The Effect Of Speed Limits On Actual Travel Speeds

Editor’s Note: The following was originally written by Jim Walker for the National Motorists Association blog, and has been republished with permission from the NMA.

I have worked closely with the Michigan State Police for several years in their pursuit of correcting as many Michigan posted speed limits to the correct 85th percentile speed level as possible. Yes, we have a very enlightened state police administration that wants to see posted limits set for safety, not revenue.

I have testified before Michigan legislative committees in support of the State Police to help explain the science involved, helped to nominate the key officers for a Governor’s Traffic Safety Advisory Committee Award which they won in 2006, and helped the police find areas of state trunk line routes (numbered highways) which should be re-surveyed because the posted limits were set far below the normal speeds of traffic.

In late 2006, the state police came to Ann Arbor and did speed studies on several state routes through Ann Arbor, parts of Business Route US-23 and parts of Business I-94. The posted limits on these trunk line routes are legally under the control of the state police and MDOT, not local authorities, but the local authorities can sometimes “push back” in the court of public opinion.

After a long period of negotiations and explanations with a city that does not want posted limits raised at all, three areas were re-posted in early 2008 with corrected speed limits raised to the 85th percentile speed of free flowing traffic under good conditions.

The City Council even passed a resolution opposing these safety-oriented changes, but they do not have legal control over state routes, so they finally agreed to the three areas to be changed.

After allowing a period of adjustment while drivers got used to the newly posted higher limits, I re-surveyed these three areas to see what changes there were, if any, in actual travel speeds.

The huge study done in 1992 by Martin Parker says there would be little change in the speeds people actually drive.

This was, of course, the result.

Actual travel speeds changed by a maximum of 2 mph in some parameters, not at all in others, and some speed points were lower with the higher posted limits. The actual traffic speeds remained the same as they have been for 23 years.

One thing did change. As was expected, the vast majority of safe, sane, competent drivers who go along with the normal flow of traffic are no longer arbitrarily defined as criminals, and no longer subject to big ticket fines and even bigger insurance surcharges.

One of my key goals is to get a reluctant Ann Arbor city government to adopt the proven practices to set the safest speed limits as described in the Institute of Transportation Engineers Engineering Handbook, the Michigan Manual of Uniform Traffic Control Devices, and the revised set of Michigan traffic laws that went into effect in November of 2006.

It is an uphill battle, because of two reasons.

First, the city makes so much money from traffic tickets that safety practices take a back seat to the revenue.

Second, the flow of misinformation and deliberate disinformation that has come out of Washington since the early 1970s has convinced many citizens that lower numbers painted on the speed limit signs means lower actual traffic speeds and safer driving.

Anyone who has read the scientific literature knows this is totally false, but a lot of education is needed to repair the damage and correct the false beliefs many people have about posted limits.

Hopefully the City Council members and others who read the charts will see the proofs that actual travel speeds do NOT rise with corrected 85th percentile posted speed limits and that will remove one counter argument for posting 85th percentile speed limits to maximize safety.


Definitions included at the bottom of the page.

History Of Speeds On North Main Street (Northern Section)

Data is from the middle of the section where the posted speed limit was corrected to 45 mph in 2008, from the former 40 mph. Data is taken at Points 1 and A on the MDOT Traffic Control Order Map.

Survey DateSep 2006Aug 2008Posted Speed Limit40 MPH45 MPH% of Vehicles Obeying Speed Limit33%71%50th Percentile Speed43 MPH43 MPH85th Percentile Speed47 MPH47 MPH90th Percentile Speed49 MPH49 MPH% of Vehicles at 50 MPH or Higher8.4%8.6%Fastest Speed Recorded55 MPH54 MPHTotal Range of Speeds29 to 55 MPH33 to 54 MPHMaximum Difference in Speed26 MPH21 MPH

History Of Speeds on Washtenaw Avenue, Near the City Club

Data is from the middle of the section where the posted speed limit was corrected to 40 mph in 2008, from the former 30 mph. Data is taken at Points 5 and P on the MDOT Traffic Control Order Map.

Survey DateSep 2006Aug 2008Posted Speed Limit30 MPH40 MPH% of Vehicles Obeying Speed Limit8%86%50th Percentile Speed35 MPH36 MPH85th Percentile Speed40 MPH40 MPH90th Percentile Speed41 MPH42 MPH% of Vehicles at 45 MPH or Higher0.7%1.7%Fastest Speed Recorded47 MPH49 MPHTotal Range of Speeds28 to 47 MPH28 to 49 MPHMaximum Difference in Speed19 MPH21 MPH

History Of Speeds on Washtenaw Avenue, Ann Arbor, Michigan

Data is from the middle of the section where the posted speed limit was corrected to 45 mph in 2008, from the former 35 mph. Data is taken at Points R & S on the MDOT Traffic Control Order Map.

Survey DateSep 2006Aug 2008Posted Speed Limit35 MPH45 MPH% of Vehicles Obeying Speed Limit4%79%50th Percentile Speed42 MPH43 MPH85th Percentile Speed47 MPH46 MPH90th Percentile Speed48 MPH47 MPH% of Vehicles at 50 MPH or Higher4.7%2.6%Fastest Speed Recorded58 MPH52 MPHTotal Range of Speeds33 to 58 MPH34 to 52 MPHMaximum Difference in Speed25 MPH18 MPH


50th Percentile: Speed at which 50% of vehicles are above that speed and 50% are below.

85th Percentile: Speed at which 85% of the vehicles are below or right at that speed.

90th Percentile: Speed at which 90% of the vehicles are below or right at that speed.

National Motorists Association
National Motorists Association

More by National Motorists Association

Join the conversation
2 of 46 comments
  • Bishop2020 Bishop2020 on Aug 06, 2010

    Back in college I was an Traffic Engineering intern for a city in California. We were required to set the speed limit to the 85th percentile and the survey had to current within 5 years or there was a risk it could be contested in court and any fines thrown out. So it was my job to go out and update these surveys armed with my trusty radar gun and city van. But here's where it got tricky, to be valid the survey had to be taken on an average day with no mitigating factors such as rain, etc. I did not however prevent they city from having a cop sit on that road night & day for a week before we did the survey. So the engineering dept would let the city know what roads we'd be updating and the cops would be dispatched to scare the regular drivers there into 5-10mph slowdowns compared to their usual driving habits, and then I'd be sent out to do the survey. If it was still too far outside the desired speed we'd scrap it and try again. So even when the law is the 85th percentile there are still ways they city can artificially reduce the speed limit unfortunately.

  • H Man H Man on Aug 06, 2010

    On a related note: There was a story posted here a few months back about an Oregon man who was on a state-long walk to protest the stupid-low speed limits here. Curious what became of him... Edit: I really could have just googled it first. ugh.

  • Pau65792686 I think there is a need for more sedans. Some people would rather drive a car over SUV’s or CUV’s. If Honda and Toyota can do it why not American brands. We need more affordable sedans.
  • Tassos Obsolete relic is NOT a used car.It might have attracted some buyers in ITS DAY, 1985, 40 years ago, but NOT today, unless you are a damned fool.
  • Stan Reither Jr. Part throttle efficiency was mentioned earlier in a postThis type of reciprocating engine opens the door to achieve(slightly) variable stroke which would provide variable mechanical compression ratio adjustments for high vacuum (light load) or boost(power) conditions IMO
  • Joe65688619 Keep in mind some of these suppliers are not just supplying parts, but assembled components (easy example is transmissions). But there are far more, and the more they are electronically connected and integrated with rest of the platform the more complex to design, engineer, and manufacture. Most contract manufacturers don't make a lot of money in the design and engineering space because their customers to that. Commodity components can be sourced anywhere, but there are only a handful of contract manufacturers (usually diversified companies that build all kinds of stuff for other brands) can engineer and build the more complex components, especially with electronics. Every single new car I've purchased in the last few years has had some sort of electronic component issue: Infinti (battery drain caused by software bug and poorly grounded wires), Acura (radio hiss, pops, burps, dash and infotainment screens occasionally throw errors and the ignition must be killed to reboot them, voice nav, whether using the car's system or CarPlay can't seem to make up its mind as to which speakers to use and how loud, even using the same app on the same trip - I almost jumped in my seat once), GMC drivetrain EMF causing a whine in the speakers that even when "off" that phased with engine RPM), Nissan (didn't have issues until 120K miles, but occassionally blew fuses for interior components - likely not a manufacturing defect other than a short developed somewhere, but on a high-mileage car that was mechanically sound was too expensive to fix (a lot of trial and error and tracing connections = labor costs). What I suspect will happen is that only the largest commodity suppliers that can really leverage their supply chain will remain, and for the more complex components (think bumper assemblies or the electronics for them supporting all kinds of sensors) will likley consolidate to a handful of manufacturers who may eventually specialize in what they produce. This is part of the reason why seemingly minor crashes cost so much - an auto brand does nst have the parts on hand to replace an integrated sensor , nor the expertice as they never built them, but bought them). And their suppliers, in attempt to cut costs, build them in way that is cheap to manufacture (not necessarily poorly bulit) but difficult to replace without swapping entire assemblies or units).I've love to see an article on repair costs and how those are impacting insurance rates. You almost need gap insurance now because of how quickly cars depreciate yet remain expensive to fix (orders more to originally build, in some cases). No way I would buy a CyberTruck - don't want one, but if I did, this would stop me. And it's not just EVs.
  • Joe65688619 I agree there should be more sedans, but recognize the trend. There's still a market for performance oriented-drivers. IMHO a low budget sedan will always be outsold by a low budget SUV. But a sports sedan, or a well executed mid-level sedan (the Accord and Camry) work. Smaller market for large sedans except I think for an older population. What I'm hoping to see is some consolidation across brands - the TLX for example is not selling well, but if it was offered only in the up-level configurations it would not be competing with it's Honda sibling. I know that makes the market smaller and niche, but that was the original purpose of the "luxury" brands - badge-engineering an existing platform at a relatively lower cost than a different car and sell it with a higher margin for buyers willing and able to pay for them. Also creates some "brand cachet." But smart buyers know that simple badging and slightly better interiors are usually not worth the cost. Put the innovative tech in the higher-end brands first, differentiate they drivetrain so it's "better" (the RDX sells well for Acura, same motor and tranmission, added turbo which makes a notable difference compared to the CRV). The sedan in many Western European countries is the "family car" as opposed to micro and compact crossovers (which still sell big, but can usually seat no more than a compact sedan).