In what is promised to be the last part of her exegesis of my CityLab post, Lisa agrees with me about the problems with federal transit funding, but can’t admit to it, because of the slippery slope argument.

She writes:

Ok, so I’m torn here because I am usually the only urban scholar who says openly that walking, biking, and transit advocates have overstated their claims to global benefits in trying to make a case for their slice of federal dollars, and I applaud Levinson for even saying so. Being brilliant is easy for somebody like Levinson. Being brave enough to say something this politically unpopular with the vast majority of scholars in your field? That’s a lot harder, and I’m grateful for his not leaving me to be the only one critical of my field’s claims about our entitlement to crawl into the federal taxpayers’ pockets.

That said, I’m not sure I am on board. I seriously do not know what I think here.

The slippery slope argument is of course a slippery slope, since it justifies doing nothing any time, any where.

She writes in favor of demonstration grants:

The real pain comes in thinking about those places that don’t have deep pockets. Without difficult-to-justify federal capital subsidies, there is no Portland as it exists now, and while I die inside every time one of my starry-eyed students/philosopher-kings advocates for yet another slow light rail in Los Angeles “because Portland!”, federal subsidies have given the US truly important social experiments with transit, given how the feds shoveled out for BART, Portland, and DC’s metro. Nope it wasn’t particularly just or rational, but it sure has been interesting and transformative, and for the better. In concert with transit experiments in Europe, Asia, and South America, it’s mattered a lot to urban scholarship.

The problem is of course opportunity cost. Portland I believe could have done it themselves had they wanted to, if they truly believed in what they were doing. Not that I fault them for asking for federal money, I just fault the US for giving it to them. (Just as I don’t fault the person making the bribe – they’re just doing their job, just the one taking it – they’re not.) What did we not do because we gave money to Portland. If it’s fewer buses where that would have been valuable, or less food for the poor, or fewer large screen TVs for the taxpayer is an ethical question of what you do with the money, but by spending money on LRT in Portland, something else was not done. But, you want to fund 1 instance of 1 new technology for Science!, okay, but you only get 1. Science too has diminishing returns. (The Morgantown PRT for instance).

Even if there is a better way for cities to be, and certainly there is, there is no evidence that the Federal government knows this better than locals. I don’t have a problem with federal financing, as long as the loans are paid back, with interest, by the revenue from the project (or ancillary spillovers like adjacent development). But I fail to see any reason for the federal government to systematically fund capital intensive low return-on-investment, local benefitting investments.

Then there is the spatial equity question. Undoubtedly some places are poorer than others. But why should the jurisdiction be subsidized rather than the poor residents. It makes more sense to me to give grants to poor people in Mississippi than to jurisdictions in Mississippi that happen to have poor people. This would be demand-side rather than supply-side support, but that will lead to better transportation for the people as a whole, not just the lucky ones served by one expensive project, (and fewer ribbon-cuttings for their elect.).

Lisa writes:

I’m almost done with my responses to David Levinson’s important contribution via CityLab on How to Make Mass Transit Funding Sustainable Once and For All.

…

However, I’m going to say something that is going to make errrbody unhappy: transit’s assets are worth more as assets because we all know the taxpayer will buy them back if private sector managers allow things to go pear-shaped. If there is something that, over the course of its history, has been ‘too big to fail’, it is transit. From the municipal bail-outs of holding companies in the mid 20th century to the devastating strikes that occurred before then, disrupting transit service in the pre-auto world paid out well for both capital and labor. It was textbook Ralph Milliband. So we should think Uber-level values with a bail-out and buy-out guarantee–which is basically what just about all major infrastructure transfers to the private sector turn out to be given enough time, save for some examples in Asia.

So she is skeptical of private investment. I don’t blame her, I am not investing my retirement dollars in new rail infrastructure either. But there are many lessons here.

(1) An investment that produces a stable rate of return, even if low, is extremely valuable as a financial instrument for annuities and retirement plans. The Ontario Teachers’ Pension Plan now owns a share of the Channel Tunnel among other infrastructure assets.

(2) Investing in new infrastructure is a lot riskier than investing in already built infrastructure (thus the early financiers of the Channel Tunnel got wiped out twice, similarly the Dulles Greenway and many other privately funded pieces of new infrastructure that were either more expensive than expected, or built too far in advance of demand – yet the physical thing remains, so the public risk is relatively low if the public sector is not already pwned by the private sector).

(3) The case of London is instructive, and suggests some of the risks of what we in the US call Public Private Partnerships and the UK call Private Finance Initiatives, but are really just contracts. The London Underground set up was, to be frank, insane and could only have been designed by overpriced lawyers (I read somewhere legal costs of setting up the contract was £500 before even operating, and that still did not deal with the contingency that actually occurred) without a clue as to how transportation systems work. There, all construction had to occur between 11 pm and 5 am, including set-up and break-down, and the system had to be open and operational for the morning commute. So the concession-holder spent a couple of hours setting up, performed a couple of hours of construction work, and then took a couple of hours of cleaning up. It would be much better to temporarily close the line, do all the work you need, and do it in a relatively short amount of time (months instead of years). Convert the streets above the line to (temporary?) Super-BRT exclusive busway routes with tons of buses to serve the displaced demand. Of course one should do this to only one line at a time. Christian Wolmar in Down the Tube has a great discussion.

(4) Strikes are a lot less effective than before once you no longer have a monopoly provider and a monopsonistic labor union operating with a single systemwide contract.

(5) Obviously unproductive assets should be retired. I have a few in mind, but I suspect most fixed assets (aside from selected intercity rail, commuter rail and streetcar lines) in US transit are redeemable, and all the modern buses are as well, as they can be redeployed.

(6) I am not sure what public space is anymore, but New York’s Penn Station was of course privately built (and destroyed), while Grand Central was privately built (and preserved – due to public intervention), while exploiting air rights to get some additional revenue for an otherwise dying company. In fact Grand Central is still privately owned.

Lisa Schweitzer continues her hermeneutic investigation of my CityLab post: How to Make Public Transit Sustainable Once and For All, this time focusing on capital cost recovery and land use issues.

What Levinson is trying to get at here simply concerns the “eyes bigger than market” problem that plenty of cities get into over things like stadium deals. If you actually live in a world where land taxes aren’t distorted (which I don’t, so there’s that), letting utilities buy and develop around train stations makes abundant sense. Here’s the part of the equation that I think really matters: if you require local jurisdictions to provide a portion of the capital and operating subsidies for service, they then have every incentive (instead of the incentive they have now), to alter their local zoning and approvals process ahead of timeso that development around station areas can actually occur.

She thinks she has spotted an inconsistency (or mic drop), but I am not clear what it is. (I think she means my moving from voters have the right to subsidize stuff they want vs. jurisdictions should only do something that pencils out).

There is a qualitative difference between capital and operating, between rail and bus (though that’s not the important one), between stocks and flows, and between new and old.

Capital investments are new stocks while operating expenditures are continuing flows. From a public policy perspective, continuing with existing commitments (which may be an implied social contract) may be more important than making investments that bring about new commitments. Thus new commitments (such as new rail lines which have irreversibly embedded immobile capital) should only be undertaken if we believe at the outset (admittedly a forecast, which have problems) that they have cost recovery.

It is one thing to run a bus with some small subsidy for some indeterminate period of time.  That decision is reversible if it doesn’t work out. It is another to build a large capital investment with no prospect of cost recovery for a long period of time. The bus, being mobile capital, can always be rerouted, or reallocated to another service.

The large right-of-way (usually rail, but also a BRT right-of-way) cannot – and in a world with bond financing rather than pay-as-you-go infrastructure, imposes involuntary costs (capital repayment) obligations on future generations. This appeals to rail fans as a sign of “commitment” (ignoring the streetcars went away), but it is an inflexibility and unadaptability that means the “burden of proof” for making such a decision is higher.

I agree that requiring locals pay for their infrastructure might induce them to have socially better land use controls, since they may need the revenue that the higher accessibility may have created with more intense development to support the infrastructure that created the higher accessibility in the first place.

Lisa continues her deconstruction of my CityLab post in Part 3, discussing Farecards and technology.

She writes:

Most transit companies already do require smart card use, and I’m not sure what seasonal passes get anybody, even providers,that monthly or weekly passes don’t, except if you buy for longer periods of time, the transit company gets to use your money longer than if you buy week-by-week. The only thing I think might matter here are school-year passes that, if you purchase them in August, you can a better deal than if you buy month-by-month, and you don’t necessarily want a full yearly pass. It probably makes more sense, for example, for me to buy a pass like that than my yearly pass because I don’t commute much in the summer. Unless I am missing something, this is a minor point.

I dispute her point in the absence of evidence. Most transit agencies have smart cards, and some heavy rail systems require them (BART, DC Metro, MTA). My sense of most bus agencies is they don’t require them – Twin Cities MetroTransit certainly doesn’t, hence the time saving advantages of having them. Boarding times are reduced from 6 seconds to 2 seconds per customer with smart cards if I remember a recent term paper correctly.

I would go further and say we should have pre-payment via stop-based farecard reader, i.e. all significant bus stops should have arterial BRT like payment.  (This should be coupled with a reduction in the number of bus stops). Pre-payment is faster of course.

The advantage of seasonal or annual unlimited passes vs. unlimited weekly passes is seemingly minor (reduced transaction costs for the agency, but more importantly reduced mental transactions cost for customers). However, it is an important psychological difference. This gets back to my Club Transit post, where users should be thought of as members rather than one-off riders. Sure some riders (students, faculty) might save a little bit with seasonal vs. annual passes, but what you want (as an agency) is continuous automatic billing, not having to go through a weekly/monthly/annual process to sign up members.

I fully agree that there should be “one card to rule them all”, and further my membership in MetroTransit should be reciprocal at other agencies. But I suspect most people who use transit only use transit in one or two cities per year, so this would be relatively minor. The embarrassment of certain California metro areas not being able to standardize on fare cards across agencies is not widely replicated (fortunately), and obviously there should be intra-metropolitan inter-operability even if inter-metropolitan inter-operability is still a dream. Again this is a case where contracting out (to Visa or Mastercard, e.g.) might be valuable.

Lisa continues discussing my CityLab post How to Make Mass Transit Sustainable Once and For All.

In “Part 2. David Levinson’s CityLab discussion on transit: Competitive tendering,” she takes issue with whether Competitive Tendering was causal in increasing London transit ridership. Well, as we all know, nothing is provable (though many things are falsifiable), so we cannot prove causality. We can infer causality if we have a plausible causal mechanism and an appropriate time sequence. Clearly there is a time sequence. What is the causal mechanism? Competitive firms provide better quality of service than did the previous arrangement because they are rewarded for providing better service. Competitive firms have lower costs than long-entrenched public sector agencies.

Does this explain everything? Of course not (population increases, the congestion charge, increased total bus service, fuel prices, construction on the Underground also play a part).

Does it explain something? Probably. Can I show this statistically? Not right now since London is only one city and I don’t have route-by-route breakdowns of ridership and service quality before and after competitive tendering.

However London, even under “Red Ken” did not seriously consider undoing bus competition. They did undo the poorly conceived rail competition, so undoing policy was on the table. So I infer that it is working and one of the causes of ridership increases.

To be clear, the evidence is that differently structured (more monopolistic) franchises awarded in other UK cities did not see similar ridership increases, so the answer is quite complicated about how to configure to maximize consumer welfare, and experimentation is probably required. Just giving the system away is certainly not the answer. Having the franchises be of a limited duration (5-7 years, e.g.) is better than a 20-30 year franchise. This is feasible for buses where the capital is the ultimate in mobile capital. It would be much harder for a traditional utility where the infrastructure is expensive, embedded in the ground, and long-lived.

Some even more wonky papers on London Buses:

• Cantillon and Pesendorfer (2004) Chapter 22: Auctioning Bus Routes: The London Experience 
• Kennedy (1995) London Bus Tendering: An Overview
• Hensher and Wallis (2005)  Competitive Tendering as a Contracting Mechanism for Subsidising Transport: The Bus Experience