My understanding of the traditional conservative viewpoint is that trains and other mass transit should be private rather than government operations.

That being said, nobody but an idiot would have gone into railroads anytime over the past 50 years given the huge subsidy effectively given to cars and trucks. The columnist is completely right - the US has spent many trillions subsidizing autos, trucks, farms and a some other networks. Quite a bit of that has gone indirectly into the pockets of gas stations, oil companies, road contractors, car manufacturers and real estate speculators who bought small farms and make big suburbs.

It has rewritten the american landscape and culture in a way that would be similar in scope if not outcome as an equivalent subsidy of canals, or railroads, or bicycles. More mass transit would have resulted in higher densities along certain key lines; less diffusion, less parking lots, fewer accidents, much less emergency medical services, and a greater cohesion and intermingling of industrial/commercial/resident­ial - with the bad health outcomes that might have precipitated.

So, there would have been quite a bit of good, and some bad.

Fortunately, we are beyond having to subsidize shared, mass, and public transportation to get it to work. Information technology has opened up some new vistas which can enable shared/mass/public transportation to not only compete favorably with cars, but invade a car and truck dominated landscape by starting with the highest density areas and quickly spreading outwards. This, however, will work better and faster if the government stops the huge subsidies and imposes a few additional costs on drivers - more taxes on gasoline and cars; higher prices for non-commercial drivers licenses and more non-drivers licenses as IDs; higher insurance rates for cars and more tolls (all possible because of systems like EZ-pass, so we don't need so many toll workers).

We can make the transition. It will be the death knell, perhaps, of some communities - but by in large it shouldn't kill too many substantial towns; it will just reallocate their real estate values, much like rising sea levels will do for the coastal communities. Things change, we can't pretend to be 'fair' to everyone who took a bad bet on their home prices. We can, however, do better by the coming generations.

While it is seductive to think we can 'just let the market take care of our transportation problem' but business realities make this approach impractical.

Problem 1

Building a new mass transit system in a medium size city (the Tampa Bay area for example) will take capital expenditure over several decades.During that time period the business is unlikely to be making any profit. Business that lose money for decades don't stay in busineess.

Problem 1.a

Building a new mass transit system will require a very significant taking of land. (I know you can tunnel but it cost a lot more and in some settings it is cost prohibitive.) Lots of land will need to be taken by imminent domain. Giving (or leasing) land acquired this way to private companies is fraught with both legal and political problems.

I have in previous posts to the fray discussed viable options for creating intermediate transit systems that do not rely on fixed routes, let alone rails or tunnels, for the creation of shared/public transportation grids. If you can start by dynamically routing shared taxis, multi-passenger vans, and small busses, you can increase transit efficiencies by an order of magnitude, almost for free and quickly. Adding some trunk routes, or even dynamically routed larger buses (30 + passengers) can extend that perhaps another order of magnitude.

That's before you start putting in major trunk lines on rails, which would be profitable very very quickly if people were being actively routed through them via a dynamically optimized system that had strong predictive power as to where the lines were needed.

Part of the problem with building light rail and such in the past was that once a neighborhood needed it, it was so dense that it was impractical to condemn enough land to build the system - but to build it before it was needed doomed the project to sit mostly underutilized for many years before the community built up around it. Rare exceptions occured in NY or Boston, where outlying cities were essentially joined by rails and then then communities in between filled in, or where projects occured so early that condemning slums was considered an improving experience, morally, for the people living in said slums.

In this case, we would expect the new lines to be profitable on margin immediately and to start to pay down investment later. Not that these lines might not be built by government - just that not all the mass/shared transit needs to be spearheaded by government, as it was before.

We all will probably think first of the dynamic routing of baggage at the American Airlines terminal in JFK! Oh, the visions it brings on.

I don't question that one can improve public transit somewhat with dynamic routing - my city (town? whatever. Pop 55,000) has a number of minibuses (wheelchair accessible) which are dynamically routed to move senior citizens around. When I see one go by, the are usually almost empty. I would guess that, rather than the order of magnitude you suggest, the improvement over static routing would be about 50%, at best.

Great thread. My 2 cents to add:

The market will react best if there is a clear plan set out, with a gradual shift. By gradual shift, I mean begin by freezing the car subsidy at current levels, no new construction projects beyond repair work no new car tax breaks etc. Then progressively decrease the subsidy, while upping the alternatives subsidies (not necessarily permanently).

If the timelines are clear, businesses will be able to properly plan.

So, first, I'd like to thank you for actually commenting sensibly on the topic at hand.

Then I'd like to address your two issues;

1. dynamic routing of baggage at airports. I don't know much about it. It seems to me that baggage handling works more or less well most of the time and when it doesn't, its a disaster. But I don't know enough to say much.

2. Senior citizens; so here's the fundamental problems with this as the example: too few clients and probably too crude a routing system. If you have the majority of customers with very special and different needs (oxygen, wheelchairs, etc) things get inefficient fast. If you were to open the system to half the community, for example, you might do better in terms of filling up the buses. Especially if people could be assured that the moment they contact the system, they will learn an honest 95% upper bound on when they will arrive. Take out the fear, doubt and uncertainty - and make the cost transparent prior to commitment.

The key here is that with dynamic rather than static routing, you can cover much more territory more efficiently than with a single, say, bus line. A traditional bus needs to come at least 1 time every 15 minutes virtually round the clock (or at least during daylight hours) for it to be dependable enough for people to start relying on it and taking it. Even then, it only really covers an area a few blocks in each direction off the route.

But you still need a critical mass to make it worthwhile, and seniors are unlikely to do that by virtue of being a much smaller and less active population, despite them being dependant on the service.

There is data available and, hopefully, it could cover many areas and many years. I would assume that a lot of the older data has been thrown away, but there is also a good chance that much of it was abstracted by earlier researchers. Neither of us has mentioned this area yet, so can you guess what I am referring to?

Radio dispatched taxis! They are dynamically routed at customer demand, may do some schedule juggling to minimize mileage, and often don't return to a "home" location between trips. This is a close enough model to what you are suggesting to be useful information. One issue, which is not normally recorded by the operators, is how the use of the service is distributed over the populous. However, since both origination and destination locations are recorded, clever demographic analysis could possibly fill in that portion of the record.

Couldn't agree with you more, and actually quite a bit ahead of you on this.

One issue with Taxis is that the drivers have often resisted being 'routed' - especially with GPS.

Also, most taxis don't do 'ride sharing' and the current systems are set up in ways that discourage multiple concurrent fares.

Further, taxis currently have a huge problem - fear, doubt and uncertainty on the parts of the driver and the passenger. The driver doesn't know how much he's going to get paid before he picks up the fare (destinations change) and before the fare leaves the cab (stiffed on tips). Also, traffic can change the duration of the trip, baggage is an unknown.

The passenger doesn't know how long the trip will take, when the cab will arrive, how much the trip will cost, whether he will get to his destination on time. Anyone who has experience with dispatched cabs knows the horrors of depending in their services.

With judicious reforms of the regulatory side of things and a serious influx of technology, all of this can be radically overhauled for the better. First, the regulatory side - shared fares need to be encouraged in any way possible. Customers need to know that having someone else get in the cab is a good thing.

The other sides - passengers need to have a way of knowing, with some guarantee, that they will get to their destination in X amount of time at Y cost; and they need to get their reliably. Many times even a free ride isn't worth it if you miss the plane. =( Best option is that the customer can have several choices - some faster and more expensive; cost dependant on luggage and willingness to share.

Then the cabbie - the cabbie needs to know that the customer is where they say they are, able and willing to pay, going where they said they are going; what they are carrying. All of this can be helped with a gps phone that confirms location and time; and prepayment that includes the tip - which can be withdrawn after, say, a complaint. If a person complains too many times without cause, perhaps service can be cut off by the cab company. Just the mere act of requiring a complaint to withdraw the tip, though, would be a big gain for cabbies.

All of this can integrate with the public transit system if cabs get directed to take people to mass transit stations and pick people up from mass transit stations, so that a fare is greatly decreased when mass transit can help out - turns a $200 fare into $15 + $2.50 on the subway, for example - but it also gives the cabbies some purchase on people who would otherwise switch between walking and driving. Those people suddenly don't need cars at all... no parking, no traffic...

It occurs to me that airport buses - the ones that pick you up at your home and drop you at the appropriate terminal and later reverse that process - come pretty close to what you profer as an optimum application. However, they don't really solve the many-source-to-many-destinatio­n route requirement. That of course, could be handled by a number of scatter-gather terminals. That is equivalent to joining the source and destination airline terminals without a flight in the middle. The architecture is equivalent to a database, I think.

The model I lean toward, incidentally, is one which provides the customer (as his own property) with a minimal vehicle, maybe a smart car or, where there is little rain or snow, a Segway. They would make a short trip to a local mass transit (possibly light rail, or segmented bus) terminal, be whisked away in a large group, and disembark at a terminal near their target. Small busses (dynamically routed but departing on schedule to synchronize with the mass transit arrivals) would distribute the customers to their end targets. The whole process would reverse direction to deliver the customer back to his personal transit device for the last mile or two of his trip.

Obviously, one could counter that the personal transporters could be eliminated by providing dynamically routed transportation from home to mass transit terminal, but that would eliminate the opportunity to use the personal transport for near-t0-home tasks that don't include use of the mass transport system.

Great thread all. I like this kind of thinking that begins with addressing a socially concrete concept that seems imovable, but rests at the heart of a real problem. I like it because it's tickles my "stranger things have happened" bone and motivates me to not give up on what's perceptibly "impossible" to change.

Thanks for keeping the box open!

Finally, I like the thread for reason's that are clearly selfish: it reminds of a similar thread I started - though your posts are much more knowledgable, that's for sure!

<link>

I agree that airport terminals are one of those 1 to many places, and 1 to many, the information isn't too crazy. But many to many is an information nightmare, which is why nobody did dynamic van/bus routing a few years ago. It's too hard to plan. But now, we have big fast computers. And expensive gas.

Well, this is where a nice IT solution becomes handy.

We can do the dynamic routing as a general solution.

The MBTA in boston includes in its route planner 'how far are you willing to walk?' This is the ultimate in pairing personal and mass transportation - but we could do better by including things like cars, bikes, subways, taxis, buses, vans, etc - into the total package. If the user has a bike, fine, we can plan that in.

What I would prefer to avoid is the utopian (and generally unrealistic) option that requires everybody to buy some one thing. Most people have a cell phone; including cell phones, possibly computers, as access points to the system, doesn't make a huge leap. There exist taxis. Buses. Subways. Might as well make use of them. Can we use them better? Can we provide dispatchers with cool new routing tools? Great. Can we keep the airports, buses, subways all in our planner so that they work together better? Can we tie in dynamic routing to this well? Perfect. Do we have to give terminals to the taxi drivers? Unfortunate, but not necessarily fatal.

It's getting cheap, after all, to get computer time, database time, map access, traffic prediction data, billing, and communications. What still costs lots of money - vehicles, drivers and fuel, and marketing.

So minimize new investments in those things. Make them as cheap as possible. Don't stir up competition when you can create collaborations. Get more people into the existing vehicles...

The following idea came up in a thread someone else began about hitchhiking a couple months back, but could tie in nicely to some of these ideas.

If you took Twitter and added an ebay like user feedback system, you could easily increase ride-sharing. As a starting point it has the advantage of being independent of any given operating system. It would require a critical mass to become truly useful but requires no government intervention.

That makes sense for highways and on-ramps, everywhere else it just seems like government intruding on individual decisions. If the government isn't going to help it should just get out of the way.

Still, hitchiking is just why we had thought about it before. My point was more that we have the technology right now to begin implementing the beginning of these ideas. A cab company could have gps phones twittering there locations and scanning for nearby twitters with preset codes. Users and drivers could rate each other anonymously.

Kh, thanks for posting that link. I had forgotten who I was talking to before.

The US government denied subsidies for passenger railroads while it was subsidizing air travel. This is what helped to effectively kill passenger rail transportation in the USA.

As we know, big business corrupts democracy almost as a way of doing business and our Congress (being very well paid by special interests) does not see past the next election financing scheme.

If our backward-thinking politicians had not messed things up we would probably have effective and affordable high-speed rail passenger service today.