Hacker News Reader

Comment by triggercut | Hacker News Reader

It's a shame that the focus is always on the capital outlay and not on the cost benefit potential to be realized (both economic and non-economic). This sort of spending should always be seen as an investment.

For instance, modern/improved signaling and train control technology will allow you to run rolling stock at greater speeds with more frequency, requiring less on-track maintenance. In other systems, undergoing recent upgrades (for example) you can run a train every 3 minutes as opposed to 6. As a commuter knowing you only have to wait a max (n)minutes as opposed to max 2(n) can be the difference between deciding to take a train or using some other means of transport, especially if you can shave a few minutes off your commute as well.

It gives the operators much more control over how they design services and run their business. It's a key enabler.

View on HN

> It's a shame that the focus is always on the capital outlay and not on the cost benefit potential to be realized (both economic and non-economic). This sort of spending should always be seen as an investment.

Every major city hires economists, often from universities, to do cost/benefit analysis of major capital projects. To say the focus is always on cost is nonsense. I've had to do cost-benefit analysis work before on earmarks. One of my professors did cost/benefit in Nashville for bringing the Titans to town in the late 90s. He recommended against it, in fact I think most economists on the board did...but they ignored it because cost is more often ignored or understated (by politicians,) contrary to your post. Look at the "big dig" in Boston for instance, or the California high speed rail, both exploded over 5x their expected budget.

Those are just examples, there are many others. I am willing to bet this number is a low ball, especially when contractors, regulators, etc get involved.

I think he's talking about the media/politicians/pundits that talk about it, not the people actually responsible for making and analyzing the decisions.

"Every major city hires economists...but (the politicians) ignored (the advice.)"

I think this is what he's getting at.

>Every major city hires economists, often from universities, to do cost/benefit analysis of major capital projects.

To be clear I was referring to how the greater public are informed on such matters by the media / politicians.

I'm well aware of the projects you mention. I'm a consultant on capital program / project governance but not sure I understand the relevance, I wasn't really talking about the validity of estimates at any stage, just the focus on them.

If you’re interested in cost benefit analysis of large public projects, Megaprojects and Risk is a must-read.

https://en.m.wikipedia.org/wiki/Megaprojects_and_Risk

Yes, a fantastic read.

If you're interested in how we can avoid this happening, I can't talk to the US, but in other jurisdictions governments and authorities (and some private sector giants) are slowly adopting more mature governance approaches. Implementing a Benefits Realization methodology [https://en.wikipedia.org/wiki/Benefits_realisation_managemen...] in their Portfolio for instance, Alliance contracting procurement models and Integrated Project Delivery to more fairly apportion risk and accountability. I'm cautiously optimistic.

you completely missed his point

California high-speed rail has not exploded 5x.

When originally approved in 2008 the estimate was $40 billion.[1] The 2018 estimate for both phases is, I think, roughly $100 - $120 billion (low and high).[2] That's 2.5x - 3x.[3] These numbers are actually in line with opposition figures[4], and for other reasons it's reasonable to expect these are solid upper limits unless people continue to sabotage the project.

It's also worth pointing out that $120 billion is still significantly less than estimated costs for the alternative of expanding the highways and airports--$156-186 billion.[5] Note that those alternative costs are likely optimistic like the original HSR costs, particularly airport expansion and urban highway expansion.

One of the biggest reasons for the cost inflation is the political opposition. Some of the construction contracts have actually come in significantly under budget.[6] The real killer so far is land acquisition.[7] The authority is effectively trying to bribe farmers with huge payouts (clearly beyond market prices) in order to mitigate political pressure in Sacramento. All the political and legal fighting causes expensive delays. (Time is very costly in construction because there are fixed costs in keeping labor and equipment mobilized, plus huge costs in having to re-orchestrate schedules. Lots of interdependent moving parts.) But political pressure wouldn't be so strong if discourse was more honest and less tribal.

The opposition is creating a self-fulfilling prophecy, particularly with the hyperbole and sabotage. By all means fight for transparency, but civic responsibility means we all need to check ourselves.

[1] https://ballotpedia.org/wiki/index.php/California_Propositio...

[2] https://sf.curbed.com/2018/1/29/16946576/san-jose-mercury-hi... and http://www.hsr.ca.gov/docs/about/business_plans/Draft_2018_B...

[3] I think both numbers are year-of-expenditure, which means each sub-project is separately inflation adjusted to the midpoint year of scheduled construction.

[4] https://reason.org/wp-content/uploads/files/california_high_...

[5] http://www.hsr.ca.gov/docs/about/business_plans/BPlan_2012Co...

[6] http://www.cahsrblog.com/2013/04/chsra-selects-tutor-perini-...

[7] https://www.mercurynews.com/2018/01/17/worst-case-scenario-h...

> As a commuter knowing you only have to wait a max (n)minutes as opposed to max 2(n) can be the difference between deciding to take a train or using some other means of transport

Agreed. In my mind, the most important step to make is to increase frequency to the point where passengers think of trains in terms of frequency of arrivals, not by a specific timetable.

At longer wait times, you plan around getting to the station early, hoping the train isn't late, hoping it's not too overcrowded, etc., each of which is a mental barrier to taking the train.

At shorter wait times, all of these concerns disappear because "oh well, there's another one in 3 minutes".

This, and reliability.

If I have to leave home 15–20 minutes earlier just because it is _very_ likely that the train gets stuck in a tunnel due to signal malfunctions, or crawls from one stop to the next, I might as well use alternative means (i.e., a car).

While I mostly agree with you, I think it's right to focus on capital outlay on one level: getting to the bottom of why it's so huge. Comparable cities like London and Paris complete large projects spending a fraction of the money, and they have to contend with unions and high rise building just like New York does. Increasingly it seems like New York State politics are corrupt to the core, and the MTA is just a reflection of that.

Not at all. Each actor should extract maximum compensation for the value they provide for their efforts. Europe underpays at every level so it’s “cheaper “ but not an accurate representation of fair market value.

Fair market value is not defined uniquely in one off transactions. It's very possible that what happens in both nations is a fair market price. If this is true it then becomes natural to ask how we, the taxpayer, can get a better deal like they do in Europe.

New York state politics (and also New York city politics) are indeed, corrupt to the core.

Do a search on "new york city political corruption" to see a selection of the scandals.

I'd recommend The Power Broker, which documents how Robert Moses used a toll authority to broker an unprecedented empire of spoils so that he could ram his highways through any opposition. And that was as late as '68.

> In other systems, undergoing recent upgrades (for example) you can run a train every 3 minutes as opposed to 6.

That's only 20 trains per hour. London's Victoria Line now runs at up to 36 trains per hour, a train every 100 seconds.

(Something I think about while waiting 20 minutes for a BaRT on a weekend.)

I assume OP was just giving an example, because the busiest lines in NYC subway (4/5/6) run about 20 trains an hour at peak times now, and that's with ancient signaling.

NY trains tend to have more tracks than London’s underground which is just north/southbound, instead of north, south, north express, south express (iirc).

The London Underground has more miles of track which go in all directions, not just north and south. In fact cross rail which is the latest large expansion goes East to West. New York however has more stations and along with that 36 lines compared with London's 11

New York's busiest track pair is scheduled for 30TPH (E/F in Queens).

These scheduled service levels tend not to pan out because of dwell times causing trains to get stuck in stations. CBTC helps mitigate this by allowing trains to run much closer, so the dwell time in stations is less of a limiting factor, and the busier frequency also allows platform overcrowding to be less severe.

That's interesting! I ride the E train every day that I commute into the city (half of the weak), and I'm pretty sure that even at peak times, in Manhattan, it's more than 4-minute spacing. I once lived by the F train, and it's hard to believe that it ran enough trains to make up for that. So 30TPH seems wistful, indeed!

The 4/5 actually seems to achieve 3-minute spacing consistently, though.

It's a lot more noticeable in Queens; trains roughly do show up every two minutes.

The major issue is that all the merging and interlining of services introduces many conflict points; most subway systems try to limit the branching of services on any given track pair, and very few have reverse-branching, the combination of branches into trunks in the outer parts of the system (like the E/F, the 2/5, the A/B/C/D, the B/Q, the D/N, etc.)

Yeah, post Christie Street Connection, the branching is pretty wild. Nice for having some good options that minimize transfers, but I have to imagine it's hell for scheduling. The C shares track with the B in Upper Manhattan, the E in Midtown and most of Lower Manhattan, and the A from Lower Manhattan through Downtown Brooklyn until finally getting its own track. Going by common tracks, the A, B, C, D, E, F, G, J, M, N, R, Q, W and Z have to be considered correlated. So that's basically the whole IND+BMT, except for the L!

Here's a pretty interesting map showing TPH (and number of train cars) in 1954 for the AM rush. I believe some lines are running fewer trains today than they were then.

http://transitmap.net/image/55177865550

The scheduled service on that map is 30TPH, although the theoretical max is 40TPH; the signalling system has never been used to that extent, mostly because the extra 30s between trains is used as contingency padding in case a train is late.

The Lexington Line actually ran 30TPH at one point between the 4/5, but this has decreased because overcrowding has gotten so severe that it is now holding up trains as people struggle to enter and exit overcrowded cars. Prior to the opening of the Second Avenue Subway, the Lexington Avenue Line carried more than the transit systems of San Francisco, Chicago, and Boston combined. http://web.mta.info/capital/sas_docs/feis.htm

NYC's system has a lot of lines that basically run two lines on shared track, split them off, and then join other lines in complex reverse branching scenarios. This absolutely kills the ability to run high TPH (the Washington, D.C. metro also suffers this problem). The Victoria line has no revenue branches nor line sharing--it is much easier to get high TPH on those kinds of lines. Moscow Metro, in a similar situation, can pull 40TPH.

> London's Victoria Line now runs at up to 36 trains per hour, a train every 100 seconds.

And have thinking about pushing it even further for quite a while now.

https://www.londonreconnections.com/2017/ninety-second-railw...

The reason the focus is on cost is that New York transit has a long history of spending far more than it needs to, due to rampant corruption[0].

If you look at this like an investment, then you also need to do the same sort of due diligence you would with any other investment, including asking why the costs are outrageously out of proportion with what they should be.

[0] https://nytimes.com/2017/12/28/nyregion/new-york-subway-cons...

You can't ignore things like: whether the debt can be afforded and paid back. Because these are public ventures, you practically know they'll never produce a suitable return on investment.

Did you know that the NYC subways were originally built and run by two private companies that competed with each other? Did you know that the city froze their rates for two decades then took them over when they were no longer making desired improvements (wonder why)?

On the other hand, government spending is designed to help the public good when private investment may not make financial sense. Parks, for example, would never provide an ROI if private companies had to build them. That’s why so many parks are public owned and built by the government. The ROI there comes from happy citizens who will pay taxes.

Even if a subway wouldn’t be profitable to a private company, it can still be “profitable” to the government if it encourages more population and more businesses paying more in taxes. It doesn’t have to be profitable based solely on ticket prices.

The private sector is not investing in public transit in NYC because IT IS NOT ALLOWED TO, not because it wouldn't if it could.

There used to be a lot of private sector public transit in NYC. Today there is none. That was not an inevitability!

Many cities around the world have extensive private sector public transit. E.g., Buenos Aires has privately-run public transit buses with amazing coverage of the city and surroundings. NYC's bus system is garbage by comparison to Buenos Aires'. The difference is that NYC's bus system is run by the city, while Buenos Aires' is run by private companies.

(It's always interesting to see what parts of the economy get socialized in different countries/cities. The public transit system was the easiest part of the private sector to socialize in the U.S., so that's what got socialized early. Elsewhere (e.g., in the UK) it was trains, telcos, medicine, etc. This allows us to compare results. It's never pretty, what results from socialization.)

Public transit in the U.S., and in NYC in particular, was very much a private sector function way back when. It seems unbelievable today that the private sector could build subways, but why should it be unbelievable? $19B is a lot of money for NYC, but it's a pittance compared to the capitalization of Apple or Amazon, or Elon Musk's ventures -- if only the private sector could do it, it absolutely would build a better transit system.

But the lesson has been learned: if you build something that cannot be moved, you will be highly vulnerable to takeover by the state.

Recall that the Baltimore Colts left the city in a hurry right before the governor of Maryland signed a bill appropriating the team...

You can't take a subway system with you should the city decide they want to own it instead of you, and you'll never be compensated the true value of the system you built no matter what the applicable Constitution(s) says.

Even if the private sector were now invited to invest in NYC public transit, without an ironclad guarantee that they won't get taken to the cleaners later, why should they?

But the point stands: if they were allowed to and not threatened with eventual state takeover, the private sector bloody well would invest in public transit, and it would do a far, far superior job by comparison to the state. We know that the private sector has, does, and will invest in public transit when they are allowed to and in locations where they reasonably believe their property will not be taken by the state.

If you wonder why Buenos Aires doesn't have a private sector subway system, the answer is obvious: the capital investment required for that is so much higher than for a bus system, that a private owner would stand to lose a lot more from a state takeover of their subway line than of their bus line, therefore no sane investor would invest in Bs.As.' subways. The same applies to NYC, except that in NYC the private sector is not allowed to run buses either...

You seem to have accepted the liberal mantra that some massive projects can only be run by the state. This is very, very wrong. How much evidence do we need before you and everyone else figure this out? Why would one believe that SpaceX is viable as a private company but subways are not? Elon Musk absolutely does not think this way. Yes, yes, Musk sure likes public subsidies, but I don't believe they are necessary for his ventures, and would much prefer that he not rely on them in any case, but even with subsidies his model is far superior to the state running those ventures outright.

>You seem to have accepted the liberal mantra

Aaaannd we’re done here.

Why?

I think the inefficiency of state controlled transit is worth discussing, but I think these examples are a bit cherrypicked and don't necessarily support claims like "[the private sector] would do a far, far superior job by comparison to the state." Unregulated private sector investing destroyed the traincar transit systems in the United States, and helped set a general national attitude towards cars vs transit that is not present in countries like Argentina. There are only a few privatized subways in the world, in Japan. Your argument waves away many of the realities of building and maintaining subterranean rail transit in a profitable private company. "If the private sector could do it" is a question staked on more than just government regulation.

Actually, in Argentina the use of long distance rail for mass passenger transit is increasing for the first time in years after the systems were nationalized.

Feel free to disregard what I've written if I, too, have accepted the liberal mantra. Issuing such a diagnosis poisons the discussion into an argument unnecessarily and suggests you're chanting a mantra of your own. Some of your examples are wildly picked and distant from the subject at hand, or downright confusing: What nationalized "telcos" in the UK are we comparing notes with? Can we use the highly regulated Korean ISP market as a substitute? Does it really support this point to crow that the US healthcare system is superior to socialized alternatives in European countries?

I said nothing about having no regulations.

My use of "liberal mantra" as to one specific idea, is not an ad hominem, nor does it mean that I'm not a liberal, for example, or that I don't like liberals. It just means that that particular idea is typically championed by liberals (which I believe is true). I don't see what's offensive about that.

As to long distance rail, it's not public transit, but sure, let's discuss it. I've no idea what's going on in Argentina with long distance rail. One possibility is that the roads suck, so rail looks good by comparison -- I just wouldn't know. Long distance rail has a lot of competition (cars, planes) -- it doesn't look like a long-term winner to me, but if so, that would be true regardless of which sector is to invest in rail.

The UK nationalized the telephone system around the time of WWI, IIRC.

I did pick wildly: because I was illustrating that different countries/localities tend to make different choices as to what to socialize, and not much evidence is needed to see that this at least appears to be true. I listed some examples, but did not make an extensive study. I suspect local politics, and the ease or difficulty (socially, culturally, politically, practically) of taking over some industry or another are the main factors in the choice of what to socialize.

It really is astounding (to me) that in the U.S. it was public transit that was socialized -- everywhere you go in the U.S., the trains and buses are all state-run -- but much else was not socialized (e.g., healthcare). Telecommunications in the U.S. was -for a few decades- regulated to the point that it was not greatly different from being state-run, but healthcare, for example, was not, and neither was automobile manufacturing, and not even electric utilities (with few exceptions, such as the TVA). And it is astounding to me that Buenos Aires didn't socialize buses -- I don't know why, but perhaps it was because lots of nationalizations that happened in Argentina targeted foreign-owned businesses (always popular targets), while the bus operators must have been politically connected.

Also, I did not even remotely "crow that the US healthcare system is superior to socialized alternatives in European countries". I do think the private sector does better than the public sector at such things, but that doesn't mean that every instance of one sector doing some thing goes accordingly -- it's a generalization. Stop putting words in my writing. It seems that you feel attacked, but I don't know why -- maybe attacking an idea you hold is like attacking you? But it shouldn't be.

BTW, regulation is a good liberal idea.

The subway never had a communications-based signalling system before, yet its functionality is waning compared to recent history. Will CBTC actually deliver the benefits you describe? Possibly not, if the system is afflicted by some unidentified other problem, i.e. falling speed limits.

I'm not across how it's been implemented on those lines but I know from experience that running mixed systems where you have some lines on CBTC that then then have to interface with older traditional signalling, you can't realize many of those time-saving benefits until you change over the rest of those lines for (among other things) safety reasons.

We never had those time-saving benefits before. Why do we now need them to recover levels of service that were previously achieved without them?

Reply on news.ycombinator.com