Hacker News Reader

Comment by landryraccoon | Hacker News Reader

The 1976 BART cost $10B in today's dollars to construct. If you were alive in 1976 would you have said it was a boondoggle and the project should have been killed?

IMO people consistently underestimate how much public infrastructure is worth. The US highway system was sold at a cost of $27B [1] but cost over $100B and has cost nearly $500B today [2].

Does that mean the US Highway system wasn't worth it? Of course it was, the public just doesn't understand how much it had to cost. The Iraq war cost trillions for zero benefit for future generations, even if 50% of the cost of HSR is waste it's still a better deal by far.

[1]https://en.wikipedia.org/wiki/Federal_Aid_Highway_Act_of_195...

[2] https://en.wikipedia.org/wiki/Interstate_Highway_System

View on HN

> The 1976 BART cost $10B in today's dollars to construct. If you were alive in 1976 would you have said it was a boondoggle and the project should have been killed?

And what would BART cost today to build? $50 billion? Would it be worthwhile at that price? The problem is that costs for public infrastructure are ballooning much faster than inflation. That puts more pressure on public infrastructure projects to produce benefits that justify the investment. I fully expect many public infrastructure expenditures today won't ever justify themselves, for example DC's silver line.

At $50B, yes. BART will exist in some form for at least a century. It is so hard to understand how valuable it is to make something that so many people use every day for that long but it is really valuable.

How valuable? $50 billion is enough to create a fund that pays out $2 billion annually in perpetuity, or about $15 for every BART ride. That’s not counting ongoing operations and maintenance. Does BART generate $15 in positive value per ride above and beyond ticket cost, relative to driving? I’m skeptical.

$3 I’ll believe. Which is why the constant factors kill us relative to Europe. It costs us 3-7x to build the same infrastructure as Europe. That pushes the required value per ride that needs to be generated to unrealistic levels.

You haven't taken into account how much the bay area would have needed to spend on bridges and freeways if BART hadn't been built.

https://www.bart.gov/sites/default/files/docs/BARTfactsheet_... says (as of 2017) 70k people go through the transbay tube in each direction during peak commute hours.

http://www.dot.ca.gov/hq/esc/tollbridge/SFOBB/Sfobbfacts.htm... isn't dated but says that 270k vehicles cross the Bay Bridge every day.

These numbers aren't directly comparable because they're not measuring the same thing, but taking things as pessimistically as possible, let's say those 270k vehicle crossings all happen during the same time period as the BART numbers. If all those commuters crossed the Bay Bridge by themselves in a vehicle (because the overwhelming majority of car commutes are one person per car), that would be (270k + 70k) / 270k or a 26% increase in vehicles, when it's already ludicrously packed every single day. The bridge would need to have been widened by now (or another bridge built entirely, with feeder freeway interchanges on both sides).

I'm not ignoring it. I'm computing the per-trip capital cost for BART assuming today's inflated construction costs, and asking if the reduced congestion/need for road infrastructure would still justify that investment given today's costs.

I don't think you realize how completely bonkers the cost of mass-transit development has gotten. New York is up to $1.5-3.5 billion per mile of subway. The new Bay Bridge, massively over-budget at $6.4 billion, represents just a few miles of subway (BART has almost 30 miles).

  70k people go through the transbay tube... 270k vehicles cross the Bay Bridge every day

In other words, taking your figures painfully literally, if those vehicles carpooled just one additional passenger each on average, you'd have 4 times the added throughput that BART provides.

But this will never happen as long as we give HOV lane access to cars other than those carrying 2+ licensed drivers... for political reasons.

Let's take your own reasoning and let's apply it and see where the numbers fall.

The Bart cost was $1.586 billion (according to Wikipedia) or around ~$10 billion of today's dollars.

Let say that back then we had put that money ($1.586 billion) into an annuity in perpetuity and given the money each years to the riders so they will find alternative means of transport.

So if instead of building bart back in the 70's we had put the money into a perpetual annuity, we would be receiving the sum of $63.5 millions a year from the annuity (your own numbers seem to indicate a 4% yield). That is less than 50 cents per ride (according to Wikipedia Bart gives 129 million rides a year).

I don't think the cost is insignificant compared with the level of economic activity that the 400.0000 daily riders generate for SF and the Bay Area in general.

And now imagine 20 years from now how much $63.5 millions a year will buy you compared with the benefits of mass transportation.

Your math is wrong. You can (at least historically) get a 4% annuity that rises with inflation in perpetuity.

Where can I get a save investment that pays a 4% annuity that rises with inflation?

I'm genuinely curious. My father is retired and has some savings and that would seem like the ideal investment for him.

For things like endowments, the general assumption is that you can get a 4% yield that keeps pace with inflation, without touching principal. So you'd be up to $300 million per year today. That's about $2.50 per BART ride of $750 per year per BART rider, which is a believable number.

The problem that you're ignoring, however, is that the relative cost of building infrastructure today is far higher than it was in the 1970s. If you built BART today, it wouldn't cost $10 billion in today's dollars. NYC's recent projects are running $1.5-3.5 billion per mile of subway track. BART has 28 such miles. That's $42-100 billion, not including the other 70 miles in the system.

So what if BART had cost $8 billion ($50 billion in today's dollars) instead of $1.586 ($10 billion)? Instead of having to great $2.50 of value per ride to justify the investment, you're up to at least $12.50. Or $3,750 per daily BART rider. Is each BART trip generating that much value compared to doing the same trip in a car?

  That is less than 50 cents per ride

... if you ignore all the years' operational costs, personnel costs (including benefits and pensions), debt service, etc.... which is the lion's share of the costs.

  That’s not counting ongoing operations and maintenance.

Or debt service.

  you have said it was a boondoggle

It was in the sense that it was obsolete, proprietary technology, excessively priced, and badly equipped. It's implementation favored only western Alameda county and SF at the expense of everybody else. That's why San Mateo County and Santa Clara County bailed. Contra Costa would have, too, if they had any idea what poor service they would get in proportion to their contributions from the beginning.

Didn’t Santa Clara voters just approve a $6B initiative to support Bart service? Why do you think they wouldn’t want to be part of the system if they just voted to be part of it?

http://www.vta.org/measure-b-2016

> IMO people consistently underestimate how much public infrastructure is worth.

I think you mean “politicians consistently lie about how much public infrastructure projects will cost.”

Transport hubs encourage development. In the UK look at all the satellite industries around Heathrow and Gatwick. Look at the offices being built close to Paddington and Kings Cross stations in London, or Reading.

The sensible thing would be for the city to buy up land round the stations, then build offices and lease them. The income would offset the costs.

  Transport hubs encourage development.

In the Bay Area, compare the degree of development along the BART tracks with development well away from BART (e.g. San Mateo county, Silicon Valley, etc.)

Check out the BART tracks from Oakland to Fremont especially.

  The US highway system was sold at a cost of $27B [1] but cost over $100B and has cost nearly $500B today [2]

This is disingenuous.

First of all, the original bill estimate was just for the first 41,000 miles.

Secondly, the $27B is in 1956 dollars, while the "$500B" is in 2006 dollars. BIG difference. $27B in 1956 dollars is roughly $248 billion now, and, again, the scope of the interstate system expanded broadly.

(I'd bet that the modern figure includes both the Interstate system and the U.S. highway system, which lacks meaningful usage distinction now from the designated "Interstate" routes.)

Reply on news.ycombinator.com