The Northeastern United States Wants to Set Tens of Billions on Fire Again
The prospect of federal funds from the Bipartisan Infrastructure Bill is getting every agency salivating with desires for outside money for both useful and useless priorities. Northeastern mainline rail, unfortunately, tilts heavily toward the useless, per a deep dive into documents by New York-area activists, for example here and here.
Amtrak is already hiring project management for Penn Station redevelopment. This is a project with no transportation value whatsoever: this is not the Gateway tunnels, which stand to double capacity across the Hudson, but rather a rebuild of Penn Station to add more tracks, which are not necessary. Amtrak’s current claim is that the cost just for renovating the existing station is $6.5 billion and that of adding tracks is $10.5 billion; the latter project has ballooned from seven tracks to 9-12 tracks, to be built on two levels.
This is complete overkill. New train stations in big cities are uncommon, but they do exist, and where tracks are tunneled, the standard is two platform tracks per approach tracks. This is how Berlin Hauptbahnhof’s deep section goes: the North-South Main Line is four tracks, and the station has eight, on four platforms. Stuttgart 21 is planned in the same way. In the best case, each of the approach track splits into two tracks and the two tracks serve the same platform. Penn Station has 21 tracks and, with the maximal post-Gateway scenario, six approach tracks on each side; therefore, extra tracks are not needed. What’s more, bundling 12 platform tracks into a project that adds just two approach tracks is pointless.
This is a combined $17 billion that Amtrak wants to spend with no benefit whatsoever; this budget by itself could build high-speed rail from Boston to Washington.
Or at least it could if any of the railroads on the Northeast Corridor were both interested and expert in high-speed rail construction. Connecticut is planning on $8-10 billion just to do track repairs aiming at cutting 25-30 minutes from the New York-New Haven trip times; as I wrote last year when these plans were first released, the reconstruction required to cut around 40 minutes and also upgrade the branches is similar in scope to ongoing renovations of Germany’s oldest and longest high-speed line, which cost 640M€ as a once in a generation project.
In addition to spending about an order of magnitude too much on a smaller project, Connecticut also thinks the New Haven Line needs a dedicated freight track. The extent of freight traffic on the line is unclear, since the consultant report‘s stated numbers are self-contradictory and look like a typo, but it looks like there are 11 trains on the line every day. With some constraints, this traffic fits in the evening off-peak without the need for nighttime operations. With no constraints, it fits on a single track at night, and because the corridor has four tracks, it’s possible to isolate one local track for freight while maintenance is done (with a track renewal machine, which US passenger railroads do not use) on the two tracks not adjacent to it. The cost of the extra freight track and the other order-of-magnitude-too-costly state of good repair elements, including about 100% extra for procurement extras (force account, contingency, etc.), is $300 million for 5.4 km.
I would counsel the federal government not to fund any of this. The costs are too high, the benefits are at best minimal and at worst worse than nothing, and the agencies in question have shown time and time again that they are incurious of best practices. There is no path forward with those agencies and their leadership staying in place; removal of senior management at the state DOTs, agencies, and Amtrak and their replacement with people with experience of executing successful mainline rail projects is necessary. Those people, moreover, are mid-level European and Asian engineers working as civil servants, and not consultants or political appointees. The role of the top political layer is to insulate those engineers from pressure by anti-modern interest groups such as petty local politicians and traditional railroaders who for whatever reasons could not just be removed.
If federal agencies are interested in building something useful with the tens of billions of BIL money, they should instead demand the same results seen in countries where the main language is not English, and staff up permanent civil service run by people with experience in those countries. Following best industry practices, $17 billion is enough to renovate the parts of the Northeast Corridor that require renovation and bypass those that require greenfield bypasses; even without Gateway, Amtrak can squeeze a 16-car train every 15 minutes, providing 4,400 seats into Penn Station in an hour, compared with around 1,700 today – and Gateway itself is doable for low single-digit billions given better planning and engineering.
This should be forwarded to whichever Republican is on the committee examining the bill
The BIL has passed…
Circulating to the House and Senate transport sub committees and associated members would still spur some discussion and examination surely?
Why would it? The democrats on the committee care that a lot of money is spent on union jobs, not any results. The Republicans already have plenty of other ammo they use to stop transit they don’t really need more.
There are a few exceptions to the above – from both parties – that actually want to get good transit built. However they are a minority.
Yeah I know, was trying to be optimistic and that some of this criticism does at least a smidgen of good somewhere.
They haven’t spent the money yet. Most of the stuff you are whining about is still in the design stage. Most of the stuff the bill will pay for is still at “Major Investment Study” or some stage of environmental. People can change their minds, and will between now and 2027.
Amtrak doesn’t want to expand track capacity at PSNY for the sake of improving Penn Station.
Rather, like Gateway itself, this is all about Amtrak’s stealthy approach to its fantasy of creating an all-new, Euro-standard, HSR facility all across the NEC at a current cost of more than $100 billion dollars.
This is recklessly irresponsible because the NEC is Amtrak’s smallest market (by transportation output and market share) and its weakest (in return on invested capital and load factor). Post-covid, it is a slowly dying business.
Amtrak should be disinvesting in the NEC, not doubling down on it.
What in the world are you talking about?
FY2019, NEC carried 18.8 million passengers, out of the entire network’s 32.5 million. This is more than half.
Smallest and weakest? That should mean the long distance trains currently operated by Amtrak.
And the idea that COVID is going to kill mid distance intercity train is as ridiculous as internet is going to kill commercial air flights.
Having the NEC spun off from Amtrak would be a good thing. For the NEC. Amtrak has managed to make “Acela” a prestige brand and overcharges. Depending on which way you view the books, the NEC makes money. The yokels in the hinterlandscould have their own states pay for the land cruises.
Now I feel bad for having procrastinated blogging about this (the premium brand issue) and why it’s a bad idea for days now. Here‘s what I’m talking about (content warning for Michael James: this post shittalks SNCF’s international behavior).
There are enough people along the NEC who like to brag about how much money they are pissing away by tossing around premium brand names that the :00 can be the super express, with the premium branding and first class, that only stops at major stations, the :07 is the one that stops at a few more stations and the :15 stops at all stations. :22 can be the one that originates in Richmond and goes to Springfield, :30 can be another super express premium branding, :37 more stops, :45 all stops and :52 goes to Saratoga Springs. And enough people along the NEC to have 10 or 12 trains an hour passing through Trenton. A lot depends on how fast a train can get from Charlotte, Toronto and Montreal. Make it fast enough the all stops gets shoved off onto the commuter express tracks between Philadelphia and New York.
It keeps the rich assholes on the premium brand train where they can compare iThingies, solid gold watches and which restaurant has the best $100 burger. And staff that tolerates dealing with Karens and Kevins. Preferably with separate lounges at the stations.
Aren’t there about a hundred other forums where Amtrak land cruise enthusiasts can do their weird “it’s all an acocunting conspiracy” thing? Why yes there are.
Simply don’t reply to thread hijaackers and trolls.
You’re both missing some key data:
1. The “NEC is profitable” myth is just that–a myth, propaganda generated by Amtrak to dupe congress. Board Chair Coscia himself said that NEC train revenues “almost covered our adjusted direct operating costs” in 2019. Key words: “almost” and “operating.” So we know the NEC loses money, even if you only include half the costs. When you add ALL the costs, specifically, infrastructure (fixed facilities), the NEC’s real annual loss is close to $2 billion a year, plus the “capital” costs (more fixed facility spending) that Amtrak budgets but lacks the cash to spend. (They even report that, described as “capital spending below budget.”) That averages another $400 million a year, plus or minus, in “deferred spending” that adds up to the current “state of good repair deficit” of $41 billion, and growing. Only at Amtrak does losing $2.4 billion a year. covered by subsidy and deferral, amount to a “profit.” George Orwell would be proud.
2. Unless you run a bus commuter agency, ridership–the number of tickets sold–is close to meaningless in measuring the performance of an intercity carrier. No airline reports its “ridership.” Rather, what matters is output, how much transport one produced, measured in annual revenue passenger MILES. Look that up yourself, but here’s a hint: the NEC isn’t even close to being Amtrak’s biggest segment in its production of annual RPMs. (Or load factor, or market share, or financial return on invested capital.) It is Amtrak’s smallest, least-productive, and costliest segment.
And even it’s “ridership” is distorted and deceptive, because three-quarters of Amtrak’s NEC riders don’t even meet the standard statistical definition of “intercity” travelers, because their trips are too short or too repetitive, and are properly classified as “commuter.” (The definition of “intercity” is “non-recurring travel over 100 miles.”) In terms of genuine intercity traffic, the inter-regional trains carry at least as many passengers as do Amtrak’s NEC trains.
First, it makes no sense to compare Amtrak’s NEC train revenues with overall NEC infrastructure plus capital costs. The NEC also serves MBTA, CTDOT, LIRR, NJT, SEPTA and MARC commuter/regional trains, as well as NS’s long-distance, heavy-haul freight trains and CSX’s local freight trains. More than 90 percent of the trains currently operating along NEC are not Amtrak trains, and they need to share the maintenance/construction costs, usually via subsidy.
Second, where the hell defines “intercity” as “non-recurring travel over 100 miles”? Are you claiming that the 113.5km-long Beijing-Tianjing Intercity Railway is actually a “commuter railroad”, serving “commuters” rather than “intercity travellers”? How ridiculous it is.
I didn’t compare Amtrak NEC train revenue to “overall NEC” anything. The comparison was between Amtrak NEC train revenue and Amtrak NEC expenditures, both direct operating and what they call “capital” cost categories. Even if you add rental revenues from the partial cost reimbursements from the various state commuter agencies that use Amtrak-owned infrastructure, Amtrak is in a very deep financial hole in the NEC, which it covers with all or most of the annual federal subsidy.plus deferred maintenance and purchasing.
Where do you think the $41 billion in deferred maintenance came from? The alleged “profits”?
Consider this: In FY’19 (i.e., pre-Covid) according to Amtrak’s fake APT cost allocator, the NEC was up $500 million, the inter-regionals were down $500 (net: zero), and the other regional trains were down $100M. Net/net: down 100. (Source: Amtrak letter to congress–it’s on the website.) So where did the rest of the $2 billion cash subsidy go? Every train is already fully paid for (and outside the NEC, on an expressly fully-allocated basis). The only place left–which APT doesn’t account for accurately (source: Amtrak)–is NEC infrastructure. Plus another $400 million budgeted but not spent, i.e., deferred.
The NEC has no profit, surplus, operating margin or any other euphemism. Just losses, totalling around $2 1/2 billion a year. And a load factor of around 50% (pre-Covid). Despite the massive subsidy, they can’t sell half the inventory (of ASMs) they produce every year, and three-quarters of what they do sell isn’t even intercity in character.
In terms of gross output, the NEC is Amtrak’s smallest operating division. In terms of federal cost per annual passenger mile of output, it is by a huge margin the worst-performing segment of the national system.
The definition of the difference between the statistical categories of “commuter” and “intercity”–for US passenger trains–came from the US Dept. of Transportation Bureau of Transportation Statistics. I don’t know what the Japanese do.
You look at the books to make them look bad. Other people look at the books to make them look good.
Just for fun, I asked Amtrak for business class, on Acela for a roundtrip, tomorrow. The most expensive option was $377. For 181 schedule miles. Neat trick they have pulled off, charging people business class fares for coach seats. They are taking delivery on trains that will let them sell approximately 40 percent more of them.
It would be really difficult to avoid tolls to drive between cities in the Northeast Corridor. You want to whine about the money the Federal Government has spent on the NEC I want to talk about the money the Federal Government has spent on I-94. Grand Coulee Dam. Rural telephone co-ops. Ethanol! And how much money limiting the SALT exemptions costs.
So we can agree then that the federal government wastes a lot of money.
It shouldn’t be news that Amtrak’s NEC is an under-appreciated component of that.
It’s a vanishingly small part of whatever the government spends. You don’t want to subsidize it, I want the Federal government to subsidize adding a lane to the New Jersey Turnpike for the substitute bus services. And the 20 billion or so it would take to tear down Madison Square Garden for the bus terminal to service them.
The comment raises two more issues: intercity buses already carry more passengers along the NEC than Amtrak does, and the annual two billion that goes into the Amtrak side of the NEC has a significant opportunity cost. Public capital is being “invested” where it produces a negative rate of return and no meaningful growth (Amtrak’s tiny share of intercity travel in the NEC, which is under 2%, is shrinking even as total demand grows); it freezes out investment in other parts of the national system that have demand that is not being met by reason of lack of capital investment. Kudos to Chuck Schumer, but everywhere else in the country pays a price in lost opportunity.
Buses are subsidized in many ways. Triple decking the George Washington Bridge and double decking the Cross Bronx Expressway, to carry them won’t be cheap.
… Maintenance should be part of operating cost. As for measuring passenger-miles versus passenger count, the key is how important they are to the market they are serving. Much more people take aircraft across the country than Amtrak. Buses aren’t taking more people between cities along the NEC than trains.
BTS data says that buses carry more passengers in the NEC than Amtrak does.
For a venture engaged in intercity passenger transport, only a metric that captures the distances passengers are carried accurately measures the volume of transportation being produced. Headcounts–called “transaction volume”–don’t do that. Ten passengers carried 100 miles each do not equal (or exceed) three passengers carried 500 miles each. Headcounts are a lazy way to measure volume in an urban transit service where fares are the same or nearly the same, but not for variable-fare intercity transport. That’s why airlines never report “ridership.”
If you are looking at data like https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6933228/ , then the bus data is lumped together with automobiles aka people driving their own cars.
The fact that loading factors on the NEC are around 50% does not mean that AMTRAK is providing to much capacity. The marginal cost of a train is low (or should be low, if it isn’t there is indeed room for improvement there). Average loading on the SBB is around 30%, and that allows them to break even.
Maybe AMTRAK should a) run more trains, and b) lower marginal costs so that they could be a major player in the market.
Can we agree that Amtrak is supposed to be an intercity, not commuter, carrier? Low load factors are intrinsic to commuter carriers where traffic is imbalanced.
But in intercity applications, your comment that load factors don’t matter is true only where capital is both free and abundant. Neither condition applies to Amtrak. The annual subsidy is cash and nominally free, yet there is a substantial cost of inefficiency imposed by purely political force, and a strong opportunity cost (because available capital is not limitless) when one inefficient use crowds out other more productive uses.
Even in a socialist system, if (hypothetically) relatively free of political force, available capital is best used where it is most productive. The goal should be to generate the most annual revenue passenger miles of output per dollar of capital input. Anything less than that is wasteful and inefficient.
In our case, we have political force driving terrible investment decisions, decisions that minimize annual RPMs per dollar invested, all masked by a great myth and fog of propaganda to the effect that Amtrak’s operation in the NEC is “large,” successful,” and even “profitable.” None of those characterizations is true.
In most years, the NEC is Amtrak’s smallest, least productive and most heavily-subsidized segment (in both absolute terms, and in terms of federal cost per annual RPM). Its tiny, trivial, market share in that market is steadily shrinking, despite the investment since 1975, in current-dollar terms, of as much as $125 billion in public capital. We have spent a great deal on the NEC and have very little to show for it.
Last I checked, Hokuriku Shinkansen in Japan (the line to Kanazawa) also have an average load factor of ~50%, with trains becoming max packed only in peak travel seasons, but it still have revenue/cost recovery ratio of over 200%.
It can be said that Antrak isn’t using its money wisely&properly, solely from the existence of something called deferred maintenance, but the execution being faulty doesn’t mean the direction is wrong. A better and more efficient way to spend the money should be sought after, not reduce or stop spending money altogether.
You want to complain about Amtrak I want to see numbers on how productive the Interstate Highway system is. A quick glance at the New Jersey Turnpike Authority’s annual report, they aren’t going to be very good. Burning gasoline on the Delaware Turnpike is especially costly. Or Essential Air Services. Or rural anything whether that’s telecom or electricity. Or Medicaid and Medicare. And a very painful chat about who pays taxes and who spends them.
Low load factors in the Swiss Intercity System exist because the system is efficient. It tries to keep it stock moving, and that means that trains are often very comfortably uncrowded. If you are efficient that is possible. The low loading factor doesn’t have anything to do with traffic imbalances, or differences between intercity and commuter traffic. In quite a few countries there isn’t even a difference between commuter and intercity.
Railways and trains are expensive. That is why you don’t let trains sit idle in sidings. You run them as much as possible. The lower average loadings and higher frequencies increase passenger comfort and utility.
That’s a convenience for Swiss train travelers, but it is extremely capital intensive. All those trainsets came at a price and SBB deployed more capital than they otherwise would have had to in order to carry the traffic they do. That reflects a political judgment made by people spending taxpayers’ money, not an economically rational decision. If Swiss taxpayers are happy with that, so be it. It is also the case that most of Switzerland benefits, whereas in our case we spend huge amounts of public capital that benefits only a tiny part of the population in only a single region of a very large country, while the needs of the entire rest of the country are largely ignored.
Of course Swiss taxpayers are happy, their trains are efficient and therefore have Europe’s highest modal split and per capita p-km while earning operating profits. They do not need Americans to tell them how to run their public transport.
A single 16 car N700 Shinkansen costs $35 million dollars as per https://en.m.wikipedia.org/wiki/N700_Series_Shinkansen, I mean it’s not that much really.
If one glances at the Federal budget, the needs of the subsidy suckers in the hinterlands aren’t being ignored.
@Andrew trains come at a price but they are needed for peak level of traffic. Like peak hour, weekend, or holiday season. One can argue that extra money shouldn’t be spent on making sure passenger can use the service during the peak and should instead alleviate or divert the demand to either other mode or other time period but that is not what the passengers want. Hence you have trains that are emptier during off-peak and off-season.
Also, Northeast US have 58 million population, compares to the US total count of 330 million. This is already close to 20% of entire US population, aka roughly the percentage of resident in metro area around Tokyo compared to the entire Japanese population. Far from being “only a tiny part of the population in only a single region of a very large country”. If number of residents in other states where proposals for intercity HSR trips have been made (California+Las Vegas: 42M, Florida: 21M, Midwest states: 69M, Texas: 29M, Southeast, specifically Georgia, SC, and NC: 26M, Cascadia: 13M), this add up to 258 million people, aka almost 80% of the entire country.
Not to mention US government also have schemes like Essential Air Service to subsidize the traveling of non-urban US residents.
I didn’t intend to go ad hominem at this point. But after going through Mr. Selden’s opinion pieces posted on Railway Age (https://www.railwayage.com/author/andrewselden/ ), I figured out some of his central arguments for investing long-distance trains rather than NEC: (for fair use)
“The long-distance, interregional division of Amtrak’s business is its largest, measured by its output of annual passenger transport, dwarfing the output of the NEC … but “ridership” measures only transactions, not output. Airlines measure performance by revenue passenger-miles and load factor … The NEC therefore is a much smaller business than the interregional one,”
“… and it is also substantially overcapitalized because Amtrak produces far more inventory there than it can sell, and far less revenue than is required to maintain its fixed assets used to render the service.”
“I have noted previously in these editorials that Amtrak’s market share for intercity passenger transport in the NEC does not exceed 1.5%, and is declining slowly. (Buses and private automobiles dominate, with a share of about 93%; bus share is growing, and, for true intercity traffic in the NEC, may already exceed Amtrak’s.)”
“The market share of the interregional trains in their respective corridors varies slightly, but appears to be centered on 5%, more than three times the market share of Amtrak’s NEC operations.”
Among them are so many flaws, just to name a few:
First of all, Mr. Selden asserted that the long-distace sector of Amtrak’s business is its “largest” – measured by annual passenger miles – and will “dwarf the output of the NEC”. However, according to Amtrak’s FY19 factsheets (https://www.railpassengers.org/site/assets/files/3433/nec.pdf https://www.railpassengers.org/site/assets/files/3435/ld.pdf ), Amtrak moved in FY19:
165 * 12361534 = 2.0 billion passenger miles on NEC, and
534 * 4505575 = 2.4 billion passenger miles on national corridors.
This is definitely not a huge difference. Nor has NEC’s bussiness shown any trends of decaying as of FY19. Also, one should be aware that Amtrak’s NEC statistics did not include any Keystone trains or Northeast Regional trains that extend to Virginia – these routes, although partially operating along the NEC, are categorized as state-supported ones.
Then, Mr. Selden claimed that the NEC is “overcapitalized” because it costs more (roughly 2.4 billions per year) to maintain .
However, this is not a fair comparison. Amtrak spends nearly zero to maintain the fixed assets of its long-distance corridors, simply because it doesn’t own any of the right-of-the-way these routes operate over. The Class Is take maintenance responsbilities. The NEC, on the other hand, is maintained by Amtrak and serves, beside Amtrak intercity trains, commuter/regional trains that are an order of magnitude larger in number.
This is like saying buses are more cost effective to operate than trains, given than bus companies do not pay for roadway construction or maintenace, and should receive more public investment. And we all know that such arguments are faulty.
And finally, Mr. Selden compared the market share of Amtrak’s NEC sector and its long-distance, and concluded that interregional trains are more comercially successful.
Again this comparison also doesn’t make any sense, even if was based on accurate statistics – which I doubt whether it was or not. Travel markets are different in size, and by economics we know that travel markets shrink with travel distances. Capturing a larger portion of a much smaller market does not translate to more successful.
Share estimates are based on data from Amtrak, AAA, US BTS and similar industry sources.
The inter-regional trains have a share in their respective markets that is roughly three times the share of the NEC trains in theirs. The aggregate market for intercity travel in the inter-regional markets is larger than in the NEC, and Amtrak consistently produces 30 to 50% more intercity transport in its inter-regional markets than in the NEC, or the other shorter corridors.
We can agree to disagree on adjectives, but one market out-producing another by 30-50% consistently over decades, on a small fraction of the capital investment, is a significant difference.
The NEC is overcapitalized for two simple reasons: One, it consumes the vast majority of available public investment, achieving over the lat 25 years a negative rate of return on invested capital and crowding out much more productive investment opportunities elsewhere; and, two, the NEC is a market where for decades supply (annual available seat miles, i.e., inventory) exceeds demand (annual revenue passenger miles) by an annual average of around 100%. Amtrak can’t sell about half the inventory it produces in the NEC–it has capital resources deployed in excess of what it productively uses. It is therefore over-capitalized in this market. That is true for any enterprise where supply consistently exceeds demand by 100%. Capital is being wasted producing unsalable inventory.
The same is true in most other short corridors Amtrak operates.
In inter-regional markets, by contrast, Amtrak is under-capitalized because its inventory on an annual basis is consistently unable to satisfy demand. No business should invest to meet demand on its peak day or week–that would be inefficient–but when demand consistently exceeds supply over half the year, and has done so consistently for decades, the business is under-capitalized.
For a dramatic empirical demonstration of what a trivial market share in the NEC means in actual practice, consider the consequences of the wreck of Train 188, which closed the NEC entirely for just under a week (at its busiest point, for Amtrak). It was hard to find any. The air shuttles weren’t suddenly swamped; Delta was advertising the availability of seats on its planes that week. And there were no reports of gridlock on I-95. When a competitor with a tiny share exits a market, pretty much nothing happens.
The BTS data you mentioned and I linked above on June 8, indicate that on the NEC corridor, air carriers have taken 15.5 million passengers, trains have taken 16.5 million, and cars + buses have takens 18.5.
How is 16.5/(15.5+16.5+18.5)=33% worse than 5% in term of market share?
If you say inter-regional trains have 3× more market share than the trains on NEC, then you are saying the trains have 3×33=100% long distance travel marjet in the US? Are all the transcontinental flights as well as cars on interstates illusion?
What is it in RPMs after you subtract the non-intercity riders from Amtrak’s customers? Headcounts don’t matter–static equipment displays generate “ridership” but zero PMs.
The stats is for “The necessary data were collected in order to be able to estimate a demand model for the different modes of transport between cities in the Northeast Corridor (NEC) of the United States.”. Read again it is “transport between cities”. So it have no “non-intercity riders” counted.
Also what are you smoking that you think “static equipment displays generate ridership”? How do people ride something that isn’t moving?
Headcounts don’t matter when assessing the production of transportation. You can even run up headcounts without producing ANY transportation. Only revenue passenger miles measures the output of an intercity transportation business.
If you want to find out how many dollars they are making you have to measure dollars.
That is impossible. Amtrak has no GAAP-compliant financial performance info at the route, train or segment level, and the GAAP-compliant corporate aggregate financial statements don’t reveal that level of info either.
The fake cost allocation numbers derived from the non-GAAP “APT” system that Amtrak flauts about are not financial statements and do not reflect the financial results of operations in any given period. Any representation or implication that they do is inherently fraudulent.
> You can even run up headcounts without producing ANY transportation
By definition this couldn’t happen as the data is for passenger traveling inter-city. This is also the data source you suggested.
If the numbers are bad, why do you cite them?
For all you Amtrak fans who like to believe the NEC propaganda, it’s even worse than it appears from the half-empty trains and the massive financial losses.
Start with two data points: Amtrak’s published LF data showing the NEC trains are half empty in any given (non-Covid) year, and the empirical observation that Amtrak’s NEC trains are, in fact, full or nearly full in two discreet submarkets: PHL-NYP and NHV-NYP.
If you reconcile those two facts, it is arithmetically impossible for Amtrak’s NEC trains over the course of any given year to be more than 27% full south of PHL and east of NHV. Yes–three-quarters empty. At an annual taxpayer cost of $2 billion. Kinetic art.
What that demonstrates is that Amtrak’s NEC operation is essentially a hyper-expensive New York City commuter service that contributes next to nothing to other regional intercity mobility need, for the simple reason that hardly anyone uses it outside of PHL-NYP and NHV-NYP where the trains cannot be more than 27% occupied; and even that overstates their use because a large percentage of customers south of PHL and east of NHV are statistically classified as commuters, not intercity riders.
The Northeast may be the country’s most under-served and poorly served intercity rail passenger market, as well as the most costly..
It may just be me but twice an hour most of the day seems a lot better service than three times a week at a bus shelter by the side of the tracks where the land cruise train stops. You cook the books one way and other people cook the books the other way. No matter how hard you stamp your feet about miles and seats, people along the Northeast Corridor are going to want to travel. What wunder alternatives do you have for travel within the Northeast Corridor?
Match capacity to demand.
Stop playing fantasy games with top speeds–there is no empirical evidence of demand elasticity in the NEC with trip time reductions of single-digit numbers of minutes.
Put the saved capital to work where it will generate much greater returns on capital investment. That is just as important a goal for public capital as it is for private capital: the most output you can get per dollar invested.
Be honest about it–if you want transit welfare for NYC, go to the FTA to get it. Meanwhile, acknowledge that each dollar of capital invested into inter-regional markets (already) earns more, often much more, transportation output per dollar spent than if that same dollar were invested into the big NYC commuter machine.
Of course the beneficiaries of lavish transit subsidy mostly paid for by other people like the resulting service. Duh. Too bad more of them don’t use what they are given.
But on a national basis, the NEC costs a mountain of money, and produces very little in exchange. And, it crowds out growth that the blockheads running Amtrak are busy running off elsewhere in the country.
The population and population density in the NE is also meaningless to this discussion because Amtrak’s penetration of that market after 50 years of failed efforts to exploit it are minuscule AND declining. That is the opposite of what rail has accomplished in the same period elsewhere in the country, in long markets and some shorter markets alike.
Now you are stamping your feet about dollars! I’m not going to feed the troll anymore.
> empirical observation
> arithmetically impossible
Who gives a flying leap about seat miles versus passengers miles? If you want to know how many dollars they are making or losing you have to measure those. You cook the book one way, other people cook the books another way.
Wheres your post about another $4 trillion being given to corporations?
Thank you for you valuable contribution
with a track renewal machine, which US passenger railroads do not use
Ask YouTube the right question, “Amtrak track laying machine”, and it finds this.
It’s 11 years old. I doubt it’s a deep fake. It’s unclear if they are laying new track, laying new ties or just “undercutting”, cleaning the gravel under the ties.
Yes, Amtrak has one (or maybe two?) and they do use it for track renewals and new track, but the commuter railroads don’t have their own and Metro North in particular seems to do track renewal very piecemeal and inefficiently.
One isn’t none. The video appears to be taken in daylight and it looks like a Regional passes through on the adjacent track.
Ask YouTube a slightly different question and it comes up with this.
It appears to be a different machine. The ConnDOT commissioner doesn’t say who owns it but the television station seems to think it’s ConnDOT’s. It’s not MetroNorth, it’s CTRail. Connecticut uses machines. Whether they own it or it belongs to a contractor doesn’t matter.
Ask Google the right question and it comes up a MetroNorth machine
so there have been at least three on or in the vicinity of the Northeast Corridor.
A few key lines in the press release from 2016 re the machine (ref: https://portal.ct.gov/DOT/CTDOT-Press-Releases/2016/GOV-MALLOY-SAYS-THE-STATE-IS-CONTINUING-TO-MAKE-PROGRESS-BUILDING-THE-HARTFORD-COMMUTER-RAIL-LINE-AS)
1. “The TCM is expected to return to Connecticut in 2017 to add another ten miles of track” >> implying that the state of Connecticut, at the time, does not own the machine and is only used for laying new track not maintenance/renewal activities.
2. “Only a handful of these machines exist in the United States.” I interpret a handful as less than five, so unlikely they are in widespread use for maintenance/renewal activities
The title of the video implies the State of Connecticut owns it. It wouldn’t make sense for Connecticut to own it, they can hire contractors, who move between different railroads, to do it. It is the kind of thing they need every few decades. it makes more sense to hire a contractor, who may or may not own it. The freight railroads use them and I don’t care who owns the machinery and who employs the people operating it.
The assertion was that passenger railroads don’t use them. They do.
I wouldn’t trust the titles of Youtube videos as a source of information.
I didn’t, the television station did. And I don’t care who owns the equipment or who employs the people operating it. I didn’t look for reporting marks on the train emblazoned with “Amtrak Track Laying Machine” on the locomotive. The assertion was that passenger railroads don’t use them. They do.
Metro North is an “exception” in the industry. Not only did they make no use of TLMs, they’re also replacing concrete ties with wooden ones, by claiming that “concrete ties are not reliable”.
The MBTA went from concrete to wood on the Old Colony Lines too; their concrete installation had been defective and they generalized this to thinking concrete is inferior.
I’d like what drugs they’ve been taking. They sound like fun!
5 minutes of attempting to figure out what conspiracy theory is being promulgated, it appears a vendor, or perhaps a few, , in the 1990s, formulated the concrete incorrectly and those ties fail prematurely. Causes were still being investigated many many years ago, when they rehab’d the Old Colony lines. They may have changed their minds since then.
The company has been trying to replace it. I googled a bit earlier and lost interest, but you will find info on amtrak.com and also in the Federal Register since they needed an exemption from “Buy America” rules because there isn’t anyone in America to Buy one from. IIRC, the original one, perhaps the one in the image, had been idle for years when David Gunn took over the company and he was pissed about that fact.
Let me also add that *Amtrak* only owns the track west of NYP, and east across the New York Connecting, to New Rochelle, and then, I think, nothing until just past New Haven. They don’t own the track between the RI line and Boston either (the Commonwealth of Massachusetts does but lets Amtrak run it as if they own it). The other bits are CDOT or NY’s MTA (through MN). That segment of the NEC between NYP and NHV (and the line up from NYG and all of the CDOT stuff) seems to be the source of most of the problems on the NEC. Congress should make CT and NY offers they can’t refuse to have the National Railroad Passenger Corp buy it from them … Not that Amtrak is a hotbed of state-of-the-art best practices but they *have to be* better than Metro North and CDOT …
Sounds good to me. The high speed trains can go across Long Island Sound at New Haven and use the very very straight ROW on Long Island. Where there are a lot more people. Who would be able to get to New England without going through the Bronx. Bridgeport, Stamford and New Rochelle can have a once or twice an hour “Shore Line” train. Or once an hour Inland and once an hour Shore Line.
On the other hand, we could choose not to build a random 13 mile bridge and build something useful instead.
It’s not a random bridge. It moves more people more places than relentlessly conforming to decisions made in the 1840s. Though some of the parts on Long Island reflect decisions made in the 1830s.
Nassau/Suffolk->Greater Boston in no way represents more than 5% of NEC demand. Just not worth prioritizing.
In nice round numbers there are 8 million people on Long Island and 6 million in greater greater Boston.
I don’t know what Boston has to do with the bridge. People in Boston who want to get to New York or beyond won’t care a whole lot about what happens in the rest of New England. People in New England not-Fairfield-County-Connecticut won’t care a whole lot how the train gets from their part of New England to New York City and beyond either. There are 8 million people on Long Island and under a million in Fairfield County Connecticut.
Or are you saying the trains shouldn’t stop in Providence because it’s too small? There are more people in Brooklyn than there are in metro Providence. Or more people in Queens. Almost as many people in Brooklyn as there are in metro Baltimore. Should the trains discontinue stopping in Baltimore and BWI Airport too?
You absolutely should not count Queens and Brooklyn. This does not improve journeys to New England for anybody west of Jamaica. Specifying Boston is key because nobody going from Nassau/Suffolk to points south and west of New York (a much larger market) don’t benefit from the bridge. Boston is the end of the line.
Anyway, Nassau+Suffolk have 2.8 million people. Fairfield+Westchester have 2.0 million people. It is not worth building a 13 mile bridge to improve journey times to *one* city north of NY for net 800,000 people.
If I’m on the Hudson line and I want to go to Philadelphia I’m not going to take multiple buses to get to New Rochelle, or call a limo. I’m going to walk to the Metro North Station, take Metro North to Yonkers and change to an Empire Service Amtrak train there. And someday just take the Metro North train to Penn Station. If I’m on the Harlem line I’m screwed. There aren’t very many people in Westchester who find New Rochelle enticing. Or Stamford. They would still be able to take the two or three times an hour train to Boston or DC from New Rochelle or Stamford.
People along Queens Blvd can get to Jamaica real easy. Almost all of Brooklyn can get to Atlantic Ave, where they could catch the twice an hour train to Boston or DC. There are other cities in New England. New Haven, for instance where the New England end of the bridge could be. Or Providence for instance. People in Providence aren’t stupid, they aren’t going to take an MBTA train to Boston and get on a Amtrak train …………….that stops in Providence.. Providence is part of the 6 million people in greater greater Boston. People in New England, won’t care where the train stops other places in New England if they want to get to New York or beyond. I suspect a 13 mile bridge is cheaper to build than a 26 mile tunnel from Herald Square to Westchester Airport. Which isn’t New Rochelle or Stamford and doesn’t connect very well to New Haven Line trains. Or Harlem line trains or Hudson line trains.
It would have similar costs to the Chesapeake Bay Bridge Tunnel. There are no naval bases west of New London, Connecticut. The Navy isn’t going to insist on artificial islands and tunnels. There are never going to be Post Panamax container ports. supertanker oil terminals, liquefied natural gas docks. The Coast Guard can tell you where the channels are and how high the bridge has to be. The people with sailboats in the mighty port of Cos Cob can probably tell you where the channels are and how high the bridge needs to be. It serves more people to more places and is likely cheaper to build.
I don’t think that many railways (maybe none at all) in Europe actually own track laying machines. These machines are typically owned by contractors. There are quite a few big building firms that do a lot of railroad construction and maintenance. They typically own the machines, and supporting stock.
The contractors may not own it, they may be leasing it from someone who is in the business of leasing construction equipment. I’m a lowly bookkeeper, an accountant can explain why that may make sense. That’s digressing far from whether or not they build/maintain track with machines or not. They do. It doesn’t really matter who owns it or who employs the people operating it. And the people standing around watching them do it might be employed by someone else, that isn’t the railroad either.
This guy running for Congress sounds like he might usher in a new wind in regards to infrastructure spending if elected. Any thoughts, Alon?
After Andy Byford’s very public ejection at the hands of Andrew Cuomo, “hire the best and brightest from Europe and Asia” seems like a wildly ambitious goal. The lesson was broadcast loud and clear: if you’re an upcoming, results-oriented technocrat, you should stay as far away from the USA and especially New York as possible, because the existing political power brokers here do not want you and will pay no price at all for mistreating you.
The idea that any State Agency is going to go out and hire Europeans or Asian staff to replace their staff is non sensical. Speaking from experience visa issues could be problematic, considering that some agencies now require designers bidding for work to indicate how much is going to be offshored, you know to European and Asian office of multinational designers, and an increasing “Buy America” mentality even for design…..As referenced above the Andy Byford saga is not exactly going to help either. I know some non US citizen staff who work for various agencies, LA Metro among them but these are folk who originally came over with private firms and once they got Green Cards or Citizenship and their visa was no longer tied to employment, moved to an agency position. I’m also not sure why you continually write off consultants experience. Many of the big consultants work in multiple countries with experience of some of what you consider best practice, so surely they need to be part of the solution to get that cross pollination. Also trying to overcome the, “where has this been done before in the US mentality” is a seriously challenging issue…….. The biggest challenge though is not the engineering, but how projects are financed and procured in the US plus the legislative differences between States and who has jurisdiction. 5 year Capital Plans, a recent conversion to Design Build which is abysmally managed in the US and which most transportation agencies see merely a way to shift risk to the Contractor while being able to meddle in the design process leading to unnecessarily adversarial relationships, the endless planning process that has to be navigated to even get a project to the street for procurement……. I could go on, but all of this cost time and money to overcome.
As for the Gateway project, given its already been simplified to remote the vanity terminal on 34th St and its now twin tunnels from NJ to Penn I’m not sure what improvements in planning and engineering you think are possible to reduce the cost. The alignment is challenging especially form a vertical perspective due to the depth of the Hudson, the zero blow count material that exists in the bed of the river that would need to mined through and the buoyancy issues this creates for the tunnels, plus the need to tie in to the existing box beneath the Hudson Yards that was previously constructed for the project. There’s really not many alternates that could be considered unless you radically change the alignment and avoid Secaucus in its entirety and come up somewhere short of Newark Penn, given that the area from Tonelle Avenue to basically the Turnpike over bridge at Kearny Interlocking is all swamp/wetlands and to bring the tunnels up in that area to get on to the new higher level Portal Bridge would be almost impossible. What could be done is to get the Palisades tunnel awarded and started, after all the current version is the same as the previous one and was taken to around 80% design completion before the ARC project was cancelled. The shaft in Hoboken is needed to switch from the hard rock Palisades to the soft ground Hudson tunnels, completely different ground requiring completely different TBM’s, technology and lining designs. The ground under the Hudson is so bad that there is no way you can use Cross Passages for emergency egress so the emergency egress passage way has to be accommodated within the envelope of the tunnel which means the diameter increases leading to to an increased buoyancy problem which again constrains the vertical alignment to get enough confinement to counteract this. This means that you will be highly unlikely to be able to drag the soft ground TBM’s through the completed hard rock tunnels as they will likely be too big, hence requiring a construction shaft in Hoboken which will probably also be needed from a schedule perspective. Now on the NY side there is the riverfront diaphragm that you have to get through. This dates from the 1800’s I think and is made of timber. Its going to be a absolute nightmare to get through and will require significant marine works to install a large concrete slab over the tunnel alignment in the river to deal with the buoyancy issues. So again, what engineering decisions could be changed to provide a cheaper design?
I’ve met a great number of of the consultants doing rail transporation “planning” in California.
With the exception of only one or perhaps two people I can think of over 20 years, they were all, to a man and woman, incurious, insular, stupid and ignorant. In the few cases when they were not totally ignorant of anything outside the USA, they were actively uninterested in, disbelieving the existence of, or dismissive of any relevance of it.
These people hire their peers, to continue the culture of massively overcompensated systemic failure.
Yeah, and the “engineering” is piss-poor, sub-moronic, purely cost-inflating rent-seeking as well.
I mean look at the outcomes. Terrible terrible terrible shitty projects, undertaken at out-of-control cost, delivered years late. This doesn’t happen by accident.
But hey, nice that you think you can build the wrong shitty tunnels in the wrong places for the wrong alteratives of the wrong projects for the wrong reasons with the wrong financing and still come out thinking it’s all cool.
The real world outcomes are really bad, and that’s all that matters.
I see, well judged comments there. Now fuck off back under your rock.
Your discussion of the engineering challenges of a tunnel under the Hudson are interesting – thanks for your contribution. But I thought these challenges were already fully addressed during the old ARC engineering analysis.
1. Use of Ground Freezing and Manual/SEM excavation of cross passages under the Hudson
2. Use of Sealed EPB TBM for cutting across the Sandstone > Clay layer in the NJ > NY direction
3. Pre-removal of timber piles or use of a cofferdam to ‘sterilize’ the edges of the river prior to tunnel boring
If it could be accomplished in the early 20th century it can’t be that difficult today…
Errm no. The cross passages were removed as do you have any idea just how expensive and difficult ground freezing is under a river. When I worked on the Great Belt project in Denmark we used freezing, but the use of Freon for freezing is no longer allowed. Brine freezing would be slow and the sheer risk involved is too great.
For the TBM’s it will be a hard rocks main beam gripper for the Palisades section and some form of pressurized face TBM for the Hudson tunnel, depending on what the contractor selects either slurry, EPB or Variable Density. I doubt an EPB can tackle mixed face as easily as a slurry TBM, indeed for the Queens soft grounds tunnels for ESA slurry TBM’s were selected, although Variable Density machines are becoming more popular.
Sure you can build a cofferdam and remove the piles, but under ARC the alignment was deeper and that does not impact the need to apparently place a massive concrete slab in the river to confine the tunnel on the approach.,
And yes it was achieved using compressed air technology and multiple deaths in the early 20th C, how many deaths would be acceptable to you in 2022?
Freon is Dupont’s brand name for quite a few different chemicals or mixtures of chemicals, typically used for refrigerants but they can be used for other things. Why can’t they use refrigerants?
Years ago they would excavate river sludge and dump it in the ocean. Nobody cared about what was in the river sludge. Or dumping it in the ocean. And no one cared about how it might affect endangered species because no one had quantified that yet. Dredging stuff out of the river is much more complicated these days. At least the water doesn’t have raw sewage from most of Manhattan and large swaths of Brooklyn in it any more.
Very, the U5 extension here cost around $350m/km in part because they had to freeze the ground under the Spree to build Museuminsel.
In Stockholm, too, I’ve been told the underwater premium is substantial and that’s why Citybanan cost as much as it did (around $300m/km) even though the stations were not by themselves very expensive.
I mean, is the Hudson crossing literally harder to build than any other tunnel in the developed world? That seems unlikely, and the fact that all US transit projects are way overpriced further suggests the engineering challenges aren’t the issue. If the tunnel was simply more expensive- say German expensive, or even Dutch or Japanese expensive- cost control might not be quite as high a priority. But at a certain point a transit project doesn’t have enough value to justify itself and the money would be better spent elsewhere.
Most of the cost isn’t the tunnel itself, it’s all the stuff to make the tunnel useful. There is a tunnel that has been sitting under the East River, unused, for 50 years. Making that one useful is costing a lot of money.
No its not and in fact we are currently involved on two other underwater projects in the US, Parallels Thimble Shoal and Hampton Roads, larger diameter road tunnels with similar cover and buoyancy issues. The excavation and lining is not that difficult although its certainly challenging with the ground issues under the Hudson. Its what goes inside the tunnel that’s the problem. I mean look at the Channel Tunnel, mining was finished pretty much on time, the systems fit out took forever, same for Crossrail.
He said “the fact that all US transit projects are way overpriced further suggests the engineering challenges aren’t the issue.” and the two comparable projects you mention are also US ones, i.e. also overpriced.
The only other project you mention is the Chunnel, ironic because Alon is constantly saying that “US transit planners almost never think to compare to other countries, and on the rare occasion that they do, the other country is usually the UK”.
There is a whole world there outside the US/Canada/UK, generally accomplishing the same or better things at much lower prices, and it would be great if US consultants could show the least bit of awareness of it.
“a recent conversion to Design Build which is abysmally managed in the US and which most transportation agencies see merely a way to shift risk to the Contractor while being able to meddle in the design process ”
That sounds like the fault of the agencies, not the politicians. So likely a place where foreign expertise could help.
I agree it must be the agency staff at fault.
I know districts controlled by one political party or another often stagnate in a democracy but if it was the politicians fault someone could come in and promise to build infrastructure for 1/2 the cost and they’d win a landslide.
It’s the fault of the politicians who believe the myth that the all powerful unseen hand of the free market does all things better. And force the agencies to do it.
The big problems the public sector has especially in places of high housing cost is that it isn’t prepared to pay its top people enough compared to the private sector.
The other issues is that it has far too much bureaucracy and far too little curiousity about new ways of doing things. Whether that’s learning from other countries or adopting Agile/Lean/Six Sigma.
Haven’t you previously state part of the cost here is due to Amtrak/NJT’s very conservative choice in ruling grade for these tunnels?