Runaway Train?

Robert Moses, the titanic “power broker” who is responsible for much that is wrong (and some that is right) in the planning of modern New York, had an infamous dictum: once you’ve turned the first shovelful of dirt, they’ll never make you stop building.

Big infrastructure projects might be killed on the drawing board, but once under construction they will almost never be canceled, even by those who bitterly oppose them. No matter how high the cost overruns are or how disastrous the engineering mishaps may be (think the Big Dig), no elected official wants to abandon a half-built expressway, leaving a rusting blot on the landscape which screams testimony to the incompetence of government.

Moreover, thanks to a quirk of our psychology we humans are very loath to walk away from “sunk costs” no matter how much it is in our interest to do so. Economists counsel that past effort should not dictate future action: what’s done is done and all that matters is whether future benefits outweigh future costs; no point throwing good money after bad.

But in practice we find it very difficult to admit we were wrong, and we often feel that sunk costs should be left on the balance sheet when it comes to the calculus that determines our future actions. Indeed, great difficulties, which may indicate we should abandon an undertaking, may ironically make us more, not less, inclined to stick to our questionable course.

Government officials are not necessarily more immune to this way of thinking than the rest of us. So once they have started spending money on a project they find it very hard to stop, no matter how disastrously it may be proceeding or how dim the benefits are starting to look.

This is why we should consider the administration’s high-speed rail (HSR) network carefully before construction commences, as I advocated here. The $12 billion that has already been allocated may just be the tip of the iceberg; once the dirt begins to fly, the public may be irreversibly committed to the capital and operational support of this program in perpetuity, no matter how much construction costs may spiral out of control or how disappointing ridership might be.

HSR backers may consider this a happy circumstance (and hey, I think when all is said and done, the Big Dig may be looked at as a great project). But at the very least, the HSR program deserves more scrutiny than it is receiving, buried as it is under weightier issues like health care and the state of the economy.

That’s why HSR requires a really rigorous cost-benefit analysis. Doing one on such a large and multifaceted project calls for a very big brain, and fortunately Edward Glaeser, over at The Times‘s Economix blog, is providing one. Glaeser, an economist at Harvard, is perhaps the closest the field of urbanism has to a rock star. He’s doing a multi-part series: here’s the link to his introductory piece, and here’s the link to his first substantial essay, on direct costs and benefits. I’ll let you know when the subsequent posts come up.


Regarding "No elected official wants to abandon a half-built expressway, leaving a rusting blot on the landscape which screams testimony to the incompetence of government..."--I guess you've never been to New Haven.


@#1 - Or Baltimore. I-395 is pretty much an abandoned half-built expressway.


How about the hole they're trying to fill in lower Manhattan with millions of sq ft of office space that nobody needs? Sorta like the old Trade Towers before the state filled them with their own.


The first project that sprung to mind for me is the Superconducting Super Collider, which was shut down after 14 miles worth of tunnel, and $2 billion cost.


DC too, i live a few blocks from a pretty worthless stretch of 395. Thankfully the project was stopped short and didnt destroy any more of our city. Instead the money went to metro and our city is far better for it.

Id rather have a half-built system of interstate HSR than a fully built out system of freeways running thorough the downtowns of our major cities. While costs of HSR are high, the positive externalities and development potential are high too. The marginal returns of HSR will be enormous as long as routes are well designed and can compete effectively with air travel and bus travel.

See what has happened to the short haul air market in Europe after HSR became prevalent, and how much more successful their cities are than ours.


The only section of the country that needs HSR is Washington to Boston. The Northeast Corridor is the only moneymaker for Amtrak, yet, the $ specifically limits $ going to the NEC, because some senators in Nebraska want their train too. Poor planning. (Possibly California could use HSR too)

Eric M. Jones

In Mexico, I have been told, you don't have to pay property tax on your house until construction is finished. It doesn't take a genius to figure out what happens--houses never get finished.

Surely the most amazing piece of unfinished public works is in the Megalith of Baalbek. See:
This single stone weighed 1200 tons and is sitting abandoned. Nobody has a clue why, but the mechanics of moving such things was well within the capability of the Romans.


Yes, those of us who would like to get from the Bay Area to LA without driving would love HSR in California!

But I meant to point out that San Francisco is a case where for years there was a half-completed freeway downtown, along the Embarcadero (i.e., near Fishermen's Wharf and other tourist goodies). It was taken down and the place looks a lot better. I don't know what the cost-benefit analysis was, but it's as if the idiots who approved the New Haven freeways had had the guts to stop and take down the structures they were putting in to divide the city into an unwalkable maze of depressed slums.

Phil O.

First, I think the Big Dig was worth it. Just walk along the new Greenway and tell me you liked the elevated I-93 there more... or drive through Boston to the airport from almost anywhere. Or go sit on 93-S coming into Boston for hours as four lanes merged into two, when what you really wanted to do was get to Storrow Drive...

Second, Manhattan's famed Second Avenue subway was started, but never finished, either. They never stop threatening to start building again, but it's 30+ years and counting...


Mr. Morris,

How do you respond to the above comments offering counter evidence to your claim?


Excuse me for being rude here, but Eric Morris's posts tend to be little more than poorly argued rants. He seldom delves into any actual analysis. For instance in this post the gist of it is "HSR is bad, so stop it before people start building it because once they start they don't stop because they are idiots and don't understand sunk costs." He doesn't cite a shred of evidence that this assertion about human behavior is correct other than a quote that was made in jest by a long dead urban planner--hardly compelling.

I also think Morris misunderstands what sunk costs are. A sunk cost is one that cannot be recovered, so that behaving as if it can be is irrational. As an example: we value a bridge at $1 billion and are told it will cost $900 million, so we pass a bond worth $900 M to finance it. After we've spent $900 we learn that another we require another $200 million to finish the project. The SUNK COST IS THE $900 M! At that point we have two choices:
1. Spend $200 M and get a bridge worth $1 billion
2. Spend nothing and get nothing
Being rational people, of course we pick option #1. Had we known from the start that the bridge would actually cost $1.1 Billion, we wouldn't have spent the first $900 million, but since we didn't, it is now rational to get a $1 billion bridge for only $200 M.

Morris seems to for some reason believe the polar opposite, that for some reason we should pick option #2 and leave with the bridge 85% complete.


John Doe

The fallacies of the "rock star" of urbanism were easily destructed by the commentators.

Phil H

Eric, we get it. You don't like HSR. Please, either pick another topic or put a better argument than this.


Well, this comment is really about Edward Glaeser's essay. What I do not understand is if Glaeser is really a "rock star" of urbanism, why isn't he making any urbanist arguments? His second essay looks at the economics of high speed rail as compared to ... taking an airplane??
One of the chief arguments of high speed rail is that it provides a potentially fossil-fuel free method of taking cars off the highway. In other words, folks who would normally drive see the possibility of taking the HP rail as an alternative, perhaps because flying is too expensive. HP rail then becomes an urbanist tool to do the following, in this order:
1. Take cars off the highway
2. Provide a "reverse incentive" for sprawl. Businesses centralize near the train stations
3. Compete with the airline carriers over "short haul" routes
4. Get folks there faster
Glaeser only looks at the last two items, and does not examine the first 2 less tangible items. Perhaps he intends to in his next column.
It we look at his analysis in Texas with the last 2 items in mind, then it is clear that Texas is a uniquely bad example to use for comparison. Perhaps it's true that land may be cheaper, leading to lower per mile costs, but HP rail may introduce a profound culture change there, potentially dooming the execise to failure. Why not instead look at the existing example in California, or find another example (say Philadelphia to Pittsburgh) in which the population is "primed" for such a shift?
The French are the first ones to admit that the TGV HP rail system was ridiculously expensive when the tracks were first laid down, but who would say now that the investment was not worth it? Is there any doubt now that by our short-sidedness of relying on specious economic arguments, that the USA is worse off today? We cannot ignore costs when we build these tracks, but cost alone should not drive these decisions. After all, if it is not the role of Government to think about the our long-terms interests and direction as a society, then who's role is it?



I haven't seen any good arguments against the HSR so far that takes into account the full effect of HSR- That is, once you have HSR connecting cities, people can take more forms of public transportation, and that is where much of the savings comes from. People traveling between cities without their cars will use more buses at their location (and home) which are vastly more efficient than cars.

Also, HSR is quite efficient compared to planes in many cases, you just need good ridership. If an effective, or even just adequate, public transportation system consisting of buses and/or light rail is at both ends of the HSR, many people will choose it over driving. Admittedly the German Rail (DB) may be biased in their estimates of environmental savings, but their website allows you to compare traveling by train to car/plane on most routes within Germany. There have been a few times where the car came out on top, generally when the connection between two cities is somewhat ineffective, but most of the time the train wins by a large margin. I suggest opponents of HSR check it out at - Try punching in Berlin to Munich, roughly the equivalent in distance of Boston to Baltimore and look at how much carbon is saved!

(And do forget the comfort- Relaxing on an HSR is much nicer than dealing with the traffic, especially the traffic of I-95...)



Wow, if we never had the foresight to build big, think of all of things we wouldn't have...

Interstate Highway System
NYC Subways ( or any mass transit system for that matter)
Hoover Dam
Golden Gate Bridge
Tennessee Valley Authority..... and so on

Just because a project is controversial doesn't make it wrong. Long term issues that develop as a result can be mitigated down the road.


To Steven's point on sunk costs. It assumes that the 1B bridge, with 900M in sunk cost will be completed with an additonal 200M. It overlooks the possibility and likelihood that several additional 200M might be needed in order to complete the bridge valued at 1B. There is a classic economics classroom experiment in which two teams compete in an auction for a $1 bill. The nuance is that any bid must be paid whether the team wins the auction or not. When I participated, the bidding started at $0.01 and ended (by the instructor) somewhere at $8.00. The underlying psychology is that "we've already sunk $X, if we don't win, it will be a complete loss, better to continue bidding in order to have something to show for it."


You cannot do a straight cost-benefit analysis on public works projects, especially transportation. Transit projects almost never cover their costs without significant public subsidy. (Highways are the same, so conservatives can stop whining about public transportation).

If all the construction costs for HSR is on one side of the balance sheet, then the other side has to have not just fare revenue and subsidy revenues, but also intangible public benefits like reduced congestion, reduced carbon emissions, less expensive long-distance intercity transit, easier to implement a more sustainable land use planning. Plus there are reduced gas, insurance, repair, and registration fees to consumers who would no longer have to purchase as many cars per household.

Then, supplement that balance sheet with a comparison of marginal operating cost/per passenger-mile with marginal benefit per passenger mile. Do that for HSR, autos, buses, and planes, and I'll bet you'll see a completely different picture.



Eric, so you don't like HSR? What will you say if Glaeser's final report winds up supporting it?;)

As someone who knows more about urbanism than economics I can say that there are two distinct hopes for HSR: 1) that it will move people around from city to city more effectively, enjoyably, economically, and environmentally sustainably than people move around the country with today's current transportation system; 2) that HSR rail can be a major infrastructural element for development that is more dense (and thus less taxing per person other types of infrastructure and resources) as well as less automobile depandant than we have today.

Thus, there is a deeper economic gain that HSR suporters, like me, seek beyond the transactional difference compared to flying or driving, etc. I would be very interested to see an economic model that could study greater societal gains to having HSR. For instance, what has the economic value been to Japan to have HSR? Are you up to it? Is Glaeser? ( If I my hunch that HSR would be a good investment when everything is counted, I would give up my hopes for it, but not until then.)

I would like to see this debate out there more, so thanks for bringing it up (even if we disagree).



An old woodshop teacher once taught me, "Measure twice, cut once." It was as good a life lesson as I've ever gotten from a four-fingered man.

It seems to me this is all Morris is suggesting. So before we nail him up to the side of an unfinished highway to be devoured by crows, why not at least consider his point?

Besides, on the off chance he is right, who wants to have to listen to him say "I told you so" (assuming any of us live long enough to see the HSR completed).