Review of:
Portland's Blue Bike Lanes:
Improved Safety Through Enhanced Visibility
by William W. Hunter, David L. Harkey,
J. Richard Stewart
of University of North Carolina Hwy Safety
Research Center
and Portland, OR, City Staff

home                   facilities

1 Introduction

At ten locations where bikelanes approached intersections, Portland painted blue the length of the bikelane where motorists crossed it. At most of these locations there had long been regulatory signs requiring motorists to yield to cyclists. At the start of the study improved signs were installed that both diagrammed the lane pattern and required motorists to yield to cyclists. The object was to determine whether motorists or cyclists chose to operate in a safer manner after the blue paint and the improved signs were installed. The traffic at each intersection was videotaped for several hours duration before and adequately after painting the blue bike lanes, and the video records were analyzed. Conclusions were reached on the basis of the videotapes. (Other conclusions were based on answers to survey questionnaires; they pertain only to attitude, not to traffic operations, and are therefore not considered further.)

2 Types of Intersection

The investigators divided the studied intersections into three types. The quoted descriptions are taken directly from the report, but they are not exactly correct.

2.0.1 Group 1: Exit Ramps

"The cyclist travels straight ahead and the motorist crosses the path of the cyclist to exit a roadway, such as an off-ramp situation." Sites #1-4.

2.0.2 Group 2: Right Turn Lanes

"The cyclist travels straight and the motorist crosses over the cyclist path to enter a right turn lane." Sites 5-8

2.0.3 Group 3: Entrance Ramps

"The cyclist travels straight and the motorist crosses the bicycle lane to merge onto a street from a ramp." Sites 9-10.

3 Study Assumptions

3.1 Cyclist Right-of-Way

The investigators maintain, although not explicitly, that in all of these cases the cyclist has the right-of-way because he is making a straight movement while the motorist is making a turning movement. "However, cyclists are not required by law to use a hand gesture at any of the sites since the cyclist is making a through movement rather than a turn."

3.2 The Bike-Lane Superstition

The investigators considered that riding in the bike lane, rather than elsewhere, was a good action. "The increased cyclist use of the recommended path is also encouraging, since it should be the safest route through a conflict area."

This statement is utterly damning for either of two reasons. First, these supposed bicycle safety experts know so little about safe cycling that they can't decide what is the safe path through these intersections. Second, if they are sufficiently well-informed to know the safe path through these intersections, they daren't say what it is.

3.3 Most Erroneous Assumption

The investigators failed to consider any other method of cycling through the intersections than those already designated by the existing bike-lane stripes. The investigators failed to consider, and to observe traffic, with different locations for the bike lanes. The investigators failed to consider, and to observe traffic, without any bike-lane stripe at all.

As a result, only one kind of conclusion can be reached from this study. That is, that if you have a bike-laned intersection that is terrible, is causing complaints and/or collisions, then you might be able to lessen its dangers a little by painting the bike-lane blue. But it might be much better to redesign the intersection so that the conflicts that cause the collisions are either eliminated or made easier for the participants to handle. This option was not considered by the investigators; they have nothing to say about the design defects of the intersections that they treated with blue paint.

One would think that the study was specifically designed by the manufacturers of blue traffic surfacing.

4 Study Conclusions

I summarize the conclusions. Motorists more often acted as though they knew that they were required to yield, while cyclists more often acted as though they knew they had the right-of-way when on the blue-painted area. The investigators obviously consider both of these to be improvements, although they do express a little caution. "However, coupled with and perhaps resulting from the perception of increased safety appears to be declining cyclist caution (fewer cyclists turning their heads and signaling)."

5 Site Descriptions

The following descriptions of each site compare the investigators' grouping system with actuality. The first quotation is directly from the study. The rest is my description. For further information, I have consulted the Street Atlas USA program and have printed out large-scale diagrams (they are not exactly surveyors' drawings) of each intersection. Following each description is a paragraph listing the major design errors, and the proper action for the cyclist approaching in the described direction.

5.1 Exit Ramp: NE Broadway at Williams

5.1.1 Description

"Bicyclist heading west. Motorist crosses bicycle lane to access I-5 northbound entrance ramp."

This is a five-legged intersection, presumably over I-5. Signalization is present; phasing not known. Westbound Broadway (one-way W/B) reaches the intersection directly from the east. Williams goes north/south and is one-way N/B. At the intersection, westbound Broadway turns about 45 degrees to the left while the I-5 ramp turns about 45 degrees to the right. There is no straight-ahead path. The lane layout given is, counting from the curb, a motorist-only-right-turn-only lane, a left swerving bicycle lane, a motorist-left veer-or-right lane. No other lanes described, but are present. Motorists in the far right lane can turn onto Williams or the I-5 ramp. Motorists in the second motor lane from the curb must either turn right onto the I-5 ramp or veer left. The blue bikelane crosses from the left of the outside lane to the curb at the far side of the intersection.

5.1.2 Design Errors

Having a bikelane on the right of right-turning traffic when the bicycle traffic is not allowed to turn right.

5.1.3 Proper Action

Merge to the left early, so you have plenty of time and room to get to the left of the right-turning traffic before it gets to its turning points. Ride through the intersection, then merge back to the curb when traffic evens out.

5.2 Exit Ramp: SW Beaverton-Hillsdale Hwy (OR-10) eastbound at Bertha

5.2.1 Description

"Bicyclist heading east. Motorist crosses bicycle lane while veering off to Bertha Blvd."

This is a Y intersection. My map shows that the cyclist who continues along OR-10 makes a left veer while the motorist who wants to turn onto Bertha goes straight. The photographs show that OR-10 veers left, Bertha veers right. The bikelane follows the Bertha curb to the right into the Y, then cuts sharply left across the traffic going to Bertha.

5.2.2 Design Errors

Running the bike lane along the curb partially around the curve to the right, then sharply turning it across the right-turning traffic.

5.2.3 Proper Action

Merge to the left early, to be clear of the right-turning traffic before it reaches its turning point. Ride through the intersection, then merge back to curb when traffic evens out.

5.3 Exit Ramp: SW Multnomah, eastbound at Garden Home

5.3.1 Description

"Bicyclist heading east. Motorist crosses bicycle lane while veering off to Garden Home."

This is not what my map shows. Garden Home runs E/W. There is no traffic eastbound on Multnomah at Garden Home. The eastbound traffic on Garden Home comes to a 4-legged intersection. Orthogonal to the right (S) is SW 69th St. 45 degrees off to the left (NE) is the start of Multnomah. The cyclist intending to ride Garden Home eastward merely continues straight. The cyclist who intends to go along Multnomah is making a 45 degree left turn.

5.3.2 Design Errors

Carrying the bike lane for Multnomah through the intersection as if it were the lane for Garden Home, ignoring that the cyclist who intends to reach Multnomah is making the equivalent to a left turn.

5.3.3 Proper Action

If you want to travel on Multnomah, prepare properly for your left turn onto it.

5.4 Exit Ramp: Hawthorne Bridge, east end, eastbound at the McLoughlin off-ramp

5.4.1 Description

"Bicyclist heading east. Motorist exiting Hawthorne Bridge eastbound viaduct onto McLoughlin."

The cyclist eastbound on Hawthorne (E/W) comes to the intersection of SE 3rd (N/S). Also at that intersection is a 45 degree to the right ramp to McLoughlin, which is US 99 (SB). The bike lane curves around to the right, then swerves left across the traffic going to US 99 (SB).

5.4.2 Design Errors

Carrying the bike lane to the right of right-turning traffic. Having the bike lane curve to the right, as if the cyclist is turning right, and then have the cyclist swerve sharply to the left across overtaking traffic.

5.4.3 Proper Action

Merge early to the left of the right-turning traffic. Cross the intersection, then merge to curb when traffic evens out.

The following image is from the report.

5.5 Right Turn Lane: SE Madison, eastbound, between Sixth and Grand

5.5.1 Description

"Bicyclist heading west. Motorist crosses bicycle lane into right-turn-only lane onto northbound Grand."

It ain't eastbound, but westbound. Madison is one-way W/B. Grand is one-way N/B. The cyclist traveling west comes to Grand, which is US 99 (NB). At that point the parking lane turns into a right-turn-only lane for traffic going north on US 99.

5.5.2 Design Errors

None, really. The proper course is that which is described by the bike lane, but, of course, the bike lane is immaterial.

5.5.3 Proper Action

This is just a standard added RTO lane. Proceed straight.

The following image is from the report.

5.6 Right Turn Lane: SE 7th, southbound at Morrison

5.6.1 Description

"Bicyclist heading south. Motorist crosses bicycle lane into right-turn only lane onto SE Morrison."

Cyclist proceeding south comes to Morrison, where the parking lane turns into a right-turn-only lane. Morrison is one-way W/B.

5.6.2 Design Errors

None, really. The proper course is that which is described by the bike lane, but, of course, the bike lane is immaterial.

5.6.3 Proper Action

This is just a standard added RTO lane. Proceed straight.

5.7 Right Turn Lane: East End of Broadway Bridge, eastbound at Larrabee

5.7.1 Description

"Bicyclist heading east comes off sidewalk of Broadway Bridge onto roadway bicycle lane. Motorist crosses bicycle lane into right-turn-only lane onto NE Larrabee."

There is a bicycle sidepath/sidewalk on the south side of the Broadway Bridge. (Don't know about the north side.) At the point where Broadway widens to provide a right-turn-only lane for access to Larrabee, the sidepath dumps the cyclist right between the through lane and the RTO lane.

5.7.2 Design Errors

Having a sidepath or bicycle sidewalk. No matter how you manage such facilities, they always present severe conflicts with motor traffic. Probably the only solution is to provide Dutch-style traffic signals to prevent competing movements.

5.7.3 Proper Action

Don't ride on sidepaths or sidewalks. If you do, be very careful to yield before entering the roadway.

5.8 Right Turn Lane: SW Terwilliger, northbound at I-5 Entrance Ramp

5.8.1 Description

"Bicyclist heading north. Motorist crosses bicycle lane into right-turn only lane onto I-5."

According to my map, confirmed by what one can understand from the photograph in the report, northbound Terwilliger curves gently to the right, this curve continuing to become the ramp for I-5. At this point, Terwilliger veers leftward.

5.8.2 Design Errors

Running a bike lane through the right-hand side of an intersection where the cyclist will be turning left, then swerving the cyclist sharply across the path of overtaking traffic.

5.8.3 Proper Action

Merge left early to join the path of the traffic turning left toward Terwilliger, and out of the way of the traffic turning right to I-5.

5.9 Entrance Ramps: East End of Broadway Bridge, westbound at Interstate

5.9.1 Description

"Bicyclist heading west from roadway bicycle lane onto Broadway Bridge sidewalk. Two lanes of motorists from N Interstate cross bicycle lane to use Broadway Bridge westbound."

The cyclist is proceeding west toward the sidewalk of the Broadway Bridge. Two lanes of motor traffic, that was southbound on the Pacific Highway and is turning right to go west on the Bridge, come in from the cyclist's right.

5.9.2 Design Errors

Having a sidepath or bicycle sidewalk. No matter how you manage such facilities, they always present severe conflicts with motor traffic. Probably the only solution is to provide Dutch-style traffic signals to prevent competing movements.

5.9.3 Proper Action

Plan not to ride on the sidewalk or sidepath. Stay enough left in the line for straight-through traffic and for the entering traffic to enter, stay in that line until traffic has settled down, and then merge right to the normal position.

The following image is from the report.

5.10 Entrance Ramp: NE Weidler, eastbound at Victoria (I-5 northbound off ramp)

5.10.1 Description

"Bicyclist heading east. Motorist exits I-5, crosses bicycle lane as he enters eastbound NE Weidler."

Cyclist is proceeding eastbound on Weidler. He comes to a 4-legged orthogonal intersection. To his left, the north side, is Victoria. From his right enters one lane of right-turning traffic (N/B to E/B) from the off-ramp from I-5 (N/B). The ramp and Victoria are one-way N/B. Weidler is one-way E/B.

5.10.2 Design Errors

Keeping the cyclist to the right of traffic entering from the right, and swerving him across that traffic to do so.

5.10.3 Proper Action

Stay enough to the left in the line for straight-through traffic and to give room for the entering traffic to enter. Stay in that line until traffic has settled down, and then merge right to the normal position.

6 Analysis

As stated above, this study is not a study that demonstrates the value of bicycle lanes. All that this study could be intended to do is to see whether or not some blue paint and improved small regulatory signs can make dangerous bicycle lane designs less dangerous. However, this study has produced some additional, unintended, conclusions.

6.1 Well-Known Conflict Points

It has been recognized since 1972, published for all in 1977, that bicycle lanes create conflicting movements between bicycles and motor vehicles in two situations: when cyclists turn left and when motorists turn right. Each of the intersections studied in this investigation presents one or both of these conflicting movements. Nothing at all was done in Portland to reduce these conflicts. All that was done was to try to get motorists to behave more safely at the conflict points created by the bicycle lanes.

In a way, the experiment succeeded: motorists more often yielded to the cyclists making the conflicting movements. However, the cyclists also reduced their frequency of yielding at the same conflict points.

6.2 Failure to Recognize Right-of-Way Conditions and Right-of-Way Law

This change was that desired by the bike-lane designers and investigators. That is, they wanted to demonstrate that in all of these situations the cyclists have the right-of-way to which the motorists should yield. I repeat the most relevant quotation from the report concerning right-of-way: "However, cyclists are not required by law to use a hand gesture at any of the sites since the cyclist is making a through movement rather than a turn." That not very specific statement is the most specific statement that the investigators were at all concerned with analyzing right-of-way at these locations. In short, the investigators thought that the subject was either so unimportant, or so cut-and-dried, that no more specific mention was required.

6.3 Failure to Understand Traffic Law

I doubt that Oregon law says what the investigators say it does. If it does, the investigators should have criticized it for being dangerous. I repeat the relevant quotation again: "However, cyclists are not required by law to use a hand gesture at any of the sites since the cyclist is making a through movement rather than a turn." What the investigators are saying, in this vague statement, is that what matters when right-of-way is being considered is not the immediate movement but the final destination. The investigators are saying that if a driver intends to leave the intersection traveling in his original direction, then that driver has the right-of-way over all turning drivers, regardless of whatever swerves and wiggles the straight-through driver makes on his way through the intersection. That is not the law in any state that I know of. If it is the law in Oregon, then whatever has been investigated about blue bike lanes in Oregon is valid only where such a foolish law exists. Obviously, the investigators are arguing that their study is valid nationwide because this is what the laws say all across the nation.

6.4 Actual Traffic Law

I refer to the relevant provisions of the Uniform Vehicle Code as these are generally the law throughout the land. Variation in these particular provisions is very small.

  1. 1: "Right turns: Both the approach for a right turn and a right turn shall be made as close as practicable to the right-hand curb or edge of the roadway."
  2. 2: "Left turns: The driver of a vehicle intending to turn left shall approach the turn in the extreme left-hand lane lawfully available to traffic moving in the direction of travel of such vehicle. ..." UVC 11-601
  3. 3: "Turning movements and required signals: No person shall turn a vehicle or move right or left upon a roadway unless and until such movement can be made with reasonable safety nor without giving an appropriate signal in the manner hereinafter provided." UVC 11-604

Obviously, the laws regarding the starting points for turning movements can apply only where actual turns are possible. That is, to intersections and similar places, such as where driveways exit the roadway. However, the right-of-way provision covers both turning a vehicle and "mov[ing] right or left upon a roadway."

The person turning the vehicle has to yield to other traffic. However, the turning position laws say that the turning driver does not have to yield to traffic coming up behind him, because he is required to start his turn from a position in which he cannot be lawfully overtaken by traffic on the side towards which he will turn. This has been very carefully thought out for over seventy years, and has not been found defective.

The driver moving right or left upon the roadway has to yield to traffic in his new line of travel. That's the meaning of: "until such movement can be made in reasonable safety." Quite clearly, this law applies both to turning movements and to changing lateral position on the roadway. That movement can occur almost anywhere in the highway system; it is not limited to intersections or similar places.

6.5 Bikelanes Generally Disobey Traffic Law

Quite obviously, the designs of most of the studied intersections violate the requirements of the relevant traffic operating laws.

6.5.1 The Single-RTOL Exception

The only studied intersections that do not disobey standard traffic operating laws are the two in which right turns are allowed only from an added right-turn-only lane where the parking lane becomes the RTOL. At these locations, the driver intending to turn right merges (moves laterally) into the new lane before making his turn. If he performs the merge properly, he yields to whatever traffic is already in that lane. I have always recommended this design because it gives the motorist the time and distance to merge across the cyclist's path rather than, as at an intersection without such a lane, swerving across the cyclist's path. Of course, without an RTOL, the motorist is required to start his turn from a position in which there is insufficient room for a cyclist on his right. However, motorists often fail to obey this closely, and motorists driving long vehicles cannot, no matter how hard they try, turn from such a position.

6.5.2 Bikelanes That Disobey Traffic Law

All the other studied intersections disobey one or more of the standard traffic operating laws. Just look at the descriptive paragraphs above for confirmation of that. They generally place the straight-through or left-turning cyclist on the right of right-turning traffic. Some make the straight-through or left-turning cyclist follow the right-hand curb part-way around the corner, and then require him to make a sharp 90 degree left turn across the right-turning traffic. Surely, such a movement by the cyclist counts as either, or both, a left turn from the right lane or a movement turning right or left upon a roadway. One is unlawful per se, the other requires that the cyclist yield before making the turn.

Two intersections, of course, have the chaotic traffic pattern imposed by the requirement (or is it merely the encouragement?) that the cyclist enter or exit a sidepath.

6.6 The Investigators' Guilt

6.6.1 Generally

Despite the obvious fact that all of these intersections (except the two with added RTOL) require traffic to violate the basic traffic operating laws, and that these violations are the root cause of whatever problems might be there, none of the investigators chose to make such a comment. More than that, the investigators chose to try to cover up the design deficiencies by advancing the clearly absurd argument, never supported by traffic law, that only the final direction counts in considering right-of-way, not the swerves and wiggles in reaching that exit from the intersection. The investigators advanced this absurd argument because they believed, either that cyclists did have the right-of-way in these intersections, or that cyclists should have the right-of-way over motorists. They were reduced to advancing this absurd argument because there are no reasonable arguments to support their belief.

So much for the legal absurdity. There is also the physical absurdity shown by the physical facts about the intersections. Often, the cyclist following the bike lane is expected to curve to the right, following the right-hand curb, as if making a right-hand turn, and then, suddenly, without any distance for merging, to turn 90 degrees to his left, across the right-turning traffic. One cannot reasonably consider this a straight-through course. At some other intersections, what the investigators describe as this straight-through course is actually a left turn. In the case of Garden Home and Multnomah, even the street-name changes.

Now consider the traffic operating problems caused by these violations of standard traffic law. We have motorists trying, despite the other demands on their attention caused by the intersection, to observe, and to yield to, traffic overtaking them. They just cannot devote much attention to that part of the task. This has been formally described since 1972, published nationally in 1977, but this knowledge was obviously a strong factor when the original vehicle codes were being written in the first third of this century. We have cyclists encouraged to ride straight on into this dangerous situation, which would normally be unlawful, in the hope that the motorists will both understand this unusual duty and have the ability to perform it. We have cyclists encouraged to make sharp turns and swerves, otherwise unlawful, without yielding to the affected traffic, in the hope that this traffic will understand their plight and avoid their movements.

There is nothing mysterious about these matters, legal, physical, or operational. I am sure that had Portland's highway engineers designed such lane patterns for motor traffic, the investigators would have soundly criticized the designs, if, indeed, the designs had survived long enough to justify an investigation.

6.6.2 Specific Investigators' Guilt

The plain fact, and it cannot be doubted, is that either one or both of two conditions exist.

  1. 1: The first is that the investigators are bare-faced liars who are willing to write any report for which the money is forthcoming, or, more likely, which supports the desires of the powerful organizations which support their research programs.
  2. 2: The second is that the minds of the investigators are so discombobulated (this seems to be the appropriate word for this condition) that, whenever they think of bicycle traffic, they become incapable of applying the traffic-engineering and motoring knowledge that they, as traffic experts, must be presumed to possess.

7 The National Bicycle-Traffic Program

7.1 The Dangers of the National Bicycle Traffic Program

What we have is one more example of the desire of the most powerful highway organizations to control cyclists as rolling pedestrians instead of as drivers of vehicles, and of their willingness and ability to pay for pseudo-scientific propaganda to support that desire. What we have is one more example of the complete mental inability, by traffic professionals, to comprehend the standard traffic-engineering principles that they supposedly know, whenever they start to consider bicycle traffic. The immediate result, in the case of this investigation, is unjustified praise for reducing the dangers of bike-lane designs that should never have been produced in the first place.

What we have, in the long run, is subversion of the traffic-operating knowledge that allows our traffic to operate smoothly and safely. The superstitions embodied in bike-lane design encourage both motorists and cyclists to operate in dangerous ways that contradict the standard traffic laws, in the mistaken belief that these are safer.

The subject report is one more piece of evidence that the National Bicycle-Traffic Program is being run according to the Cyclist-Inferiority Phobia, which provides its only pseudo-intellectual support.

7.2 How to Rescue the National Bicycle-Traffic Program

There is only one known way for cyclists to operate safely and efficiently in areas where there is other traffic. That is, for all that traffic to operate according to the traffic laws for drivers of vehicles. We have made that impossible in the case of pedestrian traffic, so that it is extremely dangerous for cyclists to operate in the normal manner when around pedestrians. We are on the road to making that impossible in the case of cyclists and motor traffic. That course must be reversed. The National Bicycle-Traffic Program must be changed to facilitate and require that both motorists and cyclists operate according to the standard traffic laws for drivers of vehicles, with facilities designed accordingly.

We must replace the present national bicycling policy of the Cyclist-Inferiority Phobia with the Vehicular-Cycling Principle.

Return to: John Forester's Home Page                                     Up: Facilities