Hacker News Reader

"a massive impact in terms of social equalization."

The folks who don't like that, can prevent the equalization by careful scheduling and have already done so. I live in a nice civilized town/city and the bus routes travel within one block of every square inch of the city hourly. So average transit time is one hour, max theoretical (assuming not missing the bus, etc) is a bit less than three hours. Unfortunately I can drive in my car between any two points in 10 to maybe 15 minutes worst case.

I can, and have, repeatedly out walked the city bus by walking direct pt-pt rather than meandering all over the city. Its really more of a bad weather service, or a service for people who can't walk.

Also I live in a civilized town/city but work in the neighboring uncivilized dump of a city, which has slightly better bus service, but schedules have been "optimized" to keep us separate such that my commute via bus would seem to only take about two hours each way, but careful optimization of schedules increases that to at least three hours with lots of standing around at the transfer point. Of course in a car its about 20 minutes, maybe 40 at rush hour.

And that's how you keep social groups separated even with low subsidized (or even free!) bus fares.

It sounds like your city would be a great candidate for decent bike lanes.

That sounds quaintly conspiratorial, however it is incredibly complex to optimize bus routings, especially given financial limits most systems operate under. Any change will always be at the detriment of someone else, and the last mile(s) problem is enormous.

Logistics is hard in general, is my impression. In fact, most of the interesting problems logistics seem to reduce to NP-complete problems, which has some unfortunate consequences for finding optimal solutions.

The generalized form of most logistics problems reduce to NP-complete, but we are also almost always working in a system with much more structure than the general problem, so a polynomial time solution may be available. Also, many NP-complete problems can find 'near' optimal solutions in polynomial time.

For example, consider the traveling salesman problem in a metric space. We can construct a minimum spanning tree in polynomial time, and we know that traversing this tree takes no more than twice as long as traversing an optimal solution to the traveling salesman problem. By using the triangle inequality, we can further optimize our solution, still in polynomial time.

There is also a bit of a self-reinforcing system in play. Good access to mass transit makes an area more desirable and more expensive, so rich people end up living there. Bad access to mass transit makes an area less desirable and less expensive, so poor people end up living there. But then there is less incentive to expand service into those areas, because it's just a bunch of poor people...

Sounds like London. I live in London and these days I just cycle everywhere, but I´m sure that walking wouldn't be much slower than the bus.

London, UK? Which happens to have one of the best and largest underground system in the world covering almost all areas of the inner city. Then you have the train service to get you out to the outer parts, and then the buses?

Admittedly, once you get to rural parts of the UK the public transport sucks, but day or night, I can get myself around London without too much of a wait.

Even just purely by bus, London is night and day different from any American city. You rarely have to wait more than 10 minutes for a bus, and often they come just a few minutes apart.

Now, like the GP, I also cycle around London, and it is by far the best way to get through central London during rush hour, but bus drivers here are aggressive and they tend to be quite a bit more efficient than their sparsely populated American counter-parts who are brow-beaten into submission by aggressive American sedan drivers who've never heard of public transportation or sharing the road.

Some routes are very sub-optimal though. But you can't satisfy everyone's needs.

Prices also help keep the foul-smelling derelicts off the bus.

Maybe you think "well, foul-smelling derelicts deserve to ride the bus." And you could well be right.

But other people won't ride the bus with the foul-smelling derelicts.

If your goal is have a bus system you can feel happy about, this isn't a problem. If your goal is to have a bus system that people actually use, this is a big problem.

So in a nutshell, your argument is that if the public transportation were free, nobody would use it because it's too crowded?

It takes very very few derelicts to make the bus system too crowded to use.

when it becomes free how do you keep people from living on it?

Give them a better option?

Crazy talk. Clearly they have housing and jobs, they just like living on the bus!

I feel like you would like this, to punish society for not "giving them a better option".

Whatever the merits of the argument, that's obviously not it.

A bus full of people commuting to jobs is getting more use than a bus a quarter full of rank homeless men with lice dropping off of them and otherwise empty because no one wants to be near them.

I would argue, though, that a bigger problem is that prices don't wholly fix even that problem. Evidence: a casual ride on the non-free San Francisco area BART, which is full of homeless people who ride it all non-peak hours because it's more comfortable than sitting on the sidewalk.

When I've been in Asian or European cities, that's simply not an issue.

I've gotten off buses because of stinky people.

The fix is to provide sanitation and hygiene resources.

I would never want to deter, filter, screen any person (who is not harming others) from public services.

Tolls, fares, fees, onerous parking and impound rules, etc are a way to separate "us" from "them". It's becoming more acute as inequity continues to increase (in the USA).

stinky people aren't caused by lack of access to sanitation and hygiene resources in America. It's much more likely to be a personality or mental disorder.

Mental disorders are caused in large part by lack of access to jobs and housing[]. Do you not think that a lack of access to jobs and housing contribute to a lack of access to sanitation and hygiene?

http://www.nationalhomeless.org/factsheets/Mental_Illness.ht...

Is their stink harming you? Honest question.

Does something have to cause physical harm to be extremely unpleasant?

The point was that maybe someone who's extremely stinky is harming others, and thus could be legitimately

> deter[ed], filter[ed], screen[ed] [...] from public services.

Making transit free isn't going to solve homelessness. Also, plenty of people still ride transit even if someone they dislike is on said transit because they depend on that transit, unpleasant ride or not.

We don't need to encourage people who depend on the bus to ride it, because they depend on it. We need to encourage people who are otherwise riding cars to ride the bus.

Sometimes the strategy for that is to make driving cars suck, but that is a political loser before it even starts.

Or just add congestion fees for cars entering the city center, and use those revenues to fund citywide public transportation.

"Funding" isn't the issue. You want the bus system to be good enough that people want to use it.

I'm sure there are several dozen people who live within 2 miles of where I live that work within 2 miles of where I work. Batch us into groups by pickup time, dispatch small buses to my neighborhood to pick us up, sending texts to give estimates of arrival, drive us to drop off spots. Make the bus climate controlled and with power adapters and WiFi so we can work or play or read while this is happening.

I'd take that now.

It's already here! Just get yourself hired at Google.

I would argue that prices don't really keep them off the bus.

Originally from Chicago with a pay-per-ride system and moved here to San Francisco to a (weak)proof-of-payment system, sure it's faster but far more people take the risk of not paying in the proof-of-payment system (even those who can afford it).

Pay-per-ride keeps them off the bus, but as the article pointed out, the latency of the bus stopping until the bus moving is much higher and unacceptable in larger suburban areas.

Proof-of-payment does not keep anyone off the bus. Especially the "foul-smelling derelicts". 1) they game it by getting on and then getting off before the proof needs to be presented and 2) if they're caught, the fines are too steep for them to afford anyways. A night in jail may cost the city as much as the fine that would've been issued.

Free public transportation has a lot of benefits, but the elephant in the room is "Who is going to pay for it?". In cities that have a lot of tourists, the residents may complain and prices at businesses may be higher due to passed down costs to consumers. On the other hand, I pay for a monthly pass on a recurring schedule, it might as well be the same as tax, as it would make no difference coming out of my wallet anyways.

I think there's also an argument that it's not particularly unfair for drivers. One current trend is to charge drivers for access to congested areas: congestion charges in city centers, toll roads along congested routes, and fast-track toll lanes on non-toll highways. An alternate approach is to spend that money on getting other people off the road: subsidize transit so that some percentage of my would-be fellow drivers get off the road and onto buses or trains, leaving a less congested road for me.

Exactly. I've never understood why proponents of public transit don't say this more often. The more people that use public transit the fewer people who drive. Which means less traffic for everyone else still driving.

I read about a study in the book Traffic[1] that seems to negate this view, for buses at least. The findings were as follows:

1. More people use buses along a certain stretch of road.

2. Said stretch of road sees less congestion. Bus riders save time!

3. Noticing that this road has less congestion, more drivers take this route.

4. Route becomes congested due to increased interest.

5. Buses take just as long as cars to get to destination.

6. Bus riders migrate back to cars due to lost advantage of bus.

7. Road is more congested than before.

Overall, Traffic is a fantastic read and really opened my eyes.

[1] http://www.amazon.com/Traffic-Drive-What-Says-About/dp/03072...

Is this a case study, or a way something "could" go bad? Specifically 6. Bus riders migrate back to cars due to lost advantage of bus seems to encompass some assumptions that aren't necessarily true:

1) Bus riders have cars

In SF, many people don't own cars because the city isn't particularly car friendly. If buses get too slow, I can see some riders using alternate transportation, but I doubt very many would choose to use a car (and most probably don't have one to use).

Note: I don't live in SF, but was born, raises and continue to live an hour North. My primary mode of travel there is in a car, and it's not ever pleasant.

2) Bus riders ride buss primarily because it's quicker

What about being cheaper? What about convenience (it's a different type of convenience, but it's liberating to not have to worry your vehicle is safe and whether you've forgotten keys somewhere).

Think of a suburbs -> downtown commute situation. Houston comes to mind here.

I believe that was the study. I'll have to go look at the source to be sure, though.

Speaking of convenience factor, I can't wait for self-driving cars in this regard. I know it's a topic change, but widespread autonomous transport will clear up much of our traffic problems. Indeed, humans are the weak link when it comes to traffic. Not roads, signs, capacity, mass-transit, etc. but humans. We're just not capable of coordinating our behavior in traffic.

> Think of a suburbs -> downtown commute situation. Houston comes to mind here.

> I believe that was the study. I'll have to go look at the source to be sure, though.

That makes more sense, although I have to imagine some prioritization of public transit would alleviate this somewhat.

> Speaking of convenience factor, I can't wait for self-driving cars in this regard

Imagine self driving mixed with public transportation. Instead of large buses (or in addition to) we could have more, smaller vans, seating 6-10, depending. We could have multiple per prior bus serviced route if replacing a bus, and they could be more accurately dispatched base on load (having a bus service a route when there's few to no riders is a waste). Additionally, these could be used to add less commonly used, but still beneficial routes between farther points.

To really make it next gen, you could allow people to reserve seats online for a small fee, which would give useful information on route usage and upcoming demand, to allow reserve units to be dispatched accordingly.

If the cost of the drivers is removed, and the vehicle cost and repair can be brought down, a lot of really interesting things could be done with public transit.

I found the relevant section in my copy of Traffic:

  ...congestion pricing can help reverse a long-standing
  vicious cycle of traffic, one that removes the incentives
  to take public transportation. The more people who choose 
  to drive to work, the worse the traffic. This raises the
  time the buses must spend in traffic, which raises the
  cost for bus companies, who raise the fares for bus
  commuters -- who are being penalized despite their own
  efforts to reduce total traffic. As the bus becomes less
  of a good deal, more people defect to cars, making things
  worse for the bus riders, who have even less incentive to
  ride the bus. (p. 167)

The author cites congestion pricing as the solution, with money raised going to pay for buses.

And this brings us full-circle to the argument that those buses should be free for riders. If a city charged cars to drive during congested times and subsidized bus fares with that money, this might be a solution to ease up congestion.

5 => 6 seems odd. The advantage of the bus is basically never that it gets you to your destination faster. Buses stop every couple blocks; they're not going to beat cars taking the same route and not stopping.

The advantage of the bus is 1) it's (hopefully) cheaper; 2) I can put the time to better purpose than staring at the road and listening to NPR; 3) I don't have to be frustrated by parking; 4) if I meander from where I started I don't have to make my way back there; 5) (addendum to 1) for my commute I can spend pre-tax money on it, making it cheaper still.

For me, I find 2 the most important, though sometimes 3 is substantial.

You're forgetting one of they key points of buses in dense places: Parking!

Here in DC I can out-drive the bus on any route in the city, but when you factor in 20+ dollars/day for parking downtown, or spending 30 minutes finding a street spot, the bus wins hands down.

Parking was number 3, though I'll admit I may have underplayed it.

When I was attending university, there was a big problem with parking. So they built lots of parking space, more or less they tripled the parking lots. Net effect: there was still the same problem with parking, but now there was also a bigger traffic problem. On the other hand, buses were less crowded :)

While it adds little but perhaps a smile for a few readers, relevant reading - http://www.theonion.com/articles/report-98-percent-of-us-com...

The onion was the only major newspaper to pick this one up.

"98% of US commuters favor public transportation for others"

http://www.theonion.com/articles/report-98-percent-of-us-com...

Transit subsidies are actually what the generated toll money is supposed to be used for. The point of congestion pricing is to induce a portion of the population to switch transportation modes by making public transit a more attractive option in terms of price and quality. When London implemented the its congestion charge, there were transit improvements in place on day one: http://en.wikipedia.org/wiki/London_congestion_charge#Public...

In the US, the toll money often goes toward expensive, politically-motivated capital improvements, not operation expenses. In the worse case scenario, the money is used to build more toll lanes, which isn't going to encourage mode shift.

>In the US, the toll money often goes toward expensive, politically-motivated capital improvements, not operation expenses.

There isn't a transit system in the US that isn't subsidized. What difference does it make that toll money goes somewhere else if subsidies come from general funds? Money is fungible.

Being explicitly "paid for" from a specific source vs coming from the general fund is a powerful political symbol, which has significant, real consequences whenever deficit hawks go looking for things to cut. It's easier to convince the public that you should cut something when you convince everyone that their money is being spent ("wasted"), as opposed to having a specific revenue policy to counterbalance specific spending. It's a lot easier for people to assume that massive, massive amounts are being spent/wasted when it's coming from general funds.

Where I live politicians don't demonstrate any unwillingness to raid funding sources like tolls and property taxes whenever they run out of money. That's probably how we got here in the first place.

Personal anecdote. I am a student. In the town where I live, university students (and faculty and staff if I'm not mistaken) can ride the municipal buses for free. This has had an enormous impact on my behavior, even versus a very small charge like 25 cents. I don't often carry cash, and maintaining a fare card is a hassle, so if I had to pay for the bus I probably wouldn't use it very often. But since I can hop on to any bus I happen upon without worrying about paying or figuring out which transfers I need, I use it all the time.

Personal anecdote: it isn't price that is the problem, it is "convenience". Growing up, the only bus I could take came once an hour, and usually only took me halfway to my destination. So I walked, and often beat the bus.