Why Buses and Other Things Should Be More Dangerous

Jeff Kaufman writes:

Buses are much safer than cars, by about a factor of 67 [1] but they’re not very popular. If you look at situations where people who can afford private transit take mass transit instead, speed is the main factor (ex: airplanes, subways). So we should look at ways to make buses faster so more people will ride them, even if this means making them somewhat more dangerous.

Here are some ideas, roughly in order from “we should definitely do this” to “this is crazy, but it would probably still reduce deaths overall when you take into account that more people would ride the bus”:

  • Don’t require buses to stop and open their doors at railroad crossings.
  • Allow the driver to start while someone is still at the front paying.
  • Allow buses to drive 25mph on the shoulder of the highway in traffic jams where the main lanes are averaging below 10mph.
  • Higher speed limits for buses. Lets say 15mph over.
  • Leave (city) bus doors open, allow people to get on and off any time at their own risk.

Other ideas?

Excellent recognition of tradeoffs. Pharmaceuticals should also be more dangerous.

Hat tip: Slate Star Codex.


> Allow buses to drive 25mph on the shoulder of the highway in traffic jams where the main lanes are averaging below 10mph.

If only there was some existing concept where some portion of a street were allocated for buses and other high passenger density vehicles in order to protect them from traffic jams.

Paris, Bogotá, Seattle, Istanbul, Tehran...?



wait, no, that's an error, not on the highway

Beautiful city, either way.

You imply dedicated lanes. These may be superior, but they don't yet exist. Shoulders exist and are largely unused.

Well, sure - HOV lanes and dedicated HOV highway sections are real common where Prof. Tabarrok lives. Rt. 66 between the Fairfax and Arlington campuses of GMU, for example.

Though possibly, he has never noticed such features of the Northern Virginian driving landscape.

shoulders have a bad habit of disappearing under bridges.

Dedicate lanes more even out the difference between cars and busses by slowing down cars rather than speeding up busses.

Slowing down cars intentionally would decrease fuel efficiency and increase emissions.

Yes, it does. So I guess the dedicated lanes aren't really about efficiency, are they?

Well I guess they help bus efficiency somewhat, but the real reason is to make car diving more painful. Punish the sinners.

Depends how many drivers leave the car at home and take the faster bus. However, Hamilton Ontario recently instituted a bus only lane, and reversed the decision within just a few months after observing that the effect on traffic was just too bad with only marginal benefits for the buses (not that big of a city, traffic wasn't that bad in the first place until they put the bus only lane in).

It is nuts to have a dedicated lane that get used once every 15-30 min. In my city we also added a separate signal for them. There is a wait for cross traffic, then the bus, then you go. It has basically shut down a main street in a busy downtown.

Pittsburgh has a series of bus-only roads carved from old rail right of ways. These bypass the infamous tunnels that make getting into town from the South, West, or East such a nightmare. During rushes, the main sections of the busways see busses every couple minutes. They also serve as clear routes for police and emergency vehicles.

It is nuts that a bus only comes along once every 15 minutes. Most main streets of cities I've lived in have had 15 buses each minute, not the other way around.

@Deek where on earth have you lived that buses come every 4 seconds? Ok, that is obviously exaggeration, but I think every 15 minutes is pretty typical except for main routes in very large cities.

Ha ha this is obviously not true. If a car had its own uncongested lane it would be able to go faster. Giving a bus its own uncongested lane allows it to go faster.

The claim was that cars are slowed down more than buses are sped up--not that buses do not speed up at all.

Given that a dedicated bus lane is artificially lowering the utilization of that lane, it seems a quite defensible point.

"Allow buses to drive 25mph on the shoulder of the highway in traffic jams where the main lanes are averaging below 10mph."

The shoulders are used by emergency vehicles to respond to accidents. Which is probably why the main lanes are moving so slowly. Having buses use the shoulders would be dangerous to the entire highway system.

"Which is probably why the main lanes are moving so slowly."

To explain, if there's an accident then the highway slows down. At that point the shoulders are needed for emergency traffic. So you wouldn't want buses blocking it in those cases.

There are no bus stops on highways, and not enough buses on the highway to fully occupy the shoulder (i.e., not enough buses on the shoulder to lead to slowing down). I don't imagine this would affect the speed of first responders much, if at all. If there are enough buses on the route for this strategy to lead to congestion on the shoulder, then a dedicated bus lane is warranted anyways.

It only takes a single bus to dramatically slow down an emergency response vehicle, as there won't be room for the bus to get out of the way if there's a traffic jam to the left.

This works until the first bus breaks down on the shoulder.

I'm generally in favor of buses, but be realistic about what you are proposing.

@JWatts, but there are plenty of urban highways where there is no breakdown lane, as well as urban highways where the breakdown lane can be used by auto traffic during peak periods. Also, the breakdown lane is used for breakdowns, which block the lane far more than a bus that can move out of the way.

There may be a small cost to the proposal caused by increased emergency response times, but I don't think that would be decisive when added to the cost-benefit analysis.

"Also, the breakdown lane is used for breakdowns, which block the lane far more than a bus that can move out of the way."

I think this point is a little weak. First, breakdowns blocking the shoulder are reasonably rare, and if they aren't rare in a given area then you can't use the shoulder for buses in any case. Secondly, a loaded bus is a far larger obstacle to get around than your average commuter vehicle. The bus is large and will have all 4 wheels squarely on the asphalt.

"... but I don’t think that would be decisive when added to the cost-benefit analysis. "

I do agree that the objection isn't a deal breaker. The pragmatic response would be to analyze the specific roads and allow the buses to use the shoulders in areas where the shoulder is wide and/or the emergency vehicles can drive past the bus in the grass. Or to create wider shoulders that won't take up as much room as a dedicated lane, but would still keep the current functionality of the shoulder largely intact.

"Allow the driver to start while someone is still at the front paying."

I've ridden buses in three cities in the US--DC, Pittsburgh, and NYC--and this already happens in all three. I doubt that they are all outliers.

Very bad for the elderly.

Pittsburgh bus rider here: They do pull away while people are paying, but it's not a hard and fast rule it the person still standing is elderly, or is managing a passel of children, or some such. They're allowed to do it, but they use reasonable discretion when doing so.

I love discretion. Unfortunately, some people are retards (to be fair, all of us at some times or about certain things), so the rest of us have to live in a world full of rules which constrain us unnecessarily 99% of the time.

Why do they even need to pay at the front? Why not use a pre-paid system like most of the developed world?

Most US bus systems have a pre-paid swipe or NFC card payment system with a cash backup system. The pre-paid transaction still takes time because some people need to fumble around in their wallets or purses to find the card. That's what the author is talking about; the author isn't simply talking about cash transactions, which are fairly rare because the US is in fact part of the developed world.

"some people need to fumble around in their wallets or purses to find the card."

That's one of my pet peeves whenever there's a line to pay, as on a bus or at a cashier. People should have their frigging bus pass or credit card out and ready when they get to the front of the line, instead of standing there like a zombie and only taking action when it's time to pay.

'Most US bus systems have a pre-paid swipe or NFC card payment system'

Which still is probably not the point. In Germany, and many other European countries, you simply by a month ticket that provides access to the entire system (bus, streetcar, train) for the number of zones you travel to normally travel (or you can buy a ticket with a number of zones representing unlimited use of the system).

This month card simply has a date - the driver simply checks the ticket as you enter - no swiping or handy involved. In part, because the ticket machines are often at the stops themselves, at least when the streetcar/train is involved with bus travel, as it generally is - the streetcar has its own 'lane' in a city, carries more people with more exits and entrances, and travels at least as fast as a car. Why anyone would want to use a bus in a typical city setting escapes me, to be honest.

But this much is true - car drivers who think they have more right to the street than a streetcar tend to learn quickly that the streetcar rules the city street, not some puny piece of metal that only weighs a couple of tons. Call it punishing those unable to follow the rules, who may or may not be considered sinners.

In my experience, most of European cities do not involve driver in this ticket hassle at all and rely on ticket inspectors instead.

London used to have buses with open access (no doors) at the back
The new versions have the same design but with more controls

London buses do pretty much all these things to some degree. There are special bus lanes, drivers will often speed off before everyone has finished touching their oyster card to the reader, some have entrances at the back without a door to close them and, frankly, the majority of London bus drivers seem to consider speed limits optional (indeed I remain convinced there's a prize for the first amongst them to succesfully powerslide a fully laden bus).

Being able to slide a bus successfully is actually part of the training for a London bus driver. Google "Chiswick skid pan" for more info...

It might be interesting to compare with London, where they've been actively trying to increase bus use with considerable success (just because all other modes of transport couldn't be expanded)

Reducing time at bus stops seems to be key. Key elements are banning buying tickets on the bus at all, having multiple doors (and accepting that people going on the back might well be fare dodgers), and using contactless payments so that the ticket validation that does happen is very fast. I suspect that hybrid buses have faster acceleration than the old ones, which also help.

Interestingly (and annoyingly) they've got rid of buses where you can hop on and hop off between stops.

Why on earth does a bus need to open its doors at a level crossing? What is that supposed to achieve?

I've heard that the original intention of stopping buses and opening the doors at crossings was to ensure that a train could be heard when visibility was too poor to be seen directly.

Cut railroad and highway labor costs to boost profits at railroads and cut taxes.

Ask instead "why do cars and trucks and buses need to give way to kilometer long trains?"

Input of labor easily eliminates all grade level crossings in areas with cars, trucks, buses, with the added benefit of creating many middle class jobs while the capital investment is made.

US grade level crossings are pretty non-standard in advanced economies.

"US grade level crossings are pretty non-standard in advanced economies."

What advanced economies don't have large numbers of grade level crossings for trains?

This section of Germany (call it the Rheintal, Mannheim-Basel) has spent a few years removing all grade level crossings (there might be a few remaining or in the process of being replaced further south) - but then, unlike the U.S., the stretch from Stuttgart to Strasbourg is where both ICE and TGV high speed trains run. It likely will be a long time before the U.S. has a single high speed train system working, much less two running on the same tracks.

And really, when the ICE gets up to something 240kmh (150mph) between Mannheim and the Frankfurt airport, the idea of any grade level crossings at all is funny, in a macabre way.

So, to clarify, Germany still has large numbers of grade level crossings for trains?

"Why on earth does a bus need to open its doors at a level crossing? What is that supposed to achieve?"

I've only seen this for school busses, where the signals on the bus are tied to the door controls.

School busses are required to stop before rail crossings to make sure they don't attempt to cross without sufficient clear space after the crossing to clear the tracks, due to actual incidents where a train tagged the tail end of a bus that was overhanging the tracks.

I've never seen a requirement to stop for transit buses.

Just space the bus stops further apart. If people are willing to walk a mile to the subway, they will be willing to walk another 100 meters to the bus stop.

Yes. In my part of DC, the stops are spaced about every 2 blocks along a major artery (Pennsylvania Avenue), which dramatically slows down the bus. The speed limit doesn't matter if the bus has to start slowing down for the next stop as soon as it gets up to speed.

That's because the bus is designed to feed the subway.

Sure, many vigorous people will walk the 2000 meters to the subway rather than mess with the bus, but the system is designed for people who cannot / don't want to walk.

During my decade in Chicago, an important realization for me was how to use the bus system (generally, buses on the major grid streets at every half mile). Many people, even so-called dedicated urbanites, just forget about the bus and concentrate on the El.

Especially helpful was when they added texts / apps / web pages that tell when the next bus would arrive. I could time my exit from my house / bar to expeditiously catch the bus rather than wait around for 1-45mins.

This is the primary problem I've noticed with mass transit. The problem is, one 80 year old who can't walk two extra blocks outweighs a hundred other people who arrive at their destination 45 seconds later than they otherwise would.

Would the world be better if 100 people had an extra 45 seconds of spare time, but one elderly person was unable to leave their house? Cost-benefit analysis needed.

The Coasian solution is to hire an Uber to bring her to the bus stop.

This is a perfect example of an issue that is unsolvable by economics, but which economists won't hesitate to confidently proclaim a "solution".

I will say the uber idea is cute

Proportionate to other risks, cars are far too dangerous (& in the US, they're far more dangerous than in Europe). Better to make them safer before making buses more dangerous. Suggestions don't make much reference to experience anywhere else, as other comments note. Promoting distracted driving by allowing driving while taking fare is crazy, and unnecessary; the answer is to sell tickets via machines before boarding. Higher speeds might cause lower bus use, because it will increase risk to pedestrians. The last suggestion is actually second least crazy; we already do this in London. It gives people freedom to take their own decision on risk, rather than impose it on others. I'd say second suggestion is craziest of all.

"cars are far too dangerous (& in the US, they’re far more dangerous than in Europe). Better to make them safer before making buses more dangerous."

From what do you conclude that? The original post from TC cites evidence that most people would gladly trade safety in order to get speed.

People aren't taking conscious decisions about risk/reward trade-off. They're following heuristics that accept massively higher risk from cars than from, say, aircraft or industrial chemicals. It's cars that are the real outliers. If you take that as your base line, you'd be willing to make more or less everything more dangerous.

Original post from AT not TC, btw.

If we took cars as the baseline, the TSA would be shut down tomorrow. There would be a few hijackings a year, maybe, but we would shrug our shoulders when the airforce shot them down, pointing to statistics on car accident deaths.

I would like the option of TSA-free airports, and see who attracts the greatest number of fliers (save time, money, hassle).

While I'd agree about the absurdity of our airport security theater, the tradeoff is not so simple. A more hijackable airplane is a risk not only to those aboard, but to anyone who lives or works anywhere near anything into which a highjacker might choose to crash his plane. Compensating them for this risk would be impractical at best.

A hijacking is very difficult these days, since the passengers no longer sit back and wait it out. And hijackers have responded appropriately to this new dynamic as well, by not trying any hijackings.

Or simply lock the cockpit door, like has become the new normal. We could roll back to 1990s security methods if we just locked the cockpit door securely.

Hey, you could even roll back to 1950s security and not to any screening at all. Someone might blow up the plane, but that's a risk taken by people on board the plane.

A plane flying into a building or crashing into city residences will merely kill people who drive cars unsafe lyrics or on roads with unsafe drivers, so they will be no more at risk from no TSA than they are getting around by car and not flying.

And it is a huge burden on flyers to pay the high taxes dedicated to funding TSA just to provide benefits to car users on the ground who are paying nothing for their increased safety that places massive burdens on flyers.

Fundamentally, the risks of driving and flying are different. When we consider the risk of driving we include in those statistics the results of people who drive recklessly. We weigh those risks against extremely unlikely airplane crashes. The fundamental difference here is that there is nothing an individual can do to make his trip on an airplane safer, whereas it is certainly within every individual's capability to make their trip in a car safer by orders of magnitude. While the irrationality of the average person is a fun topic to poke at, I believe this discussion speaks more to a misunderstanding of statistics by the people commenting.

You talk about what people can do to make their own car trip safer, but much of the risk is imposed on pedestrians and cyclists. There are some things you can do to make your walk or cycle safer, but risk is determined largely by design of roads and behaviour of drivers. That's not so different from the airplane example.

As other have pointed out, moving while people are still paying is common in many cities. Certainly, it's the norm in Pittsburgh.

I suspect it's against the rules, though. It certainly should be. You can't expect driver to be counting money and issuing tickets and concentrate on the road. Automatic ticket machines are the obvious answer. In fact, following European cities that already have much higher bus modal share would be better than making up ideas about what might work. Turns out increased risk isn't required.

Oh, no tickets are issued or change given. Routes and paypoints are structured in such a way that there is no ambiguity about the fare. If you’re going from the suburbs to downtown, you pay the full ride and, if you want to get off before the terminus, well, it was an expensive few blocks. Money is fed into a bill taker that does not give change. If you only have a 20 to pay the $2.50 fare, you’re out of luck. Drivers will sometimes let people on anyways or let them ask passengers near the front if they can make change, but they are not allowed to do so themselves.

Great ideas, but passengers say the problem is the reliability of public transport not the speed. http://gothamist.com/2015/04/23/stats_prove_that_select_bus_service.php

I agree heavily with this. I have a bus that could take me to work and home again with a total of 4-5 blocks of walking each way (something I do not mind even in the rain). That being said I cannot sit and wait FOREVER for the bus to show up.

While I support the idea of making the buses faster as well (our biggest issue in our town is not having an easy cashless system, so most riders pay with slow cash), they have got to be on time.

While I support the idea of making the buses faster as well (our biggest issue in our town is not having an easy cashless system, so most riders pay with slow cash), they have got to be on time.

Perchance might these two issues be related?

I wouldn't mind a bus ride that is longer than my driving commute, if it were reliable and I could concentrate on something else.

This is how the Google Bus is so effective for its passengers.

This was my hope when I started working downtown. My city has (near) realtime bus tracking and the total commute isn't much longer, if I time it right.

What I didn't account for was the lack of space. Often, I don't get a seat on my route and, if I do, I usually don't have the elbow room to easily work on my laptop.

If you're ever considering such a move, do a trial commute or two before making the trip.

So, maybe the biggest Bus improvement would be seats with airplane style drop down trays.

However, I expect a lot of bus systems have a more vandalistic passenger than air lines do. And furthermore, their ticket margins are far lower. So this would only work for specific examples aimed at middle class users.

When you try to design a bus system that middle-class people would use, you quickly realize that it involves keeping undesirables off the bus, which leads to gnashing of teeth.

What's the biggest complaint about the Google Bus from its community? Poor people can't use it.

Daniel - I don't imagine undesriables will be driven off of public transit by more comforable and spacious seating.

Technology can help. Some cities provide downloadable apps which show live bus times (tracked via GPS). If you know that your bus isn't coming for another 20 minutes, you can stay inside in the warm instead of waiting outside in the cold.

That's little comfort when you know you'll be late for a meeting or when you have to board a plane/train.

That's my situation as public transport user. All I gain if the bus makes 20 minutes instead of 30 minutes between home and the office is 10 minutes. However, I lose a lot if the bus is not punctual. The greater time optimization arises when you know you can take a bus that will arrive 5 minutes before your train connection leaves or 10 minutes before the job meeting starts. Optimization comes from planing, not fast buses.

Yeah, because highways never get congested by accidents or water line breaks creating sinkholes, so you will never be delayed driving to work for a meeting....

But Republicans don't want to pay higher taxes for any reason at all, including to fix roads so you don't get delayed by congestion caused by dozens of known problems that occur frequently.

"But Republicans don’t want to pay higher taxes for any reason at all, including to fix roads so you don’t get delayed by congestion caused by dozens of known problems that occur frequently. "

Yes, Republicans are responsible for the state every bus system across the entire country and probably the entire planet.

Ever considered moving to small city? No traffic, no pollution, friendly people. Highways are used only to go to another city, planes for longer distances If you get bored, travel to an interesting big city with "dozens of known problems" for 4-5 days and then go back to happy quiet life.

Republicans are like 6000 Km away, thus...no worries.

Isn't that among the other nice features of Bridj?

"Isn’t that among the other nice features of Bridj?" No, mate, that's one of the nice features of MBTA, the state-ru public transportation system that serves any inhabited street of Boston. I tried to see bus times for Bridje in Dorchester, Boston's largest borough, and it keeps saying "ETA when hell freezes over".

This whole exercise assumes that speed is the primary thing that keeps people off buses. I think if you poll people about why they don't use buses you will get a variety of responses, ranging from speed to reliability to one of the many variations on the stigma surrounding buses.

Jeff Kaufman doesn't seem to know much about buses, passengers, management, or transportation systems.

Consider also that most all local bus systems are government owned, designed, and managed. Thus all decisions relative those bus systems are ultimately political. Government management is notoriously inefficient and resistant to improvements.

"Government management is notoriously inefficient and resistant to improvements."

Private management is notoriously efficient but resistant to serving the unserved. For example, recognizing that the disabled already massively got the shit end of the stick, we throw them a bone from time to time, like rules which slow down transit for others but which make transit safer (or at all available) for the disabled/elderly. Ever seen a private sector coach with room for a wheelchair, or which lowers to ground level so elderly people with bad joints can get on without as much pain?

Mainly, I think it needs to be easier to discipline drivers for things that they would be disciplined for in the private sector.

Also, while the private sector tolerates risk well through insurance, the public sector may "over-insure" through safety rules because policy makers don't want any accident hanging over their head for the rest of their careers (for the rest of their career, their political opponents will introduce them as "Mr. X, the guy who presided over the Agency when that horrible disaster happened, and failed to respond in a way that would 100% ensure it would never happen again"). We should be more tolerant of public officials who preside during a time of a major accident, so long as flagrant violations are not evident (e.g., Flint water case).

The reality of mean politics suggests that the public sector is probably far more risk averse than optimal, which probably explains most of the "resistance to improvements" you mention. E.g., Option 1: Do nothing. If anything bad happens, just say it's because the old policy is bad, then tweak something minor and say you took action. Option 2: Reform. But if anything bad happens whatsoever, even if not particularly related to the reform, but definitely bigtime if in any remote way related to that reform, then you wear it as your crown of shit for the rest of your life.

Is the purpose of mass transit provide an efficient service for the population or is it to provide redistribution to groups the left favors. Let's have some honesty in our discussions. If mass transit is shit so that you can provide dedicated shuttles to the projects, you can't very well blame other groups for not signing up.

Most public transit systems would do better to just pay for an Uber for the disabled passengers.

Thomas - transit often is justified in terms of improving the quality of matching between employers and employees by lowering transportation costs for those who cannot afford them - also, it can be a good investment by taxpayers because it provides cheaper transportation for some and less congestion for those who continue to drive. Good public transportation can easliy increase the pool of candidates for a job from a few dozen to many hundreds, making it ten times more likely to find a better candidate for any given position. For centrally located businesses, it allows you not only to have a larger pool of candidates, but you can also hire cheaper because people who can't afford to live centrally and are willing to work for less money are then also able to get to the workplace. Also, I think most people are pretty OK with ensuring that old people, poor mothers with young children, etc. can make it to doctor's appointments, grocery stores, play dates and the like.

Indeed, many left wing policy preferences are not so much geared around handouts to losers for the pure sake of redistirbution, and primarily serve to improve the economic potential of the disadvantaged - the purely redistributive aspect is a clear plus for many supporters, however.

I'm not familiar with such ultra low grade public transportation as you refer to ...

Daniel - quite likely you're right. Applying similar logic, in Burlington Ontario they decided to cut non-peak service for a lot of routes, and instead if you call the transit authorities they send you a taxi that costs the same as bus fare - if you're going far, it delivers you to the normal point of transfer and the taxi driver issues a transfer that you can use on the bus. On the specific matter of Uber for the disabled, the transit system that I'm 2nd most familar with (Hamilton) would be a definite exception, because it is a major centre for health services specializing in many kinds of disability, leading a lot of people to move there to locate themselves nearer to the services.

Given Uber intentionally designs it's services to exclude the poor and disabled, are you arguing government should dictate the way Uber operates?

It's not the speed of the vehicle. It's the speed of getting where you want to go. I take a bus to the airport because it's "faster," but only because it doesn't have to park or pay tolls. I don't take the bus anywhere else.

In the exact same city, I know some routes that are routinely 10 minutes early and 10 minutes late (frustratingly, often the same routes which run 1-2 times an hour, so for many this means you have to plan to get to work 30-40 minutes early to be sure that you're not late), and others where drivers 100% of the time will idle for so little as 20 seconds in ensure that they don't get ahead of schedule.

The actual problem seems to be advocacy (well, actually it's the bus drivers' fault), where low income people don't realize you can pick up a phone and raise a little hell for problems, whereas other people who are accustomed to higher standards will raise hell for the least inconvenience in customer service.

"In the exact same city, I know some routes that are routinely 10 minutes early and 10 minutes late (frustratingly, often the same routes which run 1-2 times an hour, so for many this means you have to plan to get to work 30-40 minutes early to be sure that you’re not late) "

Apparently this is not a Japanese city...

Hamilton Ontario. Haven't lived there for a while.

"where low income people don’t realize you can pick up a phone and raise a little hell for problems"

Where do you live? Around here low income people are very accustomed to raising hell for any perceived problem or slight.
Trouble is that they, as well as well off people, their complaints are ignored by whatever region transit authority there is.

Cut the number of city bus stops significantly. They are often just two blocks from each other.

I assume you don't live somewhere where winter is harsh or rainstorms common. A good middle ground is to run express buses on the same route which only pick up every 6-8 stops, something that is virtually ubiquitous in most high-traffic routes in Canada.

Optimizing the wrong case.

I think busses as a paradigm (large vehicle, professional driver who isn't drinking or in a hurry) are inherently safe. Given that, the trade offs are on a different curve and not generally worthwhile. Except the railroad tracks one. That's dumb.

A professional driver caused the track damage when crossing tracks that caused the Amtrak derailment which the train driver saw, but too late to stop.

But you are correct, there should be no grade level train crossings. Fixing that problem would create perhaps a million jobs for a decade.

But I bet you would object to creating a million jobs to eliminate grade level crossings to eliminate buses, mostly school buses, stopping at rail crossings. I'm sure that 90-95% of buses stopping and opening doors at grade crossings is done by school buses which collect no fares.

In terms of reducing the number of stops, how much of the time spent at stops is taken with slowing down, speeding up, opening doors etc and how much is spent letting people on and off? Reducing the number of stops will only reduce the first set of delays (except insofar as you drive people off the buses by making the bus stops too far away). To reduce the delays caused by people getting on and off, you need to run more buses so that fewer people get on and off each one.

And let them skip every other stop so each bus only makes half the stops.

Buses skip most stops, unless in a dense area when passengers need to get past other passengers to get off and on the bus.

Fewer stops in dense passenger areas would only mean more time would be required at stops to allow more people to struggle to get off past people struggling to get on. In sparse areas, stops are already spaced out, except where an apartment complex is a block from a college. Most of the time, only one of the stops will be busy because the passengers are on different schedules.

Here are some practical approaches that are not necessarily widely implemented:
1) real-time bus location apps for passengers, so they can time their arrival at a bus stop. for example, see transit.511.org for the San Francisco Bay area.
2) transit-priority traffic signals, so the traffic light turns green for the transit bus.
3) Bus Rapid Transit (aka BRT) for high volume routes, where buses get their own fixed infrastructure, especially dedicated lanes and pull-outs, to improve performance. Usually much more cost-effective than light rail or street cars.

In London we have real-time bus location maps:


I saw something on youtube about bus transit in South America. (Sao Paolo maybe.) Bus stops are elevated platforms in the middle of the road and allow people to step directly on and off without the need to go up or down steps to access the bus or bus stop. These stops are covered and climate controlled. The location in the center of the road permits easier, faster bus connections from two directions of traffic. The location also appears not to impede traffic flow too much as pedestrians access the platforms via underground walkways.

The concept favors large cities. At least elevate bus stops so time is not wasted while people go up and down steps, especially those that may have difficulty (e.g. strollers, wheel chairs, walkers, crutches, etc.).

You might be thinking about Curitiba's Bus Rapid Transit system, which is well known for pioneering many of these techniques. The goal was to increase ridership by making bus transit as painless as possible. Curitiba has served as a model for the bus rapid transit (BRT) systems of other cities.

Here's wiki: https://en.wikipedia.org/wiki/Rede_Integrada_de_Transporte

Free buses. That will get a lot more people riding them. Or mostly free buses with congestion pricing. People pay to use buses, but only during busy periods to the extent necessary to prevent buses becoming overcrowded which would increase wait times and put people off using them.

Sure free buses would still cost money to run, but building roads to handle private transport costs money, as do accidents, and pollution from private vehicles. Electric or natural gas buses can eliminate a significant amount of pollution exposure in cities.

But since that suggestion is not particularly dangerous, how about a running board at the back people can sit or stand on so they can hop on and off as they wish. People who are late to catch a bus can sprint and grab on the back. Adds new meaning to the phrase, catch a bus.

Homeless people will use free buses as shelter. This will put off other passengers.

Clearly a guaranteed minimum income is required to get buses filled to capacity.

In San Francisco, the homeless hang out inside subway stations and sheltered bus stops. It can be very upsetting to tourists and commuters.

A better solution is to periodically cull the homeless population but the mayor's office refuses to respond to my letters.

Cooper, instead of ending your letters with "God bless" try writing "Allahu Akbar" instead. I'm sure that will result in you getting some kind of response out of the mayor's office.

More likely, it would land him on a security blacklist.

Free buses, and charge all other (motorized) vehicles to enter central city districts.

They already pay for the streets, why would you charge them?

Everyone pays for the streets, whether or not you use them. Charging them to enter central districts reduces congestion, and then the richest just pay anyways and everyone else gets better public transit. It's seem win-win for most if the population to me, but clearly not many voters like this - the idea has been proposed in all sorts of cities and kiboshed in many places where variants of it were introduced.

Turn cars into buses.

In Cuba you have to pick up a hitchhiker. If you had the same rule for cars elsewhere driving would now have many of the disadvantages of public transport. Making buses relatively better

Yes, following the lead of a brutal dictatorship and one of the poorest countries in the world [depending on Venezuela for oil!] for is just what we need.

Any tips from South Sudan?

Yes introduce guinea worm. Its a serious disease was the no treatment. Once people have it they won't spend as much on non serious diseases that have treatments. Thus I f we copy south sudan well save medical costs.

People rode the trolley and it was really dangerous (I remember my first year torts case book having several gruesome trolley cases), so Tabarrok's point is a valid one even if, or because, it is counter-intuitive. A bus is a trolley designed by committee. A better designed bus (i.e., a trolley) would have way more riders even if it's less safe. Of course, with urban sprawl, bus service, any transit service, is difficult. Higher density would be the first step toward greater use of transit service; but better transit service is necessary to attract more residents to a higher density. You see the dilemma. Anyway, I'd limit trolley service, I mean bus service, to high density areas, give the "bus" the right of way ("get out of the way, here comes the bus"), and make it easy and relatively cheap to use (such as a deeply discounted annual pass). In my new city, the bus is the only transit and it's both inefficient and unappealing. Ho hum. When I first moved to the place almost 40 years ago, I rode the bus. My superiors would glare at me as I left the office at 6:30 to catch the last bus. I'm not sure if I rode the bus because I wanted to do my part as a good citizen or because the last bus left at 6:30. Probably the latter. A memory forever etched in my mind is that of the middle aged fellow who daily would dash out of the Ohio Bar (why Ohio I don't know since it was far from Ohio) to catch the bus just as it was leaving, his cocktail from the bar in hand.

"A bus is a trolley designed by committee." Precisely: a government committee. Drivers are public employees with zero incentive for customer service. I say liberalize the rules so that private companies can better meet the demand for transit and respond to the market, rather than dumbing it down via central planning.

Why is there always an assumption about government employees that they have no incentive to provide good service? We can argue about the stringency of rules protecting public sector workers and where the line between protecting workers from frivolous complaints and being able to quickly and decisively discipline those who are not doing their jobs is. (Which vary so much across the world’s thousands of public transportation systems that it is impossible to generalize.)
Still, drivers sit in a very identifiable place and are surrounded by placards explaining how patrons can report poor service or misconduct on their part. Cameras and sensors on the busses provide reams of metrics against which they can be measured. Depending on the municipality, of course, drivers can be and are incentivized to provide good service. Certainly my experiences with them have been generally positive.

I think it mostly comes from people who can't imagine that there exists such a creature as a human being who actually cares about the general public. This leads them to believe that every person working in the public sector is a lazy and utterly self interested sponge who would never show up for work again if they could do so and still get paid.

The above argument relates to potential inefficiencies of non-market planning, however, which is a rather different story.

Busses being 'designed by committee' isn't really true. They are designed by private companies that then bid on contracts put out by government agencies.

Your point is relevant, and public sector procurement can absolutely be very good with respecting all sorts of free market stuff.

But the specs for the bid are indeed determined by committee. The private sector bids on fulfilling those requirements (and in a sense technically does the design), but do not generally have much influence on the specs. This means a reality of "feature A at all costs" in some cases because the costs are not fully internalized into the decision process. The bugs can be heavily overplayed by anti-government types, but the bugs are definitely there.

I know someone who worked in public sector procurement in health. The "doctor decides" principle of most technologies means that there is not a very strong market signal for costing. Yes, each and every procurement is 100% competitive, but the specs of the tender for bids are not often driven by cost considerations.

Short: specs come first, competitive bidding second. There is very little cost management input into the first part of the process.

Car and Driver looked through a DOT report and compiled the fatalities per mile of different transportation modes. Passenger cars came in at 1.1 death per 100 million miles and transit buses at 3.7. That's vehicle miles, not passenger miles, but it points to a problem with transit system accidents: each one involves many more people than a car accident, and every one becomes a local news item. In other words, bus passengers are more likely than car passengers to be in a vehicle involved in a fatality.


Not long after I moved to my new city, a ship hit and knocked out the center span of the enormous bridge at the entrance to the Bay just after dawn during a driving rain storm and a Greyhound Bus filled with passengers drove into the abyss and plunged into the Bay killing all the passengers. The lawyer who represented the harbor pilots (he and I were in the same firm) heard the report on his radio, immediately chartered a helicopter to take him to the ship, and advised his client (not to say anything to anyone). At the hearing, he asserted as a defense that the incident was an Act of God (a sudden, unexpected rain storm) and, hence, his client (the harbor pilot) couldn't be held responsible. It worked. I suppose riding a Greyhound Bus might be considered assumption of the risk (a complete defense when I was in law school but not under the current prevailing standard called "comparative negligence" - it's sort of like comparative advantage, for you economists - "compared to what", I might ask).

Wow, that was really stretching the law. And BTW bus drivers and bus companies (common carriers) are held to a much higher standard than others, so no "assumption of risk" nor "comparative negligence", it's almost strict liability.

It's not weighted for how many people are on the bus. You acknowledge this but then proceed to ignore it in arriving at your final conclusion.

It's plain and simple according to those stats: busses are more dangerous per mile than personal vehicles. Are you a math denier?

It's a misleading statistic. If there are 20 people on each bus and 2 people in each car, ten times as many people are transported in the same number of vehicle miles, implying three times safer.

Good application of basic math requires good reading comprehension and the abilitiy to focus on the number that matters.

If driverless tractor trailers had 10 times the vehicle accident per mile rates of cars, would you argue driverless cargo trucks are infinitely safer than cars because no people on the truck are killed?

Mulp - Yes, I would. Safety applies to humans, not machines. But, who's on the other side of that accident? If it's humans, then it matters, if it's machines, then that's "just money".

I was not concluding anything about how many passengers die. If buses are involved in fatalities three times as often as cars per mile, then bus passengers will be involved in fatalities three times as often as car passengers per mile, and there will be more of them involved since buses carry more passengers. Involved. Not killed, just involved.

Now, you may think, why should I care if a bus I'm in kills someone who isn't me? It would be unpleasant for most to witness another's fatality, and at the least, I won't be able to continue my trip on that bus.

Yeah, I thought those numbers were fishy. It's hard to reconcile the OP with stats about light-rail deaths being so high.

To Nathan W's point, though, buses mostly run empty or nearly empty. It's only certain times of day and certain routes where the passenger numbers are much higher than for passenger cars.

Buses should have no doors, leave the space open.

Buses should not stop at bus stops, it should slow down but let people jump on and off at bus "stops". This is the way it goes for trains in third world. Do the same for buses.

Key: Throw people without tickets onto traffic.

This is easy! Limit the speed of cars to no faster than the average city bus (maybe 35 mph?). People will flock to buses.

On the other hand, Buses will become dramatically "more dangerous", since cars will become dramatically less dangerous given that they're traveling slower.


In North York (Toronto), express routes only accept tap pay or you can buy the ticket from a machine at the bus stop and flash the receipt when you get on the bus. This saves a bit of time at stops without compromising safety. However, a few times while I was in the area I missed the bus because I was tied up selecting route options, fiddling with change and waiting for the receipt to print, while the bus sped away - I would actually rather be able to pay cash on the bus, especially since using the machine costs lots (20-30s) of MY time in order to save time for those who use cashless payment. The people who pay cashless are presumably similarly indifferent about my time, and are greater in number, so surely things will not change.

In much of the developing world, driving and fare collection are separate, allowing the bus driver to focus on speed and the assistant to focus on fare collection and ensuring there is no fare evasion. After busy stops, the fare collector is often literally still collecting fares even once the bus arrives at the following stop. More practical at $2-5 a day salary than $200 a day salary.

In much of the developing world, buses don't even stop at bus stops, they are still moving at a couple km/h at the stop and you literally have to jump through a moving target to get on the bus. Much faster, but I don't see it improving ridership or safety.

My city (Pittsburgh) is looking at reducing fares for card users on busses going from the outer zones into Downtown ($3.75 vs $2.50). This will have a secondary benefit of standardizing fares and reducing the importance of structuring routes to segregate those who are only going within the Downtown zone and those who are going to the suburbs.
I think that this is the best approach. It incentivizes using the quicker, less error and argument-prone cards, but doesn’t shut out either casual riders or those who forget their cards.

Adding another government employee per bus at 50k per year plus benefits plus pension to connect fares, is about the worst idea yet.

I suggested adding another employee at $600 per year and against doing so at $60,000 a year, because the second would indeed be dumb. Reading comprehension, meet math.

Increasing gdp by $50,000 by paying wages is really terrible. Better to increase gdp by only $20,000 by providing $15,000 in consumption spending with $5000 of government welfare labor costs!

Paying people doesn't increase GDP unless they're producing something. Unless ... we're talking about a major AD recession and redistribution to high marginal spending poor targets AD.

Busses in my city have fold-down bike racks on the front. These must be manually operated by the cyclist. This can take several seconds, especially on cold days when the mechanism tends to freeze. After removing their bikes some cyclists either forget or ‘forget’ to fold the rack back up. This requires several more seconds of hornblowing and confusion on the part of the driver and cyclist.
A powered version, which the driver could lower as he approaches a stop where he sees someone waiting with a bike and raise after taking off would save considerable time and confusion.
Signage asking cyclists to please be aware and queue up before their stop so that they are at the front of the line exiting the bus would be low-hanging fruit. Might not help that much, but it would be cheap and simple to implement. If it got one person per day to the front of the line at a busy stop, rather than at the tail end because they had tuned out while looking through Facebook, it would be worth it.

The essential problem with buses is that they can seldom deliver a single-seat-to-destination solution.

A typical bus trip is a walk to the bus stop, wait, transfer to another bus (or a train) after another wait, then transfer to a third bus and walk to destination. A three-bus trip will get one from just about anywhere to anywhere that lies within an urban transit system, but multiple vehicles extend door-to-door travel time far beyond what would be required if one traveled by car or taxi.

A secondary problem is the difficulty/inconvenience of making intermediate stops and then transporting parcels on the bus, especially when transfers are required

Perhaps some form of small robotic vehicles will eventually eliminate the multi-seat problem, but until that happens public-transit promoters could at least recognize the problem.

For an example of how bad it can get in the other direction, Quebec City is a good example. Because winter is very long and very cold, accompanied by massive amounts of snow, it makes sense in a way. But it's literally faster to walk quite a few of the routes than to take the bus (the university to party district bus is an important exception), because they are winding up and down all over the place on some very long routes Several routes have massive overlap and are similarly winding, with some minor differences in point of origin and final destination, and would be much much faster if they covered more different territory and included a couple transfers.

Short: One-seat objectives can be taken too far.

What do long, winding, inefficient, "access based" routes have to do with A to B single-seat objectives?

There is overlap between the winding routes to achieve single-seat objectives. Average travel time would be lower on some routes if they routed differently and included a well-timed transfer. A to B single seat objectives can be taken too far and introduce inefficiency.

None of these suggestions seem to account for the delays that will be caused by accidents, whether someone slipping as they exit a moving bus, or the bus itself being involved in a collision.

Interesting. Even ignoring safety, would you rather a commute that's 2 minutes longer a day or one which is an hour late once a year?

Probably the latter, though I have no idea if that's actually the tradeoff.

They should privatize buses, like they do here in the Philippines (buses are called "jeepneys" here, Google it). And Jeepneys are unsafe at any speed, but popular, as AlexT might understand.

As may have been mentioned (skimmed) changes that make the buses more dangerous to others (drivers, pedestrians) do not actually improve ROI. Lawsuits are expensive.

Just a nod to the fact that in Chicago, some buses are already allowed to drive on the shoulder under certain circumstances:

Allow the driver to start while someone is still at the front paying

Reminds me of Medellin. The drivers there hit the gas while simultaneously digging in their pocket to give you change, and oftentimes before everyone has even gotten on the bus. No worries, people just yell "stop!" and he waits a few second longer for everyone to get on.

Is that to stop the sicario from assassinating you?

* Don't have the Obama Administration shutdown cheap Chinatown buses over dubiously sourced safety claims?

Yes, but those were private buses run for profit. So, clearly there was a need to shut it down for 18 months due to safety concerns. /sarcasm

"Fung Wah is now defunct. In March 2013, the Federal Motor Carrier Safety Administration (FMCSA) forced the company to cease operations, claiming it was an “imminent hazard to public safety. ...

On May 31, 2012, the Federal Department of Transportation shutdown 26 bus companies in a single day, and since then it has forced an additional 15 closures. Many of those companies were owned by Chinese immigrants. The American Bus Association, a trade association that primarily represents the large corporate carriers, has cheered the government on.
Contrary to the claim that the company became more dangerous in its last few years, in the three-year period from 2010-2012 the company's buses didn't have a single accident. In its 12 years operating as an official bus service, the company wasn’t involved in any fatal crashes—with the exception of a 2008 incident in which an out-of-control dump truck struck a parked Fung Wah bus, triggering a chain of events that led a pedestrian to have a fatal heart attack. It’s hard to imagine a scenario in which a bus company could be less at fault.


I love the aspie economists at MR. Only here can you see an extensive discussion on this topic with hardly a mention about why most people don't like public transport in the US: it's potentially full of losers, low-lifes and scum bags that nobody wants to sit next to.

Curitiba, Brazil has long ago solved the problem of how to make buses faster, basically by turning them into above ground subways. Their approach:

1. Reserved streets just for busses.
2. Most of the side of the bus opens, so passengers don't have to board one at a time
2. Special enclosed bus stops where passengers pay, so the bus is not held up waiting for passengers to pay.

Curitiba sounds like a nice place. It's about 1000 meters in elevation in Southern Brazil so the weather is mild, the population is highly European, and the long time mayor, a brilliant Jewish guy, was a real civic improvement dynamo.

Have Google engineers come up with a driverless car/uber type app . One enters on their smartphone where they are, where they want to go, and how much walking they are willing to do. The overlord "cloud" tells you what bus stop to go to, when you should get off, where you should walk to to get on another bus and so on. The algorithm takes into account traffic patterns, delays, passenger loads and automatically pays fares. Every commute might be entirely different depending on daily conditions. It would tell you how long to wait, as well as tell the bus drivers how fast to drive. Treat commuters as industry already treats production controls.

CityMapper basically does this in a few major cities. It works great in London- tied into tube delays, live bus arrival times, etc. It even gives you the cost so you can make the trade-off for yourself. Bus vs. tube vs. Uber - each with an estimated time and total cost shown.

That was fantastic, your best post yet Alex.

Might I suggest further improvements

Passenger load/unload points should be prepurchased by using an app or web page.

Busses should be given roght away at all times.

Buses should nothave stops at all, passengers should embark through a system based on the already mature mail hook/cather pouch trchnology developed for railroads.

Disembarkation should involve specially configured catapults with receiving nets at passengers stop.

All accidents involving buses should just be ignored until designated periods during which busses be taken out of service for maintenance, as it is not just buses stopping at accident scenes that consume time, but also the presence of emergency vehicles and clean up that delay matters.

Buses need to be run more frequently to reduce wait time. They need to be smaller and cheaper to operate. Having unionized drivers makes costs too high for small buses. If all bus drivers were paid like school bus drivers many more buses could be run.

This post was clearly written by someone who doesn't ride the bus

thread winner

Denver-Boulder does use Shoulder Lane for busses


In my close-in-city neighborhood, the biggest reason people don't take the bus is
1. Charter schools.. Very few families send their kids to the neighborhood schools, so you have to drive them to school.. thus you need car.
2. weak communication - the real-time app that was just released is bad, but is superior to nothing
3. subsidized parking

Require busses be self supporting and the passengers pay full fares rather than the tax payer subsidizing them.

Minnesota already allows busses to use the shoulder during traffic jams.

Plant bombs on them that go off if Sandra Bullock drives under 50 mph?

Allow riders to opt for a cruising-speed debusment in exchange for a 75% reduced fare.

This thread is the best example of full blown Ivory Tower stupidity from the intelligent I have ever seen. What is wrong all of you? You are so full of stats and numbers and calculations you can't see what's in front if you, unless it is simply being avoided because you can't answer it.

The problem is this: My daughter, 18, has to occasionally ride the bus to work, a distance of three miles on one line, in suburban Mesa, Arizona, a place with low crime. In the few months she's been doing this she has regularly been stared at and even followed by prison tatted crazy-eyed Manson-like scumbags, who are WAY OVER-REPRESENTED in the population of bus riders.

So how about the braver ones here, if any, go off reservation and actually address this simple fact? And I dare you to do this without the backup of studies and stats to show exactly how many young women feel threatened or at least uncomfortable riding them. Because every one of us knows this is a problem. Shoulders, speed, yada, has zero to do with it.

Yes, that's a good point. None of these issues deal with the safety aspects of taking the bus. Particularly for women at night.

A few Canadian cities allow women to ask to be let off at any location betwen normal stops after 8pm. Were any man to jump off behind them, surely people would not be retarded and take an extra minute to pay attention to make sure that nothing abnormal was going on.

The suggest that we should answer "exactly how many young women feel threatened" without resorting to statistics doesn't make any sense to me. Like "if you can't provide accurate statistics without using statistics, then I pre-call you full of baloney." Doesn't make any sense.

Here's a specific solution I saw recently for women's safety. It's not 100% safety guarantee, but it's a live feed app through a mobile stream that women can have ready if they feel they may be threatened, guaranteeing that the perptrator would be caught on video (even if they steal the phone after). Sorry, don't remember the name of it, but if this is a concern for your daughter, it might help her feel more secure (but I'm concerned that it might also encourage her to live in more paranoia than is maybe good ... hey, I'm not her and I'm not on that bus. If a women says the men feel kind of rapey, I'd take their word at face value even if there are 1000 false calls.) Probably you'll find this idea offensive, but allowing a local hooker service might promote women's safety if properly regulated.

Aside - it amazes me how much skin is shown in East Asia, but the men here seem very able to deal with it respectfully, presumably making women feel very comfortable with showing even more skin. Seems win-win to me. Don't ogle them too much and they'll feel comfortable revealing their beauty.

I think there is research out there- probably pretty old research, because I think Christopher Alexander might have mentioned it in Pattern Language back in the 70s.

If I remember correctly, the research pointed to smaller buses, possibly van sized. More of them, faster, presumably more routes run per day too. Might have been something about incentives for drivers too, but can't remember. It was supposedly more efficient than the larger more 'modern' bus systems.

Of course, nothing was said about danger, as far as I can remember.

A post on why people don't ride buses with zero mention of low status signalling and/or safety of the rider from other low status passengers and the general unpleasantness of being on the bus among that crowd? My impression (I could be wrong) is that double decker buses in London are relatively nice city buses and ridden by higher status people than typical city's riders therefore they are more popular. I doubt they're any more or less on time or speedy.

. Put wi-fi on all buses and at bus stops so the ride can be productive, useful. Have apps that let people be reminded of their stops - changes, so when phone engrossed, they don't miss their stop. Also have charging stations. Thus integrate transportation and communication. Human logistics. Phones should also be able to pay fares or renew weekly passes. What people want most is a consistent, reliable trip time. Problem with SOV on un-freeways the one hour regular commute can vary too much. Reliable time is much more useful than speed here and there.

2. People need a door to door solution - from home to work. In the ultra low-density suburban car-city environment (looks densely built but very few people per dwelling unit - thus per block/acre) it is hard to find someone with a close match. Odds are the cubicle mate will not have the same route. To get around, it is necessary to chain a number of shuttles, with walking at both ends. There are plenty empty seats on the road - question is how to fill them without too much diversion for a car-pool option or to get to the bus scheduled route. Seamless pay and/or credits for providing rides should be able to be packaged. The ride may vary some from day to day. The passenger/traveler as packet.

So not enough people are using subsidized transportation because it's too safe. Wow. Time to order more pizza, bro. No doubt it's easy to make a huge thing travelling on the ground on roads with many people on it that stops all the time faster than small things that share the same road that don't stop at all. What could be hard about that? The size makes it safe. The subsidy makes it cheap. Both make it slow and unpopular except for among those whose time has little value, and don't mind spending time with other whose time also has little value. A fast bus is called...a car.

Back in the 1970s, the WSJ Editorial Page was always praising the Pinochet regime in Chile for privatizing the city buses. On our trip to South America in 1978, my father and I rode a bus in Vina Del Mar. As the last passenger was climbing on to the bus, the profit-incentivized driver floored it, the would-be passenger fell off and hit his head on the curb and was knocked out cold, maybe dead. (He didn't come to in the 10 minutes we waited for the ambulance to show up.)

My father said, Yup, that's what happens when you privatize city buses. I tried to remember a WSJ editorial explaining why he was wrong, but I held my tongue instead.

Dangerous, hmmm. How about a first class section on the bus? Some will pay more for bigger seats and more legroom, or simply to avoid lower class passengers. Great way to get the rich to subsidize transit for the poor, and it already seems to work on airlines which are private businesses. And trains.

It may seem at first like buses in the emergency lane is a bad idea, but they're equipped with radios/GPS and could be tuned in to first responder channels to get out of the way, unlike regular commuters who don't even respond to sirens and lights until they see them.

Certainly a broken down car in the emergency lane will block the bus, but it's already blocking a potential emergency vehicle, though less frequently.

Because the buses are not stuck in traffic, they can scan many more miles per minute of traffic and report back, although cameras probably do all that today.

Lastly, when a car is in a wreck there are usually just a couple witnesses. A loaded commuter bus in a minor crash means more than 40 witnesses, stuck there until cops interview them all. No so much increasing danger to passenger but driving up cost of bus ridership.

Motorcycles take huge risks of serious bodily harm but get correspondingly large rewards of time by riding between lanes, legally in many states.

In Los Angeles:

"By the end of October, there had been 1114 serious crimes on MTA buses and trains, the agency said, up from 963 during the same period in 2011. There were 291 robberies, up from 205 last year, and 185 assaults, the highest in at least five years and 22 more than during the same period in 2011.

There were at least three rapes during the period, including one at Union Station. Those numbers do not include a woman who was assaulted at a bus stop earlier this year, because the agency only counts crimes that happen on board a transit vehicle or at one of its stations."

Don't let women who stand waiting for a bus and who then first open their purse,fumble around inside for their change purse,and then fumble around inside that for exact change only after they have gotten the bus ,EVER RIDE BUSES .

In Washington DC the problem is that the government exhibits lexicographic preferences for parking over buses. On my most common bus route (the 42) buses take ten seconds to navigate one turn that could be made much simpler if they would simply prohibit parking close to that corner. To save a handful of drivers 10 seconds (there is a parking garage literally 65 feet away) they cost 10,000 bus riders 10 seconds of their time. The fact that drivers unfamiliar with this intersection will illegally park right behind the last legally parked car and block bus traffic means that a couple of buses will be stuck for five or ten minutes each day until the driver returns from their coffee run.

Two blocks north of this intersection are two more intersections, twenty feet from one another because of the state avenue slicing there. The bus stop is before the first intersection and on the 20 feet of road between lights the city has two parking spots. If they weren't there the bus could easily go from the bus stop into traffic, but without that buses must wait for traffic to pause, which invariably necessitates waiting for a red light, adding a minute to the route.

There's at least one other spot in this route where a few seconds could be shaved off the bus time by eliminating ONE parked car, but free parking is sacred in this city, and charging a marked clearing price for it is inconceivable. So instead we take 10,000 minutes of commuters' times to save drivers 10 minutes.

Very concrete and actionable. Do you know where to send letters? Maybe a local paper would have an interest in publishing such a letter ... in my experience local papers tare very open to publishing concrete and easily actionable suggestions for local roads, traffic, etc., which isn't as easy to ignore as the letter to the people who should pay attention to these things in the first place.

On the flip side of the safety issue: over 20 years ago there was a debate about requiring school buses to be equipped with seat belts. Which brought up the question of how safe or dangerous is a school bus? The answer, at least back then: 10 fatalities per year on school buses. Which pretty much should have provided the answer then and there: with so few fatalities, there's no way that a costly universal requirement would have a positive benefit-cost differential. Especially given that seat belts would only save a fraction of those 10 students per year, and there's no way that a bus driver can make sure that the students are actually wearing their seat belts. We can't reduce fatalities below zero, and 10 is getting pretty close to zero.

I didn't keep up with the subsequent legislation and I don't ride school buses but I think I read that some states did decide to make seat belts mandatory on school buses.

The primary factor that slows my bus commute (entirely urban) is stopping. Moving stops from every quarter mile to ever half mile would probably be the biggest change that could speed things up, but that comes with reduced access for the mobility challenged, and, selfishly, makes it harder for me to walk ahead of the bus instead of waiting at the stop.

Raising the speed limit for the bus doesn't seem like it would do anything given that it has to stop all the time anyway, or would be blocked by other traffic.

I can't imagine that rail crossing make much difference system-wide, as the frequency of at-grade crossings where there's sufficient density to make transit viable can't be that high.

I think comfort's at least as important.

After the Great (Washington, DC) Metro Crash of '09, trains were required to manually come to a halt at each station, instead of stopping automatically (and much more smoothly). The resulting jerks (no pun intended) made some riders queasy -- enough that the Washington Post actually officially noticed.

How many of the queasy riders decided "Forget this, I'm driving from now on," and how many of those got (themselves and others) into accidents?

A simple truth about transportation is that you can either have speed or access, but not both. High speed of an interstate is enabled by limited access. Thus design speeds of 65+ are possible when the traffic volume is below design and weather permits. The more entry points and presence of a complex environment - the slower the safe speed. With trains, elevators in the World Trade Center, buses, there is the express run with few stops, or the milk run with many stops. A rich, complex environment - a city - will be slow. Taking advantage of the vertical - with the use of the elevator, or simply walk-ups - which gives health supporting exercise at no cost but a little time - is what makes cities efficient.

Comments for this post are closed