Metrorail vs. Uber: Travel Time and Cost | Hacker News Reader

Metrorail vs. Uber: Travel Time and Cost

districtmeasured.com

50 pointswallflower44 comments

Where cost = financial cost to the end user, not the cost to society when private corporations suck revenue out of public transportation systems for short-term market gain.

I agree with your first statement, but not the followip.

Public transportation systems are woefully inefficient. If we made them operate more like private corporations, I would agree more.

The UK is a good example of the difference between public and private corporations controlling public transport in buses. In London they are publicly controlled and very popular while outside of London they aren't and are terrible. London has service through the night and very high frequency while outside of London buses stop running very early and are very infrequent. They are also significantly more expensive outside of London.

London isn't the only city in the UK.

I can't compare prices but buses run all night in Manchester and at high frequency during the day.

Manchester also has publicly controlled public transport.

Most of the buses are operated by Stagecoach though.

But they don't decide the schedule or price or anything like that. The key thing is that the people who decide that aren't motivated only by profit.

Nothing inherently more efficient about private companies over public ones. Private companies have all sorts of inefficiencies that public organizations don't have, like profit and other requirements.

You can send mail across the country via USPS for $.50. Or, you could pay $8 to send it privately via UPS or FedEx.

> You can send mail across the country via USPS for $.50. Or, you could pay $8 to send it privately via UPS or FedEx.

This is literally the worst example you could use. USPS has a legally enforced monopoly on first class mail.

Usually monopolies enable higher prices rather than lower ones.

It's a different service. USPS has a monopoly on letter mail. FedEx is allowed to ship documents but only as an express delivery (the USPS service that is as fast as FedEx at document delivery costs roughly the same).

>>Nothing inherently more efficient about private companies over public ones.

>>You can send mail across the country via USPS

USPS has to service everywhere. This is usually the case with all publicly funded "companies." This is inherently inefficient. It is more fair, but that's not the argument. Private companies are absolutely inherently more efficient.

Private companies are absolutely more efficient if you discount externalities.

Of course, externalities are the entire rationale for having public services.

Effiency is about archiving a goal with the least amunt of resources, not about setting the goal. The USPS has the goal of delivering mail to everyone and is obviously pretty efficient at that.

That’s kind of the point of public services. They have to service everyone equally. If that makes them inefficient, so be it. It makes them more fair. The efficiencies of the private companies you speak of also disenfranchise portions of the population. If we privatize schools, poor kids wouldn’t get an education, because their parents couldn’t afford it. As a society, we’ve decided that isn’t fair, so education is a public service. And that analogy can be applied to a lot of services. Sometimes some of us have to give up a little quality so that others can have anything at all. That’s my 2cents.

Edit: I don’t think I had enough coffee before I wrote all that, because I didn’t read the above comment well enough. Part of what I wanted to say is that efficiency doesn’t have to be about how well it serves an individual. It can be measured by how well it serves a population. Which then factors in fairness. We can’t measure the efficiency of a road based on how many cars can drive on it, whe what matters is how many people can move on it. (Buses and bikes allow for higher throughput.) The mail system can’t really be considered efficient if only some people can use it. It may get one letter from point a to b fast, but the systems effciency depends on serving everyone.

Except for in cases of monopolies, externalities, societal + shared goods and utilities.

There are good arguments for at least several of those cases apply to uber in even a simple analysis: uber use creates pollution, traffic is less safe and causes accidents + deaths, uses more fuel, and few markets charge vehicles for road use by the meter/use. There are high barriers to entry and sunk costs to infrastructure to competing because if you do not like the road between two points, you cant just open up a rail on top of it that traverses the same route, and attempts to do so lead to massive inefficiencies due to the duplication in the relatively expensive infrastructure costs externalities (imagine the cost of trying to lay two simultaneous water networks, for example).

Additionally, while uber may be more efficient in some routes, given population densities and given infrastructure and desired travel routes and given transport network patronage, uber trip cost/benefits don't scale. The marginal cost/benefit of you taking an uber trip rather than catching the train is not the same as the cost/benefit or qualities of taking uber if everyone whom the train was transporting was attempting to take uber.

I'm not sure how a larger service requirement plays into the efficiency of completing that action. The question in this case isn't about that, but instead whether or not a public entity or a private entity tasked with the same service surface is inherently more efficient. That said, I don't know the answer to this question offhand.

In the UK we had public bus services which were privatised in the 80s.

Prices massively increased. Services were removed from rural areas.

Stagecoach engaged in unethical practices -

Going into towns and putting free / low cost buses each side of the local ones until they destroyed the local service. At that point they would raise prices and reduce frequency.

Self-replying since I have many down-votes and a long multitude of replies.

A few rebuttals:

* it's not sufficient to evaluate the quality of service -- e.g. look at this nice public bus service -- you must consider QoS/Cost

* you cannot simply look at the cost at point of sale -- e.g. a postage stamp is $0.50 -- you must incorporate the government subsidies paid for by taxes

* I concede that merely being a private organization is not a panacea to optimal QoS/Cost

>If we made them operate more like private corporations, I would agree more

What specifically would you change to make them operate more like private corporations, and how would it help?

many reasons, the first being that public entities would not be in a rush to spend a fortune on ill advised light rail systems which have run some systems into the billions with even more in deferred maintenance. light rail is great for connection distant suburbs or major points of attraction like an airport to the city center. however it cannot adjust for changes population needs. only NYC has ever shown real ridership increases in the last fifty years. US light rail has a nearly sixty BILLION dollar backlog of maintenance because it cost to much to build, run, and maintain. what private company would even be allowed into that mess while serving the public?

it also probably would not take the NAACP for force a privately run bus line for a city to serve minorities but it did in LA against the government run system.

want to extend it further, most public systems are bastions of nepotism where friends and families of politicians place them in positions they are not qualified for to earn big checks and pensions.

The tokyo system seems to be a fairly efficient mix of public and private metros and train lines.

The tokyo system isn't really subject to market competition, because employers pay for employees' tickets.

From what I've seen it's not any more subsidized by employers as it is in the US with commuter benefits.

That being said, I'd say the NYC metro is probably the best subway system in the US and it's nowhere as good as the Tokyo area transit system.

Are the private metro and train lines more efficient than the public ones?

Not really. There's a lot of interoperability involved (all transit systems can use each other's IC card and you can charge it at any company's terminals). The IC card can also be used at most merchants.

This is not a fair comparison:

1.) The real cost for an Uber ride is opaque, at best.

2.) Unlike Uber, public transportation becomes exponentially better with infrastructure upgrades. These upgrades would cease (in both mind and in implementation) to exist with the proliferation of Uber and other "car sharing" rides.

3.) Road travel is more subsidized than rail, anecdotally. I hate to say I don't have hard data for this, but I'm pretty sure it's right (sure, there's a gas tax, however cars are subsidized both implicitly, with parking lots and real estate, and explicitly, with wars and other foreign policy to manipulate the price of gas).

> Road travel is more subsidized than rail, anecdotally

New York spends between $1.11 and $7.34 subsidizing each ride on our Subway or LIRR, respectively, [1]. Assuming 2 trips a day on each of a year's 260 workdays, that comes to $577 to $3,810 of subsidies per commuter per year. The article is from 2012, so we'll use BLS data [2] to adjust those figures to September 2017 dollars: $618 and $4,080, respectively.

"The statewide average of highway spending per capita in 2009 was $233," [3] or $266 in September 2017 dollars. That said, this isn't a fair comparison since we aren't adjusting for usage. On the other hand, most of roads' wear and tear comes from trucks [3]. And New York taxes the hell out of trucks [4].

TL; DR Trains work, but in the United States we tend to treat them as jobs programs first and public infrastructure second.

[1] http://www.wnyc.org/story/283927-mta-suburban-passengers-get...

[2] https://www.bls.gov/data/inflation_calculator.htm

[3] http://facweb.knowlton.ohio-state.edu/pviton/courses2/crp776...

[4] https://www.tax.ny.gov/bus/hut/huidx.htm

You're comparing different values here - subsidies per commuter vs subsidies per capita. To be fair you should recalculate the road spending to be paid only by those who use it for commuting. Not to mention that number only includes highway spending and not regular roads.

I would like to add a couple of points of clarification to your comment:

1. Subway rides for most commuters are money-makers for the MTA, it's rides late at night or at the ends of lines that lose money (full trains are obviously more cost effective). I suspect LIRR costs work out similarly though I have not seen enough of their data to know.

2. The NYC bus system is massively subsidized. The subway could operate (less frequently, and to fewer places) and make a profit, the bus system could not.

I can probably find my old sources for that if anyone is terribly interested.

In NYC, the cost of the land area that roads take up is far more significant than the cost of the actual road itself.

They have been planning to bring more light rail and more buses for decades. I’ll go with the solution that works today. When the transit authority and politicians get off their ass and install the lines we agreed and paid for then we can revisit the perfect hypothetical.

Cars are mostly subsidized by way of devoting the majority of publicly owned land to roads. If the government were to auction off the land under every street and highway, that would pay off the national debt in no time.

Just because I don't have a car doesn't mean I don't use roads (eg. as of now, the groceries I buy weren't delivered to the store by drones)

That doesn't change the fact that the cars and trucks that drive on them are subsidized. What if those goods were delivered by train?

Very useful. My experience, however, is that unless you live or are going to an event/work at a metro station, the "last mile" benefit of ride sharing tips the scales towards road travel in most instances.

Modeling metro delay as 0 minutes is completely unrealistic. For example, the scheduled trip time from New Carrollton to Foggy Bottom (all on the Orange Line) is 36 minutes. In reality, it often runs well over 50 minutes.

This is great. I wish there was consensus that data was the basis for opinions.

Just like with Internet service, the real cost (time) is in the last mile, as bkohlmann refers to. Uber picks you up at your door. Did the study figure the average walking time to a Metro station into their statistics for random addresses? I doubt it.

https://en.wikipedia.org/wiki/Lies,_damned_lies,_and_statist...

If you look closely, you'll see they did:

"A trip from Metro Center to Bethesda, for instance, takes 33 minutes, and that includes time to wait for the train and walk to and from the station."

Just to note. In DC the train system is called the metro. Metrorail is the name for the miami transit train system. The author uses metro, metro train, and metrorail interchangeably. Left me slightly confused as if they were comparing DC to Miami.

In DC, the mass transit agency is usually referred to as Metro, and offers two main services - Metrorail and Metrobus.

I’ve only referred to the train as the metro and the buses as the bus. I’ve only seen in one city the locals refer to the train as the metrorail. And that was Miami.

In colloquial use, you are right, but in written media I see references to DC Metrorail all the time.

Reply on news.ycombinator.com