are less fatigued at hour nine than at hour ten and being tired can be just as bad as being drunk.

Western Infidels writes "The auto industry itself has a long track record of resisting regulations, including safety regulations of almost every sort. Modern cars are relatively clean and safe because government mandate requires them to be. The automobile industry is pretty much a perfect poster child for the value of government regulation. "

Though not blameless the auto industry is just giving people what they want, safety rarely sells. Increase the cost and your customers bail to the model that doesn't include seat belts, antilock brakes and plastic windshields. Chrysler and Ford offered seat belts in many models starting in 1956, the uptake was less than overwhelming. Lincoln offered anti lock in the 60s. Cost shelved the units. 25% of Americans still don't wear their seat belts.
posted by Mitheral at 12:44 PM on July 21, 2006

Heh. A space-limited op-ed in a newspaper falls short of a rigorous scientific study, and teece foams at the mouth.

But apparently nary a drop of ire for Ralph "history will never forgive Congress for this assault on the sanctity of human life" Nader or Joan "I kill babies" Claybrook.
posted by Kwantsar at 12:47 PM on July 21, 2006

*checks mouth for foam*

Nope, no foam. And please don't try to read my mind, Kwantsar, you suck at it. Didn't agree with Nader in 1994/5, don't now. Thought his argument was pretty shitty. Now, shittier than this argument in the Journal? I don't know. But seeing as how this post isn't really about Nader... I'm confused why I need to comment on him.

But you're obliquely right, probably by accident: straight up propaganda meant to mislead the casual reader, proffered in a paper with much respect, yes, that gets me very annoyed. Not afraid to say it. Curse words and heated opinion don't make it "foaming at the mouth."

But you knew that. You were playing a stupid rhetorical game that you thought was crafty. Oops, I guess I shouldn't read your mind, either. But it was stupid and not crafty.
posted by teece at 12:53 PM on July 21, 2006

The WSJ Opinion Journal is a wretched hive of scum and villainry. Would someone please shit on this thread and also scream in MetaTalk about the one-link-op-ed from a biased, crappy source being Bad For MetaFilter? That's the done thing, right?

Oh wait only asswipes do that; nevermind.
posted by fleacircus at 1:03 PM on July 21, 2006

It's quite easy to imagine how increasing the speed limit could decrease the number of crashes per mile. As the speed limit increases closer to the top speed people generally want to go, the relative speed between the slower traffic and the fastest gets smaller. That means less motivation to pass other cars, fewer surprises and aggravations, and thus fewer accidents. The benefits of this get smaller the faster you go, eventually being outweighed by the obvious dangers of going very fast and the growing proportion of people at the opposite end of the curve who prefer to drive significantly below the limit.

As everyone else said, automotive technology has improved somewhat over the past few decades. Cars perform better at higher speeds now than they did ten years ago, so naturally the speed at which people will on average want to drive them should also be higher. Therefore the optimal speed limit is higher now than it used to be. By how much, I have no idea.

The appropriate speed limit is, I've heard, often estimated by measuring the speeds people actually drive. The estimated optimal speed limit is derived from the median speed on the road. If it's higher than the posted limit by some amount, then the speed limit needs to be increased. If it's lower, it's probably safe to say the limit should be decreased. One would want to measure at the fastest section of the road under consideration, since the limit is meant to be a maximum. The same formula won't apply everywhere due to the variable presence of and respect for law enforcement.
posted by sfenders at 1:10 PM on July 21, 2006

I guess I'm a pretty big boy as well. 6'2" and 240lbs, and I have downed a 6 pack of beer (Pyramid Ales Hefeweizen) and a few shots of Captain Morgans and only blew 0.06.

This was on a camping trip, and a friend brought a Breathylizer with him. Yes, he was trained & qualified to use it.

But drinking & driving is kind of drifting off topic. Speed limits are so widely ignored anyway. Texas just raised theirs to 80mph on some highways. Once you get out of the Metro areas in California (and out of CA entirely) the speed limits are often 75mph.

It's people that really don't know how to drive that are the biggest problem.
posted by drstein at 1:12 PM on July 21, 2006

"straight up propaganda"? Good God.
posted by Kwantsar at 1:23 PM on July 21, 2006

re: the Autobahn, it's also obsessively well-maintained. not a good comparison

Agreed. And add the fact that there is actual lane discpline which is uniformly observed.
posted by Neiltupper at 1:28 PM on July 21, 2006

"Cars are better, but not that much"

Well, that all depends. If you're comparing cars from 1973, when the 55mph speed limit was enacted by congress, and now, you're dead wrong.

Late 60s/ early 70s automobiles are ridiculously dangerous compared to modern automobiles with

Antilock brake systems

electronic traction control

crumple zones and specially engineered bumpers designed
to absorb the kinetic energy of impact before it reaches the passenger compartment

air bags

shoulder harness seatbelts

impact triggered fuel cutoff systems

etc etc.
posted by stenseng at 1:30 PM on July 21, 2006

Cars engineered to be safe in a crash is the big factor(crumple, cabin integrity, knee bolsters etc.). And one your missing is tires. Tires are ridiculously better now than they were in the 60s.
posted by Mitheral at 1:33 PM on July 21, 2006 [1 favorite]

The faster you drive, the sooner you get to your destination. Less time on the road = less time for accidents. So, of course, accidents would decrease when the average traffic density decreases. How's that for logic for ya?
posted by blue_beetle at 1:43 PM on July 21, 2006

Cherry-picked data, the favorite method of the anti-government crowd. Here are the statistics (on p. 24), and you can see that there has been a pretty steady decline in fatalities per mile driven since 1989, well before the repeal. In fact, it appears that the decline slowed in 1995 to my eye, although I haven't done a statistical change point analysis on the data.

Original sources are so helpful in unmasking this type of deliberate deception.
posted by Mental Wimp at 1:48 PM on July 21, 2006 [2 favorites]

"straight up propaganda"? Good God

Kwantsar, if you can't see through that transparently stupid propaganda, you're going to want to head down to the community college and take a class or logic on rhetoric or something. Because the guy's not even good at it.

Good God, indeed.
posted by teece at 2:03 PM on July 21, 2006

teece is dead on, but others obviously are on the same page.

Discussion of this topic without considering increased car safety and reliability as the chief component is absurd.

Speed does kill. Consider if EVERYcar were electronically governed to 5mph on the freeways. How many fatalities would we have? A handful, because it is impossible to anticipate the creativity of some idiots, but certainly nowhere near the deaths we have now.

Now, consider the alternative. Consider every car were electronically governed to go NO LESS THAN 120mph on the freeway, with driver education, training, and ability remaining the same. Would fatalities go up?

Speed does kill. Cars have gotten safer. Some safety programs are being at least modestly successful.

In other words, it's complicated. But saying speed is not a factor in accidents is just asinine. As mentioned above, the Autobahn is safer due to many reasons that differ from America, and DESPITE the higher speeds, not because of.

Amen about the tires Mitheral. There truly is no comparison. The tires on your car today are probably superior to the finest racing tires available back then. This is even more true for motorcycle tires.
posted by Ynoxas at 2:05 PM on July 21, 2006

I recall reading about how graduated licensing had a pretty big impact on driving accident statistics. Getting teen males off the road until they smarten up makes a big difference.

Also I suspect the majority of accidents don't happen on highways. They happen at intersections where speed limits are not the biggest problem.

Tough drunk driving laws, better road technology, and such improving auto safety features as power steering and brakes are all proven life savers.

We are often told, by nanny-state advocates, that such public goods as safety require a loss of liberty. In the case of speed limits and traffic deaths, that just isn't so.

Did he just call himself an asshole and then punch himself in the face for calling himself an asshole?
posted by srboisvert at 2:06 PM on July 21, 2006

People, look at the actual NHTSA data the WSJ cherry-picked and I linked to above. You don't have to make any arguments about safer vehicles or changes in driving ages to rebut the story. Fatalities per mile have been dropping since well before the law was repealed. Look for yourself.
posted by Mental Wimp at 2:12 PM on July 21, 2006

But saying speed is not a factor in accidents is just asinine.

Did anyone actually say that? If so, I missed it. But anyway, thank you for once again demonstrating, perhaps more clearly than any of the other worthy efforts here, that one common side effect of driving a car is a strong feeling that you know everything there is to know about traffic engineering.
posted by sfenders at 2:54 PM on July 21, 2006

Touting the increase in relative car safety misses the point that the number of car crashes themselves has gone down. Mental Wimp did the legwork, and shows that the WSJ editorial is really worse than being merely disengenous.
posted by doozer_ex_machina at 3:00 PM on July 21, 2006

The speed limits are set for revenue. QED
posted by cellphone at 3:10 PM on July 21, 2006

sfenders: It's quite easy to imagine how increasing the speed limit could decrease the number of crashes per mile. As the speed limit increases closer to the top speed people generally want to go, the relative speed between the slower traffic and the fastest gets smaller.

Why on Earth would that happen?
posted by Western Infidels at 3:17 PM on July 21, 2006

The Facts are simple. although numbers of crashes per mile travelled have been going down for a decade. The increase in speed limits did not cause an Increase.

Is it seriously the implication of some of you people that cars have become so safe that the otherwise screaminly dangerous 70 is no longer as bad?

I find it likely that speed is not much of a factor in safety on highways.

on streets and roads, sure, but on limited access highways between urban areas? minor.
posted by Megafly at 4:24 PM on July 21, 2006

Why on Earth would that happen?

Well first off, of course it wouldn't happen no matter what the speed limit is. The effect will vary across the range of highway speeds in a way that I can't describe all that precisely. Also, things are complicated by the median speed people would drive in the absence of a speed limit being influenced by the speed limits they're used to. And I don't actually know the answer from any reliable source here, I'm just reasoning on my own; I have on good authority that it does happen, but have only my own opinion as to why. I think it's not that hard to figure out, though.

When the speed limit is "way too low": some people will obey it, some will ignore it entirely, and the difference will be very large. The latter group will be larger than it would if the limit were higher.

When it's "way too fast": most will drive much slower, but some people will try to exceed it anyway, and again the difference is very large. Those who want to go very fast will not be constrained by the speed limit. The latter group will be larger than it would if the limit were lower.

Somewhere in the middle, the distribution of speeds that people go will be closer together. That seems inevitable. Contributing to the effect are the facts that people are on average more willing to exceed the speed limit by a little than by a whole lot; but the less "reasonable" it's perceived as, the more it will be exceeded, while there are always some people who will obey it. And if the speed limit is just slightly faster than someone would otherwise drive, they'll often speed up to match it if they'd otherwise be holding up traffic. So if you set it in the middle, it'll pull the speed of those who are faster down, and speed up those who are slower. If you set it far to either side of the speed people actually want to go, it'll likely skew the distribution of actual speeds in some other, less desirable way.

It's all very approximate of course, because for example the mix of traffic also contributes to the complexity, as big trucks have a different response to speed limits than do passenger cars, etc. So that's why the way to go is to measure it experimentally rather than try to calculate some kind of equation.

But anyway, the 401 west of Toronto is a classic case of the speed limit being apparently set "too low" for most of the people who drive on it: I've frequently observed that the median speed is about 15-20% over the posted limit, and the occasional bit of traffic obeying the speed limit regularly fouls things up so you have fast-moving traffic coming up on much slower groups. Meanwhile there is the usual crowd who don't worry about the legal limits at all and do 160% of the supposed maximum. Raise the speed limit by 15% and it would speed up the great majority of those now obeying it, and have very little effect on most of those already going faster than that, since they are already near the limit of what is safe and economical. So there would be a decrease in relative speeds, most of the time.

Whether that would actually contribute to safety depends on whether it would be outweighed by the detrimental effects of the increased average speed; in that case I don't think it would, but every road is different. The collective instinctive reasoning of traffic is better able in some places than others to judge how fast it's safe to go.

Finally it may be worth considering that gasoline is getting increasingly expensive. If that trend continues, the argument that lower speed limits are beneficial to conserve fuel soon might not seem so irrelevant as it does today to the WSJ. IMO it's already of much greater significance than any argument to safety on the big high-speed motorways we're talking about.
posted by sfenders at 5:23 PM on July 21, 2006

I have a theory that more pedestrians are hit crossing at intersections than jaywalking. Anyone have anything to back me up?
posted by Alvy Ampersand at 7:11 PM on July 21, 2006

I've always theorized that the mph you are going is closely correlated with the amount of attention - expressed as a percentage - you are paying to the road. So if you're doing 80 mph, most of your attention is on the road. At 35 - 50, you're playing with the radio, talking on the cell phone, etc. At 100 mph, the road has your complete attention.
posted by Smedleyman at 7:57 PM on July 21, 2006

At 35 - 50, you're playing with the radio, talking on the cell phone, etc. At 100 mph, the road has your complete attention.

I have witnessed drivers nonchalantly contradicting your theory.

High speed may give an andrenaline rush if you're not used to it. If it's habitual, why would this work?

Do you remember learning to drive? Remember when 60 MPH on the freeway was exciting enough to sieze your complete attention? Does it still?
posted by Western Infidels at 9:06 PM on July 21, 2006 [1 favorite]

sfenders: Somewhere in the middle, the distribution of speeds that people go will be closer together. That seems inevitable. Contributing to the effect are the facts that people are on average more willing to exceed the speed limit by a little than by a whole lot; but the less "reasonable" it's perceived as, the more it will be exceeded, while there are always some people who will obey it.

I can't buy that. With a too-low speed limit, a large percentage of drivers will exceed the limit excessively, but will suddenly become law-abiding fuddy-duddies if the limit is raised slightly?

My experience driving in many of the United States has been that absolutely no one obeys the speed limit. Drivers typically go about 10 MPH over the limit, whether the limit is 25 MPH or 70 MPH, whether the road is smooth or full of holes, and mostly regardless of weather conditions. Actually driving the speed limit is an invitation to road-rage. Under such conditions, I don't see how increasing the speed limit can improve safety.
posted by Western Infidels at 9:20 PM on July 21, 2006

With a too-low speed limit, a large percentage of drivers will exceed the limit excessively, but will suddenly become law-abiding fuddy-duddies if the limit is raised slightly?

Not suddenly, gradually more of them as the speed is increased will not want to go any faster. We're talking statistics here, this only works over thousands and millions of people. If you increase the limit only slightly, it will be a correspondingly slight change in the average response to it. Increase it, and a few more people will obey the limit, and a few more will not even bother reaching it. When the effects of each are about equal, perhaps subtracting some constant margin to err on the side of caution, that's the best you can do. Or so I remember from some probably long-outdated book on traffic engineering that I read long ago, anyway. It seems to make sense. It's not like everyone would keep up to the speed limit no matter how you set it. The top speed each driver wants to go varies over quite a wide range.

absolutely no one obeys the speed limit.

In places where for whatever reason it's convention to go exactly 10mph faster than the speed limit, then you have an effective speed limit that's that much higher, and have to measure based on that. It makes a little anomalous bump in the distribution of speeds since some people will still insist on obeying the speed limit, but there's not much can be done about that without changing something other than the speed limit. The important thing to count is then how many people are going significantly more than the average X+10MPH, compared to how many are going slower than that. If their numbers are about equal, that's probably about the best you can do with that system; the speed limit is set "correctly" even if the system is otherwise rather broken.

Under such conditions, I don't see how increasing the speed limit can improve safety.

Then you must believe that decreasing the speed limit under those conditions that you find throughout the United States always increases safety. I find that very unlikely, but if it were so then why not cut them all in half just to be safe? Who can decide how much of this hypothetical public safety we're willing to tolerate? Should we perhaps take a public opinion poll to decide how fast the traffic should flow? That's exactly what is done by measuring how fast people actually go and setting the speed limit accordingly. In places where you have non-drivers with a direct interest in how fast the traffic is moving, such as residential areas, they might justifiably want to have some say in it, but out on the interstate, how else would you propose to pick a number for the maximum allowable speed?
posted by sfenders at 11:06 PM on July 21, 2006

The real story here is the intellectual dishonesty of the WSJ. My fave part:

"Americans have also arrived at their destinations sooner, worth an estimated $30 billion a year in time saved, according to the Cato study."

No real newspaper -- space-limited op-ed or not -- would allow the novel science of monetizing time-saved to appear in its pages without a thorough explanation of the (voodoo) economics involved.

Then there's the question of whether Americans are actually "arriving sooner" at all -- or whether Cato just ran some numbers based on the old speed limits and the new ones, and declared it a real-world "fact" without leaving the building.
posted by turducken at 11:55 PM on July 21, 2006

My experience driving in many of the United States has been that absolutely no one obeys the speed limit. Drivers typically go about 10 MPH over the limit, whether the limit is 25 MPH or 70 MPH

That hasn't been my experience at all. In places where the highway speed limit was 55, like the Jersey Turnpike, traffic moves at around 75. In places where the limit is 65, like most rural interstates, traffic moves at around 75. In places where the limit is 75, like the OK turnpikes or some of I-40 in AZ, traffic moves at between 75 and 80. Drive along an interstate with a 75 speed limit, and you'll find plenty of people doing 75 or a smidge more without people climbing into their trunks.
posted by ROU_Xenophobe at 1:21 AM on July 22, 2006

The appropriate speed limit is, I've heard, often estimated by measuring the speeds people actually drive. The estimated optimal speed limit is derived from the median speed on the road. If it's higher than the posted limit by some amount, then the speed limit needs to be increased. If it's lower, it's probably safe to say the limit should be decreased. One would want to measure at the fastest section of the road under consideration, since the limit is meant to be a maximum. The same formula won't apply everywhere due to the variable presence of and respect for law enforcement.

If this is true, then the optimal speed limit in Utah, even on fairly dense stretches of freeway, is about 85 mph.
posted by craniac at 6:31 AM on July 22, 2006

Alvy Ampersand writes "I have a theory that more pedestrians are hit crossing at intersections than jaywalking. Anyone have anything to back me up?"

Makes sense in the same way that more people probably get hit in NYC than Twin Falls, Idaho.
posted by Mitheral at 6:54 AM on July 22, 2006

Mitheral: [Higher speed limits could improve road safety] By reducing the time spent on long trips. People are less fatigued at hour nine than at hour ten and being tired can be just as bad as being drunk.

I think the historical trend has been to use faster transport systems to increase the distance and frequency of trips, not to reduce the time spent on them. That's where suburbs and sprawl come from.

Though not blameless the auto industry is just giving people what they want, safety rarely sells.

The auto industry has historically been one of the biggest purchasers of advertising in the country. They are responsible, to a large degree, for shaping what the people want.

One might also argue that the government that regulated the industry and forced it to develop safer cars was only enacting the will of the people, so obviously "the people don't want safe cars" is at best oversimplified to the point of being completely wrong.

And one might also respond: Who cares why the industry failed to produce safe cars on it's own? Shouldn't we be more interested in results than excuses?
posted by Western Infidels at 10:23 AM on July 22, 2006

sfenders: When the speed limit is "way too low": some people will obey it, some will ignore it entirely, and the difference will be very large. ... When it's "way too fast": most will drive much slower, but some people will try to exceed it anyway, and again the difference is very large. Those who want to go very fast will not be constrained by the speed limit. ... Somewhere in the middle, the distribution of speeds that people go will be closer together. That seems inevitable.

A wide distribution at each extreme doesn't even imply a narrow distribution somewhere in the middle, let alone make it "inevitable."

sfenders: Increase [the limit], and a few more people will obey the limit, and a few more will not even bother reaching it. ... The top speed each driver wants to go varies over quite a wide range.

So essentially: Many (and in my experience, virtually all) drivers have no regard for speed limits. Everybody has a different idea about the ideal speed. Therefore all drivers will go about the same speed if we change (in particular, raise) the speed limit.

Is this really adding up for you?

I don't know two licks about traffic engineering. But this just doesn't make sense. If it were a measured phenomenon (and maybe it is), it would be far more convincing. But I expect it would be quite a project to measure such a thing, and controlling external factors might be nearly impossible.
posted by Western Infidels at 10:48 AM on July 22, 2006

Okay, then. Enlighten me. I don't see how it's misleading, or how it's propoganda.

P1: Driving is safer than ever before.
P2: In 2005, injuries per mile was lower than ever. Deaths per mile was almost lower than ever.
P3: The Federal government (more or less) mandated a speed limit. This policy was originally justified by energy concerns, and kept in place (after the energy crisis ended) for safety.
P4: Congress killed the mandate. Some states raised speed limits.
P5: A bunch of people who were in favor of 55 said that a bunch of bad shit was going to happen if the Federal speed limit was removed.
P6: Despite more distractions, deaths and injuries did not rise. They fell.
P7: 29 of 31 states saw declines in death and injury rates since the repeal.
P8: There were other benefits. People could legally drive at speeds at which they felt comfortable. Cato said that $30B in lost time was eliminated.
P9: A lot of people still die. But nonetheless, the highway system is great for a lot of reasons. And there are a lot of reasons, speed limits notwithstanding, that highways are safer.
P10: Some argue that we need to trade liberty for safety. In the case the article discussed, this is not so.

Now, the WSJ attributed the two studies in P8. It's true that the Journal didn't describe the ideological bent of the organizations, but this is commonplace.

Is it the gratuitous Iraq remark in P9 that has everyone so pissed off?

Mental Wimp-- it looks to me as if you are attacking a straw man. Where do the authors claim that fatality rates weren't decreasing before 1995?
posted by Kwantsar at 11:07 AM on July 22, 2006

A wide distribution at each extreme doesn't even imply a narrow distribution somewhere in the middle, let alone make it "inevitable."

No, it makes inevitable that setting the speed limit somewhere in the middle is going to decrease relative speeds on average by more than setting it off either end, since it pulls both ends together no matter what the shape of the distribution. However, it actually is expected that the distribution of speeds will show a well-defined peak in the middle somewhere; in fact it usually is, as anyone with knowledge of statistics would expect, a normal distribution, although it "may be a unimodal or a bimodal curve depending upon the variation in the speed of different categories of vehicles moving on the highway" according to that study.
posted by sfenders at 11:46 AM on July 22, 2006

Regarding the complaints about the WSJ story, the important bit is this: Two studies, by the National Motorists Association and by the Cato Institute, have compared crash data in states that raised their speed limits with those that didn't and found no increase in deaths in the higher speed states.

If you're willing to believe those results, their conclusion is reasonable.
posted by sfenders at 12:33 PM on July 22, 2006

Western Infidels writes "I think the historical trend has been to use faster transport systems to increase the distance and frequency of trips, not to reduce the time spent on them. That's where suburbs and sprawl come from."

True but in the sparsly populated states (think Montana) a significant amount of time can be saved traveling between citys.

Western Infidels writes "One might also argue that the government that regulated the industry and forced it to develop safer cars was only enacting the will of the people, so obviously 'the people don't want safe cars' is at best oversimplified to the point of being completely wrong."

Maybe, goverment actions aren't just the sum of the motivations of their citizens. They are supposed to act more rationally and with less individual self interest (despite the wide acceptance of pork projects).

Here's a few different examples: You can't buy a cheap mid engine car. The cheap two seater sports car market is Miata. For a while there after GM killed the F-Body your choice of pony car was Mustang or, uh, Mustang. You can't buy a 4X4 Van (the Sprinter 4x4 is Euope only). You can't buy a small pickup like the rampage or even D50 in Canada. You can't buy the Smart in the US (the last is regulation). Why is there such limited choice in these markets? Because the car companies can't sell enough to justify production. Same thing for many safety features in the 50s, 60s, 70s.

Western Infidels writes "And one might also respond: Who cares why the industry failed to produce safe cars on it's own? Shouldn't we be more interested in results than excuses?"

Sure, and at the same time shouldn't we be more interested in how to make things better rather than assigning blame?
posted by Mitheral at 3:06 PM on July 22, 2006

Turducken write:

"Americans have also arrived at their destinations sooner, worth an estimated $30 billion a year in time saved, according to the Cato study."

No real newspaper -- space-limited op-ed or not -- would allow the novel science of monetizing time-saved to appear in its pages without a thorough explanation of the (voodoo) economics involved.

I disagree with your assertion about what "real" newspapers do, but I attributed it to you...hey, kinda funny, just like the WSJ did when it attributed the statistics to the Cato study...
posted by bugmuncher at 7:28 PM on July 22, 2006

Some points I learned in my defensive driving classes way back when I was 16:

Speed doesn't kill, at least as speeds currently attainable by human transportation. (I went into an airplane and the speed of it didn't kill me!) Rapid deceleration (or acceleration) due to impact is what kills.

Speed limits frequently create artificial or unnatural speed differential by encouraging people drive more slowly in order to comply with the law.

Faster cars approaching slower cars in the same lane are more likely to want to merge into a different lane in order to maintain their higher speed. Merges are inherently more dangerous than driving in a straight line because of blind spots and speed differential.

The safest driving speed is not the fastest or slowest on the road; it's roughly the same speed as everyone else is going. Too much higher and too much lower and you're creating high speed differential, and could collide with something.
posted by bugmuncher at 7:48 PM on July 22, 2006

bugmuncher wrote:
Turducken write:
Wow, bug. You really are a journalist, leaving your typos for your editors to fix... :-)
posted by bugmuncher at 7:53 PM on July 22, 2006

bugmuncher: I don't understand your comment -- so I'll take the opportunity to re-state mine: My assertion is obviously an opinion, while the WSJ presents Cato's assertion as fact. But "real" newspapers are supposed to check facts, even if the source is of the same political bent as the Op-Ed page. Ergo, the problem: intellectual dishonesty. A lack of fact checking -- indeed, an eager acceptance of opinion/assertion/belief as gospel truth -- is rampant on the right wing of the poltical spectrum in the U.S. these days, and it's very, very bad for the country as a whole.
posted by turducken at 7:30 PM on July 23, 2006

davy, people have been living in that region for thousands of years. You can check it out in this interesting book called The Bible. I think you can get copies for free somewhere.

Nice to see free markets properly valuing reference books.
posted by srboisvert at 12:12 PM on July 24, 2006