4-to-3 Road Diets: The #1 Thing Cities Can Do Right Away to Improve Quality of Life

I have been re-reading Clay McShane‘s Down the Asphalt Path, an absolutely fascinating history of early 20th century urban engineering and street design debates. The history of asphalt may not sound interesting to you, but the book is absolutely filled with facts that amaze and astound, particularly if you pay any attention to contemporary transportation issues. For example, in the final chapter, McShane describes how once automobiles had started to become more commonplace in US cities, engineers in urban public works departments began re-designing cities to accomodate them:

The municipal engineering manuals recommended increasing the standard street width, so repaving projects frequently included some widening, usually by cutting back on sidewalks, a relatively inexpensive expedient. New York widened 23rd Street, an important crosstown street, in 1910. Two years later, the city cut back Fifth Avenue’s sidewalks to add two new ten-foot lanes for traffic. Later the city widened Madison Avenue, a project favored by merchants on that street, but opposed by residents. Baltimore widened its streets a little after the great fire of 1904 and San Francisco did the same after the earthquake, but only on a piecemeal basis. Public works departments in other cities also cut back on sidewalks. High condemnation costs limited most building demolition to the occasional elimination of bottlenecks. Municipal engineers also increased the radius of downtown curbs to allow cars to turn more easily. This made it slightly more difficult for pedestrians to cross the street. Widening streets often meant cutting down street trees. Abutters objected, but the demands of suburban commuters and shoppers for parking spots almost always took priority. Thus, urban residents not only lost the use of the street itself for social, recreational, and commercial purposes, but much of their sidewalk and many trees as well. Even the limited success of these street widening programs suggests the political power of automobility, since abutters had only allowed widening under the rarest of circumstances in the nineteenth century.

As early as 1907, a far-sighted writer in Municipal Journal and Engineer had noted that street improvements had not eased traffic mobility, but “the result has appeared to be exactly the opposite.”

(McShane 217)

This kind of history shows you how very old current debates over street design really are, that the exact same conversations about parking, traffic, and pedestrian comfort have been going on in US cities for over a century. And it’s only recently, in the past five or ten years, that McShane’s “urban residents” (on foot, on transit, or on bicycle) have begun to win a few skirmishes over the allocation of our streets and public spaces. After one hundred years of widening streets for cars, only in the last decade have cities begun thinking about different priorities. That’s not a very good track record.


Your classic road diet before picture.

Your classic road diet before and after picture.

What Can Our Cities Do Now, Right Away?

I was thinking about this recently because of a conversation I had with a local politician a while ago, who had asked me, “What can I do in my city to help improve cities for biking and walking? Give me ideas of projects that I can push for.” (Yes, this actually happened. It wasn’t just a dream.)

It took a while to figure out an answer. So many of the interesting ideas for re-designing cities seem to be politically unattainable. (David’s post yesterday about covering urban freeways is a good example. While those projects might be great ideas, they’re all extremely expensive, and would require Gordian political and logistic feats.) Most urban design projects, from new transit investments to a new bridge, require the approval of multiple levels of government, and multiple sources of funding. Even streets redesigns become complicated very quickly because a great many streets are the jurisdiction of either the state or the county, and thus subject to federal design guidelines. (This gets wildly complex really quickly, believe me.)

So the question remains: What is the easiest, most important thing that cities can do right away, on their own, to improve quality of life?

The Answer: 4-to-3 Lane Conversions on Arterial Streets

For any 4-lane street under about 20,000 cars per day, cities should re-stripe the street to have one traffic lane in each direction, a center turn lane, and improved pedestrian and bicycling space. The costs of doing this are minimal, but the benefits for cities and the people who live in them are immense.

The main benefit of a 4-to-3 conversion has to do with safety and quality of life. While safety improves for everyone, including drivers, it dramatically improves for  the most vulnerable road users, people who live and walk along these main arterials. At first, a simple change in the lane configuration may not seem like much, but to understand why they make such a big difference, think about the different road users and how they behave in each of these situations.

First, from a driver’s perspective: how you behave on a 4-lane configuration is really different from a 3-lane configuration. With two lanes in each direction, drivers are always thinking about passing each other. Driving down the road, you’re constantly scanning the cars ahead of you to see if they are slowpokes, or if they’re making a (dreaded) left turn. Cars are continually switching back and forth between the two lanes, glancing over their shoulders to see if the person in the next lane will let them in, speeding around turning and slowing cars and trucks. (For a good example, drive down Hennepin Avenue pretty much anytime.) This kind of situation means that car drivers aren’t paying much attention to the sidewalks, crosswalks, or looking out for bicyclists. This configuration facilitates speeding, and creates lots of dangerous automobile movements especially at intersections.

Contrast that wtih a 3-lane configuration. Here, each car has to follow the one in front of them. Left turning vehicles move to the center, and drivers are never stuck behind them. Nobody is passing anybody, and traffic moves along with the stoplights. Not only is the whole situation is more relaxing, but its far safer for everyone involved. Speeds are lower, and most importantly, you don’t have lane changes occuring at intersections or driveways.

From a pedestrian or cyclist perspective, the difference is even more palpable. Anyone trying to cross a 4-lane street has to deal with multiple lanes of traffic moving at different speeds in different directions. That’s a recipe for disaster.

Conversions almost always come with more space allocated to the sidewalk and pedestrian realm. For anyone walking or biking along a 4-lane street, its a highly unpleasant experience. If our cities want to encourage walking and active transportation, road diets are an absolute minimum step to take.


Some Twin Cities’ Examples

Riverside Avenue before and after its 4-to-3 conversion

To really experience the difference between these two treatments, look at some examples from the Twin Cities. As I’ve written before, Franklin Avenue in Minneapolis is an excellent place to compare the experience of both 4- and 3-lane configurations. A few years ago, Minneapolis converted Riverside Avenue to a 3-lane configuration, and it changed a dangerous street into a comfortable one. (The design has changed again since then, and been cast in stone.) Nicollet Avenue is another case, and probably the best north-south arterial in the city for bicycling and walking. In St Paul, there are a few more examples: much of West 7th Street is a 3-lane road, and Lexington Parkway south of Grand Avenue has been converted in ways that improve the quality of life for anyone living nearby. (For anyone curious about the difference design can make, go for a walk on one of these streets.)


The Minneapolis bike-car crash map


A cyclist on the 4-lane portion of Franklin Avenue

At the same time, there are lots of examples of streets that could sorely use this treatment, city controlled streets that remain in their dangerous 4-lane configurations. In Minneapolis, you can tell where these streets are by simply glancing at the bicycle-car collision map released yesterday. NE Broadway, Lyndale, and Franklin Avenues are obvious candidates, streets that are familiar problems for anyone walking or biking in those neighborhoods.

In St Paul, there was a recent push by neighbors during public meetings about converting Hamline Avenue from a 4-lane to a 3-lane configuration. Given the need for North-South connections in the city, the university in the area, and the coming Hamline Avenue light rail stop, it would be an ideal candidate. Parts of Maryland, Cretin, and Dale Avenues are other obvious candidates. (Also, pretty much any street on the East Side.)


It can’t be done. It’s actually pretty easy.

So often during debates over transportation projects, politicians and engineers will say “It can’t be done.” Pick any project, and either there’s no money, or arcane federal regulations won’t allow it.

Well, 4-to-3 conversions on our unsafe arterial urban roads are the exception to that rule. These kinds of conversions are doable immediately. They’re both cheap and practical. Compared to basically any infrastructure project, the far and away easiest way to re-make our cities is through a different application of paint and signage.

The difference between a 4-lane and a 3-lane configuration of a road is huge. A street that is currently dangerous for drivers, pedestrians, and cyclists becomes vastly safer overnight. A street that encourages speeding becomes one that encourages people to obey the law. 4-to-3 conversions make a significant difference for anyone nearby, particularly the most vulnerable users of the road. For a hundred years, cities have been taking space away from the sidewalks, until walking and biking in the city has become dangerous and unpleasant. Its time we started reversing the trend.

39 thoughts on “4-to-3 Road Diets: The #1 Thing Cities Can Do Right Away to Improve Quality of Life

  1. Matty LangMatty Lang

    Great article Bill. In case any readers are interested in some more visual documentation of what a Road Diet accomplishes, I'll leave a link to a video I made for TLC a couple of years ago documenting the transformation of 10th Ave SE near the U campus:

  2. hokan

    If that 4-into-3 conversion includes bike lanes, a biker's life may be made easier, but if it doesn't (Eat Street) than that conversion makes things worse. Biking on a 4-lane road is less intimidating because motorists can easily pass. Not so on a road with a single lane in each direction.

    1. Bill LindekeBill Lindeke

      I have to disagree Hokan. Nicollet around Eat Street is one of my favorite spots to ride a bicycle. The middle turn lane makes all the difference. I find it to be both safe and comfortable. (Actually, I'm headed there right now.)

      Selby Avenue in St Paul is a similar case of a traffic calmed street w/out bike lanes, and I ride that all the time as well. It almost functions as a bike boulevard.

      1. hokan

        How does the middle lane help you, Bill? Do people use it illegally to pass other traffic? I've heard that in some places (Ohio for one) that it is legal to pass slow moving traffic in a no passing zone, in some situations. Should we change the laws here to accommodate that behavior?

        1. Bill LindekeBill Lindeke

          I'd be OK with that. My point is that I'd rather bike on Nicollet than just about any other main commercial street in the city. It reminds me of Portland OR commercial streets, like Hawthorne Avenue.

  3. Matt SteeleMatt

    A related discussion had been going on over at the Strong Towns network where I stood up for the usefulness of 4 to 3 lane conversions, even if the result is rather stroady. http://www.strongtowns.net/forum/topics/3-lane-st

    To sum up my position: A 3 lane stroad, as a matter of design principle, should be avoided: We can and should do better when starting from the drawing board, and there's a lot more that can be done besides lane striping to convey the nature of a facility as a street or a road. A 3 lane stroad, as a way to calm a 4 lane stroad, is obviously a cheap and pragmatic improvement that helps neighborhoods reclaim their streets.

    1. Nathanael

      I’m trying to remember where the study was, but I remember reading a study which said that stroads with 4 traffic lanes had the worst crash rate of any type of road or street….

      …except for stroads with more than 4 traffic lanes.

      3-lane stroads weren’t nearly so bad. The weaving traffic apparently induces a lot of crashes and eliminating it helps a lot.

  4. Allen

    Is the east end, the 3 lane portion, of Franklin, still a mess during peak usage in the morning and afternoon?

    1. Bill LindekeBill Lindeke

      Disclaimer: I don't drive on it, so its hard for me to say exactly. But I suppose it depends on what you mean by "mess." If you mean that traffic is moving slowly, then I suppose it is. It probably would earn a poor "LOS." But does it still function for all modes, and get people where they want to go safely? Yes, it does. It's far better than the central portion of the street, which likely has a higher LOS but is more dangerous for everyone.

  5. Reuben CollinsReuben Collins

    I agree with this almost completely, my only reservation being that I think there may be some costs you've glossed over a bit. I love 4-to-3 lane conversions, and recommend them to clients whenever possible (though they don't always go for it….).

    I'm with hokan a bit in pointing out that 4-to-3 lane conversions aren't always a dream for peds/bikes. If you're not moving curbs, you don't really end up with more pedestrian space, and 4-to-3 conversions aren't always a boon for cyclists either.

    Still, they result in some great traffic calming, and just greatly simplify things.

    There are some barriers in MN – State statute doesn't easily permit 3-lane roads on state-aid routes with 15,000 vehicles per day or greater. It can be done, but there are many hoops to jump through. Of course, there are many 4-lane roads carrying less than 10,000 vpd, which are no-brainers for conversion.

    Depending on conditions, there can also be costs associated with the traffic signals along the route. If you do nothing to the signals, they may be poorly timed for a 3-lane section, and the heads will be in the wrong spot. So sometimes a project like this will include some costly signal revisions. Still, your overall point that this is way cheaper than some other projects is spot on.

    1. Matt SteeleMatt

      WRT signals, I think this stems from overengineering of signal systems, applying high speed road signal design principles to stroads. They put heads over every lane and are obsessed with protected left turns.

      We could save a lot of money in the short run AND be more flexible in the long run if we used simplified signals. We already do this on urban streets in the core cities. Squeeze in a protected left turn lane? No need to redo the signal mast. Want to add a protected left turn lane? Just change out the head from a 3 light to a 5 light.

      Of course this bypasses the fact that with better street design (stroad avoidance), slower speeds, etc we could reduce the need for expensive traffic control devices altogether…. up to and including shared space.

      1. Sean Hayford OlearySean Hayford Oleary

        Agreed. Protected left-turn signals do have some advantages, particularly for elderly drivers who may have more difficulty judging speed and distance. However, the cost and inefficiency that result are torturous. (Even more unbearable are the City of Minneapolis timed signals that include a left arrow, whether a car is in the left-turn lane or not.)

        I do worry that simple signals need roads to match, though. If you take a large intersection (say, American Blvd and Lyndale), it could become a real issue for left-turning drivers to keep track of pedestrians in addition to oncoming traffic. You could face more pedestrian collisions due to the permissive left turn.

        1. Matt SteeleMatt

          Bloomington STROADS are a lost cause, given their existing geometry. At least Bloomington arterial streets had potential back when they were four lane undivided, before the city rebuilt so many as six lanes with double left turn lanes… Think of the unfunded liabilities!

        2. Sean Hayford OlearySean Hayford Oleary

          Yes, perhaps American and Lyndale is an unfair example. I would note that the vast majority of major streets in Bloomington are still of an older variety. Except for the free-right turns, they're mostly pretty basic 44'-or-so streets with conventional intersections. It's only major streets east of Cedar (and all of American Blvd) that are of the auto-dystopia mega-stroad type.

    2. Sean Hayford OlearySean Hayford Oleary

      I also agree with Hokan that four lanes can facilitate easier bike mobility than many three-lane roads without bike lanes. Sometimes, like on 86th St in Bloomington, the resulting bike lane/shoulder is inadequate for safe bike use anyway. While I like these conversions on certain medium-volume streets, I worry that some jurisdictions are using them on streets that don't require a center turn lane at all — taking space away from bikes or on-street parking.

      Reuben, what's the state aid requirement for >15000 ADT to be four lanes? All I could find was this, which only addresses lane width. And I wouldn't fret too much about signal heads being in the wrong place. At 70th/Portland in Richfield, the signal heads are still set up for a four lane configuration. I don't think anyone's been unduly confused by two green or red lights instead of one ;).

      1. Reuben CollinsReuben Collins

        Keep reading near the bottom of the page….

        "For volumes greater than 15,000 projected ADT, at least four through-traffic lanes are required, unless a capacity analysis demonstrates that a different lane configuration achieves level of service D or better."

        At first glance, the requirement to demonstrate LOS D or better is not too onerous (since most agencies require designers to demonstrate LOS D or better for ANY design). However, it should not be underestimated.

        1. Nathanael

          Interesting to know.

          It is incumbent upon cities to refuse the state aid when necessary. Roads with intersections and four through traffic lanes are *dangerous*. They shouldn’t exist at all.

      2. Sean Hayford OlearySean Hayford Oleary

        Ah, should have read a little more carefully. Minneapolis and Hennepin County have clearly been able to sidestep the 4-lane requirement in many existing streets. Cedar/CSAH 152 had 17k ADT in 2008 and is two lanes. And the recently redone Lyndale/CSAH 22 had nearly 20k ADT in some areas. I suppose in these cases they're able to demonstrate LOS D?

        Those are both relatively suburban/residential areas, however. I imagine it would be harder to keep a smooth flow of traffic on many of the more urban undivided 4-lanes.

        1. Matt SteeleMatt

          I cannot STAND that they call these Minneapolis streets "County State Aid Highways" — no, it is Cedar Avenue, not CSAH 152. I know this wasn't your intent when you cited the legal name for it and the county does have jurisdiction, but that's the problem. Turn it back to the city and get the county out of decisionmaking for local streets!

        2. Sean Hayford OlearySean Hayford Oleary

          Yeah, I mean, I actually give my boyfriend a hard time for referring to Cedar as "Highway 77" for the freeway portion. I use the CSAH numbers here because a.) it reminds us that it IS Hennepin County's street and b.) it reminds us that state aid restrictions apply.

          I actually don't think municipal turnback would do Cedar much good. Hennepin County has genuinely gotten much better at creating streets as valuable public spaces, and recent projects attest to that. See Lowry (pardon me… CSAH 153) and Lyndale. Lyndale's a crappy design for bikes, but they're both pretty well-designed for their neighborhood function, and create pretty good pedestrian environments. The Portland/Park projects are also good examples of positive recent change.

          Even if Cedar (north of 62) were turned back to to Minneapolis, it would likely still be an MSA route, and still subject to state aid standards. It seems like it would be minimal benefit and create new financial burden for a City that can't afford the streets it already has.

          1. Matt SteeleMatt

            Too much regulation in general. I understand why there are strings attached to intergovernmental transfers, but it just bugs me. I'd like to see Cedar reworked similar to Lyndale. I'd LOVE to see the Cedar bridge over Lake Nokomis removed, or the Crosstown interchange modified in a way that doesn't dump a freeway onto a city street. Either of these would push traffic to 35W for longer-distance north-south trips.

            Anyways didn't mean to harp on you too much about the county road names, just the concept in general.

  6. bb

    I would also like to see some sorta law regarding double yellows and allowing people to be able to pass safely.

  7. Sean Hayford OlearySean Hayford Oleary

    The vast majority of vehicles will pass a bike, even it means crossing a solid yellow line. Like some of the technically illegal things that cyclists do (rolling stops, etc), I'm happy to look the other way if cars feel the need to pass me in the center lane.

    The only cars that do not pass me crossing solid yellow are cop cars — and even that's happened once, a Minneapolis squad car (no lights or sirens) passing me on Nicollet.

  8. Matt Korner

    Even better is a road diet that adds bicycle/N.E.V. lanes that are protected by a layer of on-street parking, which, additionally, helps narrow the street and slow the traffic.

    The highest throughput for any thoroughfare, incidentally, is achieved at 30-35 miles per hour. So, a road diet can also increase capacity.

  9. Katie Urey

    Another place where pedestrians lost out to our love of cars is in sidewalks on bridges. In Portland, Oregon, (someone should confirm this for all urban bridges..) the bridges built before the 1920's have sidewalks, some wide, on both sides. Bridges built in the 1920s (give or take) such as the Ross Island Bridge and the Sellwood Bridge have narrow walkways on one side of the bridge. Pedestrians shun and fear crossing either while the older bridges are gateways between vibrant communities.

    -Katie Urey

    Portland Oregon

    1. Froggie

      Haven't quite had that problem in the Twin Cities, at least with the major river bridges. All in all, MnDOT has been pretty pro-active in adding bike/ped accommodation on major river bridges…the only exceptions I can think of offhand being the downtown bridges (due to local parallel bridges immediately next door) and the I-35W Minnesota River bridge which is still an original bridge that hasn't been touched since the early 80s.

      1. Janne

        Isn't it that they've been proactive… when pushed by other jurisdictions? I know the City of Minneapolis has been pushing MNDOT pretty hard. (And as a pedestrian, a parallel bridge "immediately" next door isn't very interesting to me.)

  10. Faith

    I don't bike on 4 lane roads if I can help it and avoid Lyndale, Franklin and Hennepin. I don't mind biking on Nicollet in the Eat St section because it feels like shared space. It's rare to see any vehicles above 30 mph and most cars and buses will pass by at about 20 mph in the middle lane, giving me a lot of passing room. When Blaisdell was really potholed in 2011, Nicollet was vastly preferrable and I still like it better since the hill near Franklin is not as steep.

  11. Froggie

    It's a funding game, Matt. The state Constitution guarantees a percentage of state gas tax/registration fee revenue to both the County State Aid Highway (CSAH) and Municipal State Aid Street (MSAS) systems, but state statutes also limit how much mileage can be put on both. This is a case where, if Cedar Ave wasn't a CSAH, the city would either have to take other streets off the MSAS system to put Cedar Ave on, or they'd have to pay for maintenance/improvement of Cedar Ave completely out of the city budget. By keeping it as CSAH 152 (a reference to the old state highway 152 of decades ago), the city can (through the county) apply for CSAH funding for improvements, and can use its MSAS mileage for other city streets.

    1. Nathanael


      I grew up in an extremely rich incorporated suburb of a fairly wealthy small town. When they wanted something done in a particular way, *they did it that way*, and if they had to pay local taxes to do it, *they did so* rather than accepting the conditions of state and federal money.

      Now I live in a less rich suburb and they’re agreeing to dangerous road widenings in order to get the state and federal money. *This is a mistake and should not happen*. If the state or federal government want you to make your road less safe, turn down the money. Lives are worth more than state money.

  12. Aaron

    I agree with everything you wrote but transit has been overlooked. My city recently did a 4-3 on a bus route with frequent stops. With the 4 lanes, buses simply stopped in the curb lane and traffic passed in the inside lane. Also buses were not delayed in restarting because they didn't have to wait to reenter traffic. Now buses have to stop in the curb/bike lane and half the full lane. And although the center lane is supposed to be only for left turns and pedestrian islands, motorists usually cross the line to pass stopped buses illegally anyway. And now buses have to yield to bikes ot reach the curb and to passing motorists to get back into the lane.

    1. Nathanael

      Well, if you’ve got a major bus route, it might make the most sense to have two moving lanes and two BUS ONLY lanes.

  13. Janne

    Just after reading this article, I rode from 26th to Franklin on Nicollet, and I got buzzed twice by cars not willing to move over to pass. I really dislike that stretch of Nicollet, if only because I have to be aggressive about taking my lane when it doesn't feel like I should have to. That's not an issue with 4-to-3 conversions, but rather sharing my opinion that Eat Street is NOT fun for me on a bike.

  14. Bill LindekeBill Lindeke

    hm. good to know. nice example of how there can be large differences b/w where people's comforts lie, even with two people (like you and me) that ride all the time in lots of different environments.

Comments are closed.