Should speed limits be higher?

Arthur van Benthem, who is on the job market from Stanford, says no:

When choosing his speed, a driver faces a trade-off between private benefits (time savings) and private costs (fuel cost and own damage and injury). Driving faster also has external costs (pollution, adverse health impacts and injury to other drivers). This paper uses large-scale speed limit increases in the western United States in 1987 and 1996 to address three related questions. First, do the social benefits of raising speed limits exceed the social (private plus external) costs? Second, do the private benefits of driving faster as a result of higher speed limits exceed the private costs? Third, could completely eliminating speed limits improve efficiency? I find that a 10 mph speed limit increase on highways leads to a 3-4 mph increase in travel speed, 9-15% more accidents, 34-60% more fatal accidents, and elevated pollutant concentrations of 14-25% (carbon monoxide), 9-16% (nitrogen oxides), 1-11% (ozone) and 9% higher fetal death rates around the affected freeways. I use these estimates to calculate private and external benefits and costs, and find that the social costs of speed limit increases are three to ten times larger than the social benefits. In contrast, many individual drivers would enjoy a net private benefit from driving faster. Privately, a value of a statistical life (VSL) of $6.0 million or less justifies driving faster, but the social planner’s VSL would have to be below $0.9 million to justify higher speed limits. The substantial difference between private and social optimal speed choices provides a strong rationale for having speed limits. Although speed limits are blunt instruments that differ from an ideal Pigovian tax on speed, it is highly unlikely that any hidden administrative costs or unforeseen behavioral adjustments could make eliminating speed limits an efficiency-improving proposition.


"but the social planner’s VSL would have to be below $0.9 million to justify higher speed limits": since my own tendency is to value life at no more than a million (dollars, euros or pounds - since we don't want implications of false accuracy here) that would mean the case is teetering on the brink.

What, I sometimes wonder, should we do about lives of negative value? If higher speed limits led to higher death rates of the sort of people who drive drunk, drugged up, or while texting, shouldn't we bung 'em up sharpish? But then perhaps such people are prone to ignore speed limits. Still, it does raise the thought that it might be worth pursuing further the question of who it is that dies.

Its a good point but an awfully dangerous question...

Drunks seem to be more likely to survive the crashes they cause, so allowing higher speeds means a drunk will be less likely to be stopped for speeding, and when he hits the family in the SUV, he will kill all five in the family while being limber and relaxed enough to survive being thrown through the windshield into the bushes with minor scratches and bruises

Isn't that fact an urban legend?

From the abstract, it looks like their thumb was on the scale.

trade-off between private benefits (time savings) and private costs (fuel cost and own damage and injury). Driving faster also has external costs (pollution, adverse health impacts and injury to other drivers).

What of external benefits? Lots of people can stack up behind cars going the speed limit...

Unless you live somewhere where congestion is not an issue, this is unlikely to be the case. Increasing speed reduces the capacity of freeways because of the increased space between cars needed to prevent collisions. See this article on "rolling speed harmonization" for why "forcing" drivers into the same, slower speed will likely result in faster travel times (as well as hordes of commenters who don't want to believe that)., that's actually not the proble (and it seems to me it's not written in the article anyway). Higher speed = higher capacity; unless there is some bottleneck - and there it does matter, once you get the cars to stop in the bottleneck, the capacity nosedives; so, it is better if the cars don't overwhelm the bottleneck in the first place - and you achive this by reducing highway capacity (with slower speed) before the bottleneck.....

Anyone who drove on German highwaya knows that high speed means less congestion!

German highways are highly congested. The average speed in some parts of NRW is among the lowest in Europe. If you want to learn German by listening to the radio in the car, the word "Stau" is one you´ll here pretty often.

Having said that, I blame the Stau in German Autobahn on the fact that there are no tolls for cars. French autoroutes are a pleasure to drive through, but you can pay about 7 euros per 100km.

It's very condescending to start a comment with "um....", especially when you don't seem to understand the issue. It's true that higher speed increases capacity if there is no congestion, and so on rural roads, you're absolutely right. (That said, it's unclear how exactly Jody's scenario of cars "stacking up" behind a car going the speed limit would take place if freeways were uncongested.)

But this it's just not right to assume that "bottlenecks" are the problem here. The problem is as you increase speed, you must increase the distance between cars. There is in fact a whole profession devoted to studying this, and so I will register my standard complaint that the world is not just a fact-free theorizing zone. We really do have information about this. To take just one example:

As congestion increases and traffic begins to move slower than 70% of the posted
speed limit, overall productivity declines and the highway supports fewer vehicles, as
highways are engineered to move specific volumes of vehicles. In addition, highways do
not operate at their maximum efficiency when moving at 60 MPH (the most common
highway speed limit in Washington State) because of the need for increased spacing
between vehicles.
Maximum throughput (optimal flow) speed is not a static number for all
highways, and can vary from facility to facility and from segment to segment depending
on conditions such as lane width, slope, shoulder width, pavement conditions, traffic
composition, and presence or lack of a median barrier for example. It should be noted
that, as cars are equipped with more sophisticated devices and become easier to
maneuver, maximum throughput speed should increase. Currently, maximum throughput
speed on a typical freeway segment in the Central Puget Sound region is about 51 mph

Well, you are saying that on some highways the safe distance doesn't grow linearly... depends on the highway... btw: have you ever driven in germany? Cars ~ 2 secs from each other, going 80mph on a highway with road works, so that one lane was 2m wide..... Yes, the bottleneck are the problem; and behaviour of the drivers. German police is very tough on measuring safe distance.

The problem is as you increase speed, you must increase the distance between cars.

That's only true in their model, not reality.

The article is talking about an unusual situation with highly differentiated speeds. If the optimal flow speed is actually 51 mph, I'm not sure a freeway with a limit of 65 mph is going to have much trouble hitting that flow speed. I've driven in plenty of traffic at 10-15 mph under the limit due to substantial congestion. I'm sure it would be a lot different on a mountain road where trucks have to drive 30 under to be safe.

Remember driver's ed? Remember being told to give yourself a 2 second distance behind the guy in front of you?

Do the algebra. With a 2 second distance, the number of cars a traffic lane will carry stays exactly the same no matter what the speed, from 10MPH to 90MPH.

Your comment is a mess. Yes, higher speed reduces capacity, but it decreases travel time. If capacity is exceeded, speeds must drop. You never have a highway at the speed limit during rush hour, so raising the speed limit would have no effect.

No arguments about rush hour limits. What's annoying is when the same (65 mph) limit applies on a perfect weather summer night on a nearly empty, flat, straight run of interstate. Wouldn't higher speed limits reduce travel time then because capacity is no longer constraining?

Bingo. The problem with the traffic engineering work that leads to conclusions such as the speed limit should be 51 MPH is that the model is nonsense, and the conclusion doesn't follow from the model.

1) Why is the model nonsense?

The model is nonsense because there is a certain amount of goods/people to transport. Some of those goods are time controlled (e.g., due to working hours) and some are not. The benefit of higher speed limits is being able to transport more goods during the non-congested period and clear that load off of the road. In this regime, the spacing between the cars is irrelevant. Its there time on the road that matters. Reducing the time on the road depends on increasing the speed.

2) The conclusions don't follow.

Just because the optimal congestion limited speed is say 51MPH on a road does not mean that the speed limit should be 51 MPH. The conclusion that we should limit the speed of the drivers in this way mistakes cause and effect.

In my experience, as load increases, the spacing between cars decreases, and as the spacing decreases, drivers slow down. In other words, if the speed limit is 80 MPH, as the busy period builds up, the average speed declines first to accomodate the load. Eventually the speed drops so low that the carrying capacity begins to decline, and we get a congestion collapse: grid lock.

In short: road capacity is self stabilizing when cars are permitted to drive ABOVE the optimal capacity speed. Its extremely important then that the cars maintain an average speed above the optimal capacity speed. Once the speed drops below the optimal capacity speed at full load, the system collapses.

As an aside, 'traffic engineering' science is one of the most blatantly politicized technical disciplines. Whenever someone starts preaching about respecting the technical analysis, I cringe.

Life in Canada has taught me that adding a lane is many times better than increasing the speed limit, if the goal is to speed up travel, reduce congestion and accidents, etc.

But these kinds of studies are really flawed. Example: If travel speed didn't increase much, then how can we attribute the increase in accidents to the increase in speed limits? Put differently, did the accidents occur because some people were speeding, or because other people refused to speed up?

These sorts of studies really are garbage.

To be fair, one can point out that the claim, that raising the speed causes more accidents, is neutral as to deeper cause (not enough people raising their speeds, as you point out).

Not everyone can raise their speed, unless such things as turnoffs are redesigned which requires taking more land are redoing cloverleaf interchanges to allow added lanes for slowdown, longer bridges, larger loops, or else people will need to brake hard, slow down well below the speed limit before the ramps, etc. And trucks can't brake hard, take the loops at high speed, and can't accelerate very fast.

If the exits are tens of miles apart, then high speed is possible for all vehicles, but if the exits are only miles apart, then the traffic speed will be a mix of very different speed vehicles.

It also seems hard to believe that an average increase of 3-4 mph resulted in "elevated pollutant concentrations of 14-25% (carbon monoxide), 9-16% (nitrogen oxides), 1-11% (ozone)". Cars pollution isn't linear, but a rise in CO of 14-25% for an average increase of 3.5 mph seems to stretch credibility. I'm curious at how they arrived at the extra pollution figures.

Ok, I read the relevant section of the paper and I hesitate to use the word junk science, but it's clearly a flawed methodology. Essentially, he uses California Air Quality stations and measures the differences over time.

This part is striking:
"To estimate the impact of the 55 to 65 mph increase on rural interstates in 1987, I de fine treatment stations as being located at most x miles away from the 10 mph change. Control stations are located at least y miles away from the 10 mph change, where y >= x"

That is NOT a decent control group. A decent control group should have been another highway where the speed limit did NOT get raised. This is a pretty significant flaw.

Also, he didn't even attempt to control for higher traffic on the highway. It's pretty logical that there would have been some growth over the period just due to population change over a 6 year time frame. Furthermore, if the speed limit was raised it's quite possible the freeway had a higher than expected growth as traffic diverted towards the higher speed artery.

He makes no attempt to factor in traffic growth over the time period at all. He just apparently assumes that there was 0 traffic growth and attributes all pollution increases to the increase in speed limits, though he does try to factor out influences any large new manufacturing facilities.

Up to 60% more fatal accidents from only a 10 mph increase? Really? Vehicles are much safer now than when the limits were raised from 55, yet even then we didn't see such a huge increase in fatalities. Multiple airbags, crush zones, sensors to help avoid loss of control, better tires, brakes, lighting etc. are evidence for anyone without atrophied common sense that the study is work of activism not science.

1/2 X m X V^2

Less time to react increasing the number of accidents combined with faster speeds increasing the energy transferred to the body (by the sqaure of the speed) in such an accident?

I can't vouch for its accuracy, but it doesn't sound unreasonable.

Actual speed increased only 3 mph, so it does seem quite unreasonable to me.

_Average_ only increased 3 mph from a 10 mph limit increase. The problem is most of the time you are either traveling to or from the highway when you must travel on slow roads, or you are on highways with traffic traveling different speeds, for example trucks climbing hills or entering and slowly gaining speed of slowing for the exit or tight turn in the road. Cars will speed up. then be forced to slow down for the truck, then speed up, then slow down unless the highway is light and you can travel at the higher speed limit in the left lane with few trucks passing another truck going up hill.

AVERAGE speed increased by 3mph!

Natural experiments are for when you can't do randomized controlled experiments. The federal government has done them. If people with power over speed limits wanted to do a good job choosing them, they would do more.

An ethics panel might be a little concerned about conducting RCTs where people die at the margin...

Well, it would be interesting, right? But if you're honestly thinking about increasing the limit to 75 mph, is it really unethical to run some experiments where that's exactly what you do?

Suggest you change your title to "Study proves Germans are all dead".

This is a case of why we have a representative democracy, and why it works.

It would be hard for a representative to lower the maximum speed limit by 10 miles and get re-elected.

Sometimes representative government is a brake, and not a gas pedal.

This is pure political activism, and entirely unscientific. It's so bad it's embarrassing.

On the other hand, speed limits under 80 or even 90 in the western united states, on vast lonely straight highways is simply a misallocation of people's time. For all practical purposes, there should be no speed limits on western and midwestern rural highways.

This is essentially the case in Montana, at least the last time I drove through it about a decade ago. I went several hundred miles at 90 mph without seeing a speed limit sign or a state patrol officer.

More real world data that rebuts the research results.

Utah: Increasing Speed Limits Doesn't Kill
Test Confirms 80 MPH Okay

"The Utah Department of Transportation (UDOT)," began an article in, "announced last week that the experimental increase in the state's maximum speed limit to 80 MPH has been a success in terms of safety.

UDOT Deputy Director Carlos Braceras testified before the state Interim Committee on Transportation that that there has been no increase in accidents as a result of the higher number printed on the speed limit signs on certain stretches of Interstate 15."

When Montana had no daytime speed limit, fatalities not only went down but Montana recorded the state's fewest road fatalities during that period.

People want to get where they're going quickly and alive. If the powers that be would set limits more in accordance with that fact, perhaps the national blood pressure – and that of drivers – would flow more efficiently and just as safely. At least, it wouldn't hurt to try it out here and there.

It gets back to that self-organizing thing we've been perfecting for thousands of years. As a herd, we will find ways on our own to live together, mostly safely. Even in the fast lane.

Okay, but the problem here is that Utah is raising the limits on the roads it believes it would be safe to do so. They are not just increasing the limit on all roads by 10 mph, which is basically what this study is looking at. Of course, that doesn't say much for this study since nobody would propose a blanket increase. That's just silly.

Well yeah, raising speed limit on roads that aren't safe would just be stupid.

Even if the research is accurate, it appears to be done at the aggregate level, as to nationwide changes and observed consequences. It does not provide guidance re localized changes and consequences. Population density varies widely within his sample. The conclusion could be true at a nationwide level but it would not prove anything about the localized level. If anything, it would prove that the decision ought not be made at the nationwide level. .

I also think the driverless car is a technological solution that, once perfected, would maximize private welfare and minimize externalities. When cars can drive inches behind one another safely, because of sensors and controls that have no human interaction, and waiting time at intersections can be minimized because the vehicles communicate with each other over a local network at the intersection to pass through as safely and speedily as possible, the need for speed limit debates, policemen with radars and breathalyzers, traffic lights, etc will disappear.and instead the labor and capital dollars dedicated to those factors will be repurposed to maintain and improve the vehicle control network. We will debate whether humans should have a fundamental right to take control of their vehicles from the network. Some states will conduct natural experiments along that line and economists will study those.results instead.

If anything, it would prove that the decision ought not be made at the nationwide level.

Agreed. Personally it drives me batty whenever someone frames the issue as the speed limit, as if one size fits all.

There is also the issue of local knowledge, and setting the speed limit on given stretches of highway according to what 85% or so of drivers are doing. For example, if the speed limit on a given stretch of road is 65 mph, and most people are driving 75, then it should be raised to 75. Then sections of the highway where 65 or lower is more appropriate because of turns etc., then it can in fact be 65 (or lower), and this will signal to drivers that the lower speed limit actually means something safety related, and this could make the highway safer.

Local issues should be handled locally, and one size fits all policies like debating or studying the speed limit are ignorant of the fact that aggregates lie.

By the way, I believe highways, especially two lane highways (each direction, four total), would be safer (and more efficient time-wise) if there was something like "normal speed limit 65 mph, passing-only speed limit 75 mph".

There are far too many "rolling roadblocks" on two lane highways, where for miles and miles there are no cars, and then for a small stretch there are about 20 or 30 cars driving dangerously close to each other.

My basic hypothesis for this is because people are paranoid about breaking the speed limit to pass each other because of fears of getting ticketed, while at the same time other cars would be very happy to pass but are unable to, so they drive close to the car in front of them because they want to pass the rolling roadblock as fast as possible, and if they don't drive close to each other, someone else with the same mindset will indeed cut in front of them, making the wait that much longer.

The filters did not like the link. Executive summary of article that shows the real world examples contradict the study results.

Speed Doesn’t Kill
The Repeal of the 55-MPH Speed Limit

In 1995 the Republican Congress repealed the 55-mile-per-hour federal speed limit law.

At the time, the highway safety lobby and consumer advocacy groups made apocalyptic predictions about 6,400 increased deaths and a million additional injuries if posted speed limits were raised. Ralph Nader even said that “history will never forgive Congress for this assault on the sanctity of human life.”

In fact, in 1997 there were 66,000 fewer road injuries than in 1995, the year before the speed limits were raised. The injury rate per 100 million vehicle miles traveled fell to its lowest level ever recorded in 1997.

I know web commenters all think they can demolish a paper they haven't read on the spot, but economists are not in fact stupid, they do in fact know that time trends matter, and they do in fact work hard in their methodology to distinguish between those effects and actual policy effects.

Do you seriously think that there aren't time trend controls in the paper? (There are.)

Increasing the speed limit by 10mph has vastly different effects depending on what the old limit was. I see no mention of what margin was considered. When the speed limit goes from 55 to 65, I generally increase my speed from 64 to 74. But when it goes from 65 to 75, I only go up to 81. And need anyone point out that there's a bigger practical difference 25->35 than 65->75?

All I know is that any time the government tries to do anything, we are all worse off. Except when the government spends money improving the particular roads I drive on.

Well, a trillion is needed to get them back to the condition in 1980, and probably two trillion to redesign them for higher speeds. In the East, many 4 lane roads are still designed for 50 mph (though few travel that slow).

Wow. What an absolutely awe inspiring triumph of modern economic analysis. I appreciate the technical proficiency of the author. Quibbling aside he actually does quite an good job of capturing the costs and even the benefits.

The paper though starkly presents the problem described by David Graeber in "Debt: The First 5,000 Years" of the rationalist attempts to plea a monetary value on life in order to provide a "scientific" patina to what is clearly an ideological program.

Put amply, Mr. Benthm decides that allowing 479.850 people to save an hour off their commute is worth 1 person dying each year (Value of Life = $7,375,305 - Value of an Hour of time = $15.37).

I am not suggesting that such a trade off does not need to be made. It does and it is made on a daily basis. What I object to is the perpetuation of the belief that by doing this analysis that you are somehow engaging in "rational" analysis as opposed to others who utilize less "clear cut" rules. As noted by many of the comments, this analysis is somewhat limited. It should just be considered a "start" rather than an end. By constraining yourself to the mathematical formula you are in fact imposing your ideology on the debate under the guise of "rationality."

Check out the guy's name George: van Benthem, he's merely promoting the greatest happiness for the greatest number!

A wonderful book dealing with the issues of speed limits and everything else related to traffic:

Traffic: Why We Drive the Way We Do (and What It Says About Us)
Tom Vanderbilt

Great book. It convinced me of the sheer lunacy of something I'd always taken for granted - allowing cars to park on the side of the road. It's quite dangerous, as they block the view around corners and can hide pedestrians about to cross the street. It's a stunning waste of resources, as we spend all this money to pave roads, which are intended to be used to move as many cars as possible in as quick a time as possible, then set aside one lane for cars that aren't moving anywhere at all. And it's ugly.

"Since the Ohio Turnpike increased the speed limit to 70 mph six months ago, commercial truck traffic and accidents have also increased slightly, according to data from turnpike officials and the State Highway Patrol. About 1,270 crashes were recorded on the turnpike from April through September, including two that were fatal, according to the patrol, up from 1,159 in the same period in 2010. "

Looks like a 10% increase for Ohio.

What we really need is a lot more congestion pricing. Traffic is a nice example of what happens to usage of public goods at zero cost.

Congestion pricing sounds good in theory but implementing it fairly and efficiently is a near impossibility. I think the problem is cultural - Americans, by and large, believe public transportation is for losers. If they stop believing that, a lot of urban congestion will take care of itself.

The report of an increase in accidents -- especially the huge increase in fatalities -- makes me question the methods and honesty of the people who wrote that paper.

The accident rate has, in fact, fallen steadily since the '70s, especially since the repeal of the national 55 limit.

The paper seems to commit several fallacies, the most obvious of which is that passing a speed limit law does not magically make everyone slow down. Instead, what happened during the 55 period (1974-88) is that maybe 5% of the people complied; and about half of those decided to perform the public disservice of blocking the fast lanes of freeways (a behavior that the police, in saner times, discourage, but encouraged then because it was the only way for them to get anything close to compliance). Result: normal drivers took extraordinary chances to get past these idiots. That's a situation no sane person wants to go back to, any more than alcohol prohibition.

Besides, speeding per se is a victimless crime. If I drive fast and cause a wreck, then the state has business prosecuting me for reckless driving. But until that happens, they should leave me alone. The whole point of freedom is that they have no business restricting it as a precaution against speculative harms that have not yet happened.

Even if your speeding does not cause an accident, it greatly increases the damage resulting from any accident (mv2).

For example in this linked study pedestrians struck by a 25 mph car have about a 4% chance of dying, while pedestrians struck by a 40 mph car have about a 22% chance of dying. So while an accident may be "caused" by a jaywalking pedestrian, a speeding motorist increases the chance of fatal consequences by 5X or more.

The obvious conclusion is that law enforcement should not have to wait until you cause an accident by speeding through a crowded city, but can enforce the law simply on the basis of the massive increase in the probability of death and injury resulting from your decision to violate the law.

We are talking about freeway speed limits; ergo vehicle-pedestrian fatalities are meaningless.

Secondly, although there is this substantial mortality difference you quote between 25 mph and 40 mph; is there a meaningful difference between, say, 75 mph and 85 mph?

Correct, most of the areas where speedlimits are to low are on roads other than freeways. The exception is freeways away from congested areas.

Most metro areas would see benefit during uncongested times, but could see more problem during congested times.

"The accident rate has, in fact, fallen steadily since the ’70s,..."

Yep, thanks to lots of government dictates and mandated extra spending on useless safety features like passive passenger restraints AND air bags AND auto bodies that collapse to absorb impact energy, and anti-lock brakes, and all sorts of other big government interference in the free market's ability to give consumers the vehicles they want.

Not to mention the taking of your liberty to drive without a seat belt with worn out brakes.

Mulp, all reasons that fatalities have gone down. None of those factors explain why the number of accidents has gone down.

Speed-limits cause congestion by causing the rush period to start earlier because less of the traffic load can clear before the rush begins. That congestion leads to accidents. Thankfully, speed limits have risen, and we've enjoyed fewer fender benders as result.

This is from an email from Aaron:

"I haven't read the paper you linked, but I'm highly skeptical.

The evidence is pretty solid that speed limits are too low and that raising them will actually improve safety.

Second, most vehicles improve in fuel efficiency up to about 65mph. What matters is the amount of distance that can be traveled before having to brake.

Electronic controlled throttle and fuel injection mean that smooth, quick acceleration combined with point one mean less fuel consumption and pollution since little fuel in needed to maintain speed and it burns efficiently.

Here's my comment that didn't post:

Yes. And we should also stop calling them limits and call them the "Target Speed". A better question is why don't we have electronics signs that adjust for conditions and coordinate with signals. It'd be a great bridge to the driverless future and save fuel and time."

I am not sure why people post incorrect and unattributed assertions such as "most vehicles improve in fuel efficiency up to about 65mph".
Actual measured data shows maximum fuel efficiency is delivered at speeds ranging from 45 to 55 mph (see link below).

Since air resistance increases as the square of velocity, it is unlikely that any vehicles get maximum fuel efficiency at 65 mph (although mpg depends on many variables, including engine design, gearing, body shape, tires, etc.,etc. so there may be some combination that results in a 65 mph maximum, but it certainly is not the norm).

Yes, but bear in mind that that's 1997 data. If you look at the trends then cars have steadily been getting more efficient at higher speeds. e.g. 1973 vehicles already hit their most efficient speed as low as 30 mph.

In addition better engineering has allowed cars to run efficiently at a wider range of speeds. I wouldn't be surprised to see cars now having efficient speed ranges creep up to the 65 mph range.

EPA data shows fleet fuel economy doesn't decline until about 65mph for each model year for over a decade.

However, my claim of less pollution is suspect, the paper found engineering data showed that pollution increases despite fuel economy improving with speed.

Before we go mucking about with speed limits, I'd like to see an analysis of the optimal licensing regime.

If you choose to look at this from an economics/physics perspective, a car that’s driving faster will have to endure higher wind resistance (drag) than a car that’s driving slower, therefore cars driving faster require a greater use of energy than cars which are driving slower making it more efficient to drive slower. According to, you lose 2 % of your fuel efficiency for each mph you drive over 55mph. Even if a higher speed limit were to be imposed, driving speeds are typically variable and depend upon the individual driving the car. If everyone drives at 65mph, there is a minimal risk of collision. Realistically though, you have law-abiding motorists who drive at or below the speed limit combined with those who prefer to drive faster, which increases the risk of collision. The question here is whether or not an increase in speed limits would cause an increase in variability. I feel that if the speed limit were increased, conservative motorists would continue to drive at lower speeds whereas the increase would cause those who already speed to drive even faster, justifying themselves with the mentality that it’s ok to drive 5-10 miles above the already increased speed limit. In this scenario variability would increase tremendously leading to a much higher rate of collision. I guess ultimately it depends on the individual and how they assess the risk of driving faster and the decrease in fuel efficiency with the value of arriving at their destination on time.

Research shows that increasing speed limits doesn't affect top speeds, but increases the lower end, improving safety.

However, I think we really need to get rid of the idea of a "Speed Limit" and start using "Target Speed".

I think a "Target Speed" would appeal to both slow and fast drivers (who probably aren't as concerned about safety) as it seems clearer that efficiency is a large factor.

Having electronic signs that vary with condition could also help prevent backups.

I agree with you, I think a target speed would be a good way to reduce variability in driving speeds.

Do we have any reason to believe people will pay heed to a toothless "target" when they routinely ignore enforceable limits?

Yes, fines for outside of target, say, 8% and high fines for 15% above target (this is where safety becomes a problem).

Actually, that isn't quite right. Optimal Speed Limit for safety is the 85 percentile of what people would be comfortable doing w/o a speedlimit. The Target Speed would be lower and enforcement range tighter than I first assumed. Fines on both over and under, fines being higher for over, and increasing proportionaly with difference from target speed.

So how is this different from current limits?

I think that just calling it a "target speed" will cause more people to aim for that particular speed rather than referring to it as a "limit". The word "limit" has the connotation of the speed being something that you can't surpass but the word "target" hints that it's a speed you should aim to reach. Then again, I do see your point in the fact that this whole "target speed" idea isn't really all that different from current speed limit enforcement.

B-K-J has it exactly. I was making the point that needn't be "toothless". And safety and efficiency would benefit greatly by addressing minimums rather than just maximums.

I actually think that lowering the speed limit would provoke other drivers to drive faster, especially the younger generation who feel the need to experience a higher sense of rebellion.

Someone wiser than I said, "Speed doesn't kill; closing speed kills."

That is a remarkably well written paragraph.

Comments for this post are closed