Speed limits as an emergent phenomenon

by Russ Roberts on February 26, 2007

in Complexity & Emergence

After listening to this podcast with Don Boudreaux on Hayeks’ distinction between law and legislation, John Hoehn responded with this fascinating example of how Michigan lets speed limits emerge on certain roads:

Hayekians may be interested to know that, in Michigan, traffic speed limits on state and county roads are "discovered".   Speed limits on the latter roads may be set initially based on engineering considerations and the general experience of county road commission personnel.  However, if a member of the general public thinks the initial limit is too fast or too slow, he or she can request a "traffic study" by the county road commission.  If the road commission hasn’t recently conducted such a study, it usually responds positively to the request–it sets up equipment on the section of road to measure the actual speeds of vehicles traveling on the road.

The purpose of the traffic study is to estimate the speed at which the 85th percentile of traffic speeds.  If the 85th percentile matches the posted speed limit, the posted limit is affirmed.  If the 85th percentile is different from the limit, a new speed limit is established that is as close to the 85th percentile as possible (though rounded to the nearest multiple of 5, I believe).  That’s how speed limits are discovered in Michigan.  The only caveat here is that the discovery approach only applies to speed limits under the statutory maximum of 55 mph to state and county roads.  The state has a brochure that details of the discovery approach:


Of course, there are other types of highways for which speed limits are set by legislative bodies.  Cities have the discretion to set statutory limits within their boundaries.  State legislation sets residential and school zone limits to 35 mph.  Freeway limits are also set by statute to 65 or 70 in most places.  It’s all covered in the brochure on the state website.

Be Sociable, Share!



12 comments    Share Share    Print    Email


Brandon F February 26, 2007 at 11:39 am

My father works for the Montana Department of Transportation, and while I can't say for sure if speed limits are "discovered" in the same manner as they are in Michigan, he has informed me of that statistic as well. It probably doesn't matter as much here though, as the posted speed limit is 70 mph on state highways and 75 on the interstate, and most police officers don't start pulling people over until they are doing about 80-85 mph.

It used to be that the speed limits on highways and interstates in Montana were all "discovered", because there was no posted speed limit. The law was whatever was "reasonable and prudent" for the conditions, which could mean up to 85 in some places. If the feds hadn't threatened to cut off highway funding, I suspect the market would still be setting speed limits in Montana.

Steve February 26, 2007 at 12:44 pm

There's one major problem I see with this methodology, the 85th percentile is "discovered" while the (potentially) ridiculously low posted speed limit is in place. I'm sure the effect of the posted limit varies, but it surely has some effect. On lightly traveled, highly patrolled roads, I suspect this effect is dramatic.

I got a speeding ticket on a road posted at 25 mph (limit set by city). I had read an article in the local paper about how the state highway a quarter mile away was going to have a traffic study conducted, and that the limit would be reset to the 85th percentile. I wanted to use that methodology to prove that the limit was too low on the road I was traveling when I received my ticket. I found out that the methodology wasn't used for city streets (esp. ones with elementary schools).

Anyway, the next day, when traveling the state highway on which the study was being conducted (posted limit 35 mph), I saw a cop car in a parking lot. Of course, I immediately hit the brakes. I probably slowed from 45 mph to 40 mph. In the absence of any chance of traffic enforcement, I would have probably driven at 50 mph, plus or minus, depending on traffic volume, weather, etc. If the traffic engineers captured me going 40 mph on that day, how is that a valid data point? There's no control group. Bah.

As an aside, how hard would it be to have limits be posted electronically, and have the limit vary based on conditions and/or time of day? It drives me nuts to pass through a deserted downtown on a four-lane interstate at 4 a.m., and see that the posted limit is 55 mph. The limit is way too low at the time when your chance of being caught speeding is at its highest.

Steve February 26, 2007 at 1:02 pm

After skimming through the pdf file at the URL provided (I guess you need to triple click the URL then copy @ paste), I see they addressed my concerns. They say that the posted limit has little effect on the speed at which a majority of drivers travel. I must be in the minority, since I can't recall EVER driving at below the posted limit on ANY road when conditions are ideal. I do, however, try to keep it below 10 over the posted limit unless the majority of other drivers are speeding, too.

The significant number is the 85th percentile. Isn't it obvious that most drivers travelling well under the posted limit wouldn't go faster if the limit were raised? The people near, at, or above the limit are the poeple who's behavior I would expect to change more often.

The other notable item from the state's brochure was that the Basic Speed Law prohibits traveling above or BELOW a resonable and prudent speed. So everyone should be going the exact same speed? How many tickets have been written for going to slow under the prima facia speed limits (not on interstates with a posted minimum)?

August February 26, 2007 at 4:07 pm

I don't know if you've seen this or not:


Some cities in Europe are giving up traffic signs altogether. Apparently, when freed from arbitrary rules, people pay more attention and get into fewer traffic accidents. It's just amazing to realize economic principles work- even with traffic!

Mathieu B├ędard February 27, 2007 at 9:41 am

Still is pretty arbitrary, why the 85th percentile?

How can that qualify as spontaneous emergence when it is thoroughly verified and approved (and dictated?) by a national agency? This is not so much a spontaneous order speed limit, but rather some form of customs and habits democracy..

How could there ever be a legitimate "spontaneous order" over something that doesn't violate any private property rights?

Flash Gordon February 28, 2007 at 2:31 am

The 85th percentile used to be the way all speed limits were set almost everywhere. I'm glad Michigan is doing this because it is the best from a traffic safety point of view. When 85% of traffic is traveling at nearly the same speed with only 15% out of sync, fewer accidents happen. The cops can easily single out the one who is going faster than everyone else.

But this all changed when the old 55 mph national speed limit was instituted and states and cities realized there was a tremendous source of revenue to be had from speed enforcement. As a result, the emphasis on speed enforcement changed from safety to revenue. Now, most jurisdictions set speed limits at whatever level will make almost every vehicle a violator and thus a potential ticket paradise for politicians and bureaucrats.

Michigan is not doing something new, it is doing something old. Good for it. Maybe this will catch on, again. But I doubt it, the revenue from tickets is as addictive as crack to politicians and bureaucrats.

Previous post:

Next post: