Keep It Simple, Stupid

I almost made this response on the twitter but thought it should be more permanent.

Trying to figure out where to put a rail line in a city where you have lots of unmet transit demand and an inadequate funding stream to do everything you want to do? IE, you live in the real world?

PUT YOUR RAIL LINE WHERE IT REQUIRES THE LEAST POSSIBLE OPERATING SUBSIDY.

kiss

It’s just that simple.

Don’t talk about disrupting traffic. Don’t talk about TOD. Don’t talk about bridges or tunnels.

If you put your rail line where it requires a very large operating subsidy, you end up having to cut bus service to make up the budgetary impact. This is what Capital Metro had to do during the early days of the Red Line. Both the best 98x buses and the 9 bus were cancelled to make up for operating subsidy overruns from the Red Line. Only today is the operating subsidy anywhere close to the original budget (and it’s still monstrously high – something like $20/ride). We’d have more buses running more routes today if the Red Line had never been built, in other words. The presence of the Red Line means that the¬†people of Austin have less transit today than they otherwise would have. This is how you can tell it was a BAD RAIL LINE.

If you put your rail line where it requires a very small operating subsidy (ideally less than existing bus service1, you end up having MORE money to spend on more buses elsewhere, or on the next rail line. The best way to find that corridor is to find a corridor where a ton of people ride the bus, and where research indicates even more people would ride the train (because it’s more comfortable and reliable than the bus is today).

Anybody who wants to make it more complicated than that is trying to confuse you and get you to support a rail line that you should not support.

Hey, you ask. What about my second rail line?

Go back to the beginning of this post and repeat. The same, simple, formula works for every single rail line your city will ever build. Pick the corridor where the rail line will have the lowest possible operating subsidy. Rinse. Repeat.

Third rail line? Is it more complicated yet? NO. GO BACK TO THE BEGINNING OF THIS POST AGAIN.

Fourth? Fifth?

NO. NO. GO BACK TO THE BEGINNING. This simple process works for every rail line – it tells you which one you should do next.

This is how you build an actual network instead of a struggling disaster like we have in Austin. Again, anybody who tells you it’s not this simple is trying to fool you into supporting something that’s not in your best interest. They have ulterior motives, like, for instance, being on the board of a community college which took over a decaying mall2. Or, for instance, not wanting to be politically embarassed about previous bad decisions3. Or wanting to make a medical school look shinier.

By the way, if you follow this process, you don’t need to lie about your conversations with the Federal Transit Administration either. Hint.

Now I’m off to Germany. Where they actually use logic like the above. Which is why their rail networks actually, you know, work.

parksandrec_micdrop


  1. One way you can tell whether your city is ready for rail at all is whether you can find a corridor where rail would lower the operating subsidy compared to existing bus service. If you have no such corridor, you might not be a good candidate for rail, yet! 

  2. Hello Highland Mall! 

  3. The real reason for no G/L is this embarassment. Future blog post will show comments about the Federal Transit Administration are misleading at best; lies at worst 

  • This is the first disqus comment to this site, ever. Woo woo woopity woo.

    • Steven Yarak

      Yay.

    • Davesix

      You were happy then, but I notice that you’ve disappeared, after calling me names and failing to rebut my arguments.

  • Define operating subsidy. (serious question)

    • Operating cost per ride (cost of service divided among boardings) minus average fare paid per ride (including passes amortized however is typical).

      Every transit agency maintains this metric and most use it in service decisions.

      The average operating subsidy in Austin for bus service is around 3 bucks last I heard. The Red Line has gotten it down to the low 20s from early, disastrous, days where it was 35 dollars a ride (each way!)

      • Sounds fair. I wasn’t sure if you were talking about the total expense of operating an entire line regardless of the number of riders. Just wanted to be clear since there’s an enormous number of metrics that can be used to measure transit performance. E.g. Compare passenger miles to ridership for SF BART and DC Metrorail.

  • Davesix

    I have better advice: don’t build the stupid rail line at all. Buses are cheaper and more flexible. At $20 per ride, it would be cheaper to subsidize people using Uber. That would also cut transit times for passengers and offer far greater flexibility.
    What, you say, there’s no Uber in Austin? If the city started selling discount vouchers for the service, it would appear immediately and Austin wouldn’t be saddled with a ruinously costly, nineteenth-century, solution to a twenty-first-century problem.

    • I’m not particularly interested in more talking points like 19th century technology from the Gerald Daugherty wing, thanks. Rail has demonstrated an ability to attract passengers that bus cannot (at least in this type of city); and Uber is useless for daily commuters.

      • Davesix

        I don’t live in Austin, and I don’t know the players there, but I’m flattered that you had to engage in ad hominem, since you have no counter argument. Rail is uneconomic and embedded in one or few routes, resulting in high costs and inflexibility.

        • But @Davesix:disqus, you ignored both the central point of the blog entry and @herbiemarkwort:disqus’s important distinction between capital and operating costs.

          I suspect most of us agree that $20-per-ride subsidies are too high. This blog entry is precisely about evaluating rail and alignments in terms of how low you can get the operating subsidies.

          • Davesix

            I assure you that I understand the distinction very well. I continue to maintain that light rail is the highest-cost, least flexible and responsive solution to commuters’ needs.

            From “Austin Urban Rail”: “Light rail alignments are successful to the extent they maximize their ridership in a given corridor. These systems are often required to return other public benefits for the large investment needed to build them. Those benefits include environmental, social equity, employment, economic development, land use, congestion relief, and connectivity to a regional system.”

            What’s really said there is that “other public benefits”, gauzy unquantifiable qualities, must be identified to obscure the desires of the promoters of light rail, which are mostly real estate and construction interests, and, of course, urban planners with delusions of grandiosity.

            Employment in downtown Austin is declining, as it is in most urban centers. Building more rail in the face of that trend is really a colossal waste of scarce resources.

          • You say you understand the distinction well, @Davesix:disqus, yet you continue to ignore operating costs.

            Please show us the data demonstrating that light rail has higher operating costs than buses.

            Your passage from “Austin Urban Rail” explicitly referred to capital costs (“large investment needed to build them”).

          • Davesix

            you can cherry-pick my comments all you want, but you’ve ignored my central point: light rail is ruinously expensive from a capital perspective, and results in a system resistaqnt to inevitable change in commuting patterns. In the long run, subsidies will have to rise in the face of declining ridership.

          • Employment in downtown Austin isn’t diminishing either. You need a new playbook; yours is out of the 1960s.

          • Cherry-pick? Not one of your comments had addressed the main topic at hand.

            We all know rail is expensive to build. You’re not contributing anything interesting or new to the discussion by pointing that fact out.

            We also know it’s myopic to exclusively consider one-time costs when recurring costs in many cases have a greater long-term impact.

            Please provide your evidence for this statement: “In the long run, subsidies will have to rise in the face of declining ridership.” Rail’s relative permanence tends to attract economic development, which brings greater ridership over time.

          • Davesix

            You don’t bother to support any of your assertions with actual data, so I won’t either.

            You wrote, ” Rail’s relative permanence tends to attract economic development, which brings greater ridership over time.”

            Go back and read my statement about who supports light rail, and then re-read that sentence. Follow the money, as they say.

            The majority of employment growth in the Austin MMSA is going to occur outside the CBD, where transit ridership carries only a single-digit share of commuters.

          • You just moved the goalposts. And transit share in the CBD could be much higher, if transit were given the kind of advantages over cars it already has in some of our peer cities.

            I’d like to see you go tell Houston that their light rail network is a failure – that their ridership is declining. They DID have a CBD with declining employment; the full trains are starting to reverse that. We don’t even have to go that far; our CBD employment continues to rise.

            Get a new playbook.

          • Davesix

            Ah, yes: Houston, where transit’s share of CBD commuter trips is 13.1%. Transformartive!

            You wrote, “…starting to reverse that.” Hardly a ringing endorsement, I’d say.

            ” And transit share in the CBD could be much higher,…” Socialism could work, of course, but it’s never been done correctly. Austin has the misfortune of having been developed primarily after the invention of the automobile. As a result, transit is forever doomed as an economically viable major player in the commute trip mix. Those high costs will drive employment, except for UT, into surrounding counties.

          • Davesix

            By the way, transit’s share of commuter trips in the CBD is lower than in the rest of the MMSA. I love that.

          • Again, find a new playbook. You’re stuck in the 1960s.

            Austin’s CBD employment has been growing the last two decades. If costs were going to drive it out, it’d have happened by now. In fact, it’s only transit that can support MORE employment in the CBD – where it produces positive externalities instead of the negative ones produced by suburban office parks. With good rail transit, you’d be seeing more office towers go up than we have now; but the trend even without good rail has been up, not down.

            When you get such a basic fact wrong, it’s difficult to take you seriously.

          • Davesix

            The 1960s? My data came from Demographia’s United States Central Business Districts (Downtowns), 3d Edition, March 2014. The data is from CTPP 2006-2010. Look at Table 5.

            Everyone knows that Austin is growing rapidly. I’m merely pointing out that transit doesn’t carry many commuters to the CBD.

            Read this: http://www.newgeography.com/content/004234-new-central-business-district-employment-and-transit-commuting-data#comments

            At 5.1%, Austin’s CBD market share is the fifteenth smallest of US major metropolitan areas.

          • It takes a particularly dedicated troll to argue this hard about a city you don’t live in to a bunch of people who know it much better than you do.

            UT isn’t included in the census definition of the CBD, BTW. There’s your first hint.

            Your second hint is that I’m aware that you keep shifting the goalposts from actual numbers of jobs to share in the metropolitan area.

            Your third hint is that I don’t care what happens in the suburbs. I am not a resident of Cedar Park; and in fact, Austinites greatly subsidize them already (they do not pay into our transit system; and they get far more back for their gas taxes than we do); I feel no need to make that any worse, thanks.

            The “stuck in the 1960s” comments are in regards to your idea that the CBD is somehow shrinking instead of growing, and that the future is office parks on suburban arterials.

          • Davesix

            This is amusing. You’re right: we’ve strayed from my original point that light rail is a stupidly expensive way to move commuters, whether you’re looking at capital coasts or per-ride subsidies, which average between two and three dollars per passenger trip nationally.

            Your numbers in Austin, by comparison, are pretty horrible, aren’t they? They’re not going to get much better, are they?

            What a service to the taxpayers!

          • Kory Strickland

            There’s no reason Austin can’t achieve much lower per-ride subsidies than the terrible Red Line, assuming a reasonable alignment is chosen, which was the point of the original blog post…

            If you feel the national average subsidy is too high, what level do you think is an appropriate per-ride subsidy for transit trips? Considering that transit riders impose much lower congestion and pollution externalities than car drivers, it would be efficient to subsidize it at some level, no?

          • Davesix

            Light rail is a financial disaster anywhere it’s been built. Any alignment will be similarly disastrous and, as I wrote, “stupidly costly”.

            I didn’t say that the national average subsidy is too high. I merely stated what it is.

            Quoting you, ” Considering that transit riders impose much lower congestion and pollution externalities than car drivers, it would be efficient to subsidize it at some level, no?

            I do not agree that, in some cases, “…transit riders impose much lower congestion and pollution externalities than car drivers”. Where I live, it is not uncommon to see a bus that gets about three miles per gallon carrying three passengers.

            That’s about as fuel-efficient as a thirty year old pickup truck, and probably more polluting. To boot, you have to ignore the per-ride subsidy if you’re going to do the comparison.

            The bus fails the test, but it is still far better than light rail. Bus transit scores especially well in terms of flexibility. It’s easy to change bus routes and bus frequency. One early response to my comments stated that people “won’t ride buses” or words to that effect, as though light rail is morally and culturally superior to buses, which attract only the underclass. I think that attitude is probably reflective of a dismissive attitude toward the “customer base”, which, one way or another, pays the bills for light rail, with its huge costs, both initially, and per passenger.

            Automobile fuel efficiency and pollution have been improving dramatically recently, mostly as a result of competitive forces.

            It’s not fashionable to say so, but additional freeway lanes move commuters more efficiently, at lower cost, and at greater volume(22,000 cars per day per lane), than light rail, and they do so in a way that conforms to consumer and taxpayer preferences.

            if that weren’t so, then the CBD of Austin would be blessed with a transit share of more than 5.1%.

          • “Light rail is a financial disaster anywhere it’s been built.”

            Tell it to Houston or Phoenix, both of whom achieved operating subsidy reductions by building rail in the right place instead of the wrong place. Which allows for more transit service overall. Which is kind of the point of this piece.

            Or just get the hell out of this blog. Nobody here wants to hear the Randall O’Toole nonsense. There’s plenty of other places you could peddle that crap. The only thing worse than the Austin idiots who think any rail is good rail are disingenuous shills like you who claim any rail is bad rail.

            Improve your game or GTFO. Thanks.

          • Davesix

            It’s been a pleasure to needle you, but you have not refuted my argument. Rail is a financial disaster. I’ve met Randal, and unlike you, he’s a cheerful guy who doesn’t stoop to name-calling.

          • He’s a smiling liar. I’ve wasted enough time in my life with guys like that to know it’s not worth pretending they’re not lying just because they’re smiling when they do it.

          • Davesix

            You don’t want me to dive into the ATA data to rebut your argument about the financial health and per-ride subsidy in Houston and Phoenix, do you, because you know it refutes your argument. If that’s not true, cite the data yourself.

            You know where it is.

          • Davesix

            “Or just get the hell out of this blog. Nobody here wants to hear the Randall O’Toole nonsense. There’s plenty of other places you could peddle that crap. The only thing worse than the Austin idiots who think any rail is good rail are disingenuous shills like you who claim any rail is bad rail.”

            So much for intellectual rigor on your side. This is really fun.

          • Davesix

            “Tell it to Houston or Phoenix, both of whom achieved operating subsidy reductions by building rail in the right place…”

            So how big are those “reduced subsidies”?

          • Kory Strickland

            It’s certainly possible for transit with very low ridership to cause pollution and congestion externalities greater than what you would see if all riders on that route drove. However, in cases where ridership is higher and transit riders are uncontroversially causing lower pollution and congestion externalities than drivers, don’t you agree that transit should be subsidized?

            Also, arguments about existing consumer preferences aren’t the best guide to optimal policy, considering that current transportation choices are distorted by the existence of pollution and congestion externalities, etc.

          • Davesix

            Kory wrote, “…congestion externalities than drivers, don’t you agree that transit should be subsidized?”

            I do, but I think that the goal in any community with subsidized fixed-route transit should be to work toward a more user-responsive model, such as private-vehicle ridesharing.

            That model might certainly be less expensive than light rail, but needs some work by private-sector vendors to further develop the model and make it even more accessible to commuters.

            Then, you wrote, “Also, arguments about existing consumer preferences aren’t the best guide to optimal policy, considering that current transportation choices are distorted by the existence of pollution and congestion externalities, etc.”

            What’s wrong about using existing consumer preferences to guide “optimal policy”? A 6% transit share exists almost everywhere. Why work to change it, at great expense for no increase in consumer satisfaction?

            Parking for cars is far, far, cheaper, even if you don’t like it.

            Congestion is its own limiter, by the way, because it drives development out of congested areas of the city and into areas that are less congested.

            That might result in a development pattern that you, whoever your are, doesn’t like, but I can guarantee that it will be more efficient economically.

            I could go on…

          • Davesix

            To the rail guys who will chime in to claim that the answer to congestion is rail, you know that rail has never reduced congestion.

            Anywhere.

          • No, it doesn’t. If it did, there would be little reason to build rail. Rail offers an alternative to congestion. And it also offers an alternative to the expense of owning and operating a car.

          • Davesix

            Everyone! wants to own and operate a car, if he can afford to do so, nevermind the freedom of movement that ownership offers. As a consequence, everyone except the most impoverished will own a car and use it for trips to Costco, as an example of consumer preference.

            Owning and operating a car is less expensive than using public transit, and everyone who can afford to do so will own and operate one, even if he takes transit to work at great taxpayer expense and subsidy.

          • This is stupid. Your claim is obviously false as there exist numerous people who are easily verified who will tell you they prefer not to own a car. This isn’t a matter of opinion; it doesn’t matter whether you think those people are dumb or not; they exist in large numbers. Just as people who do want to own cars exist in large numbers (larger, even).

            You’re just repeating talking points now.

            You’re a troll, and if you do this kind of thing one more time, you’re gone.

          • Davesix

            ” Your claim is obviously false as there exist numerous people who are easily verified who will tell you they prefer not to own a car.”

            That’s true, but in Austin, where 94.9% of commuters to the CBD do so by car, that percentage is small.

          • Davesix

            What rail really does is relocate the parking lot, since nearly all riders will commute by car to the station nearest them and park in the lot, if it’s properly designed and sized, or on the street in the neighborhood, transferring the congestion impacts to those outlying neighborhoods.

          • Houston’s first line does have a lot of park-and-ride traffic. They had the park-and-ride before; it was just served by buses, less efficiently.

            The difference is that no developers built along the bus line; they are building along the rail line, because people commuting via train to the Medical Center or to downtown have expressed interest in being able to walk to stations.

            And, of course, the demonstrated operating cost reductions from replacing the earlier bus line(s) with rail have led to increased service elsewhere. Which is what road warriors are really afraid of, in the long-run.

          • Davesix

            “. They had the park-and-ride before; it was just served by buses, less efficiently.”
            But at less total cost, I’m sure.

          • Not in the long-run. Operating costs are far lower; you typically want to invest capital to save operating costs if you are actually interested in maintaining service instead of just being a road warrior troll.

          • Davesix

            The central question is, who “invests” and who benefits, isn’t it? A second-order question is , How much is invested for what increase in the level of service, and how is that investment amortized, in a world where farebox revenue covers only a portion of the cost of providing the service.?

          • Davesix

            “…just being a road warrior troll.”
            I take that as a compliment, and I hope that readers will notice that I don’t call you names, even though I disagree with you vehemently.

          • You’re gone. I have no more tolerance for this dishonesty that’s wasting everyone’s time.

          • Davesix

            You wrote, “… typically want to invest capital to save operating costs if you are actually interested in maintaining service “. You’d like to ignore capital investments in favor of a focus on operating costs, but you know that’s an incomplete analysis. It is likely that the capital costs will never be amortized.

          • Neither will the capital costs for highways. And the whole point of this thread was where to invest the capital dollars to provide the most operating benefit, not whether to invest them at all.

          • Davesix

            “And, of course, the demonstrated operating cost reductions from replacing the earlier bus line(s) with rail have led to increased service elsewhere. ”

            That’s not typical. Perhaps you could direct me to a site where the data is accessible, because I am skeptical of that claim.

            The other problem for you is that more stops on the rail line to serve people who want to live near the line and walk to a station means lower speeds and less efficiency for the line. Certainly you’re aware of that.

          • Davesix

            This is a big document. You might at least direct me to a page within the document, and accompany that with some prose from you about the nature of those cost reductions.

          • Kory Strickland

            I like ridesharing and carpooling, and in an ideal environment, subsidies and/or Pigovian taxes would encourage it be done at an efficient level.

            As you say, the private vendors may not quite be there yet, so I think it’s difficult to assess the relative efficiency and cost-effectiveness of ridesharing vs. transit. Could you point me to some links/research to help me understand why you’re so confident it’s the better option?

            I’m afraid I find arguments about flexibility vs. fixed investment a bit nebulous in the context of specific proposed rail alignments. Now, argument that a specific alignment is a bad idea because future land use changes are going to result in -lower- ridership over time would certainly be compelling, but I haven’t seen anything like that for any of the proposed lines in Austin. In a city growing as fast as we are I think that’s a tough row to hoe.

            “I can guarantee that it will be more efficient economically.” I confess I’m a bit confused by your economic analysis. You assert that congestion costs will naturally shape driving and development choices in efficient ways, but surely you know that only some (essentially half) of congestion costs are borne by the driver, meaning that without other taxes or subsidies, we’ll tend to see systematically more driving and congestion than the efficient level?

            Further, required parking minimums in zoning codes act to subsidize parking, which would tend to encourage driving above and beyond the efficient level you would get if drivers bore the costs they incurred by choosing to drive.

            For these and similar reasons, I don’t follow your inference that the number of people who choose to drive currently is the economically efficient level.

          • Davesix

            The current level is the market level, and you’re ignoring my comment about the location of the parking lot.
            Further, if commuters prefer to travel by car, then why should infrastructure not be designed and constructed to conform to their preferences?

            You’re not going to un-invent the automobile, no matter how repugnant you find the machine to be or how much taxpayer money you spend in a futile effort to pry people out of them, in all likelihood.

          • Kory Strickland

            The current level is the market level, but the market level equals the economically efficient level only in a free market clear of distortions (externality, subsidy, monopoly, etc.) or in a market where the distortions are equal in magnitude but opposite in sign, canceling out.

            My argument is that various distortions combine to make the market level (of driving) systematically higher than the economically efficient level. I thought we were in agreement on this?

            I’m not sure that I’m the intended audience for your post on parking. I’m not opposed to supplying parking always and everywhere. I merely expect that government regulation mandating off-street parking is likely to result in over-provision of off-street parking relative to the economically efficient level. Do you disagree?

            Re: consumer preferences – shouldn’t we try to create a transportation system that’s the best possible from a cost/benefit perspective, not just try to make everyone happy regardless of cost? It would probably make many people happy to be chauffeured everywhere in a limousine, but I’m not going to advocate having the government pay for free limo rides for all citizens.

          • Davesix

            You wrote, ” I merely expect that government regulation mandating off-street parking is likely to result in over-provision of off-street parking relative to the economically efficient level. ”

            Go back to my comment about who pays for that parking. If the code requirement is for too much parking, based on usage data for existing facilities, developers will immediately use statistics for the use of existing parking facilities to argue for reductions for new projects.

            “Re: consumer preferences – shouldn’t we try to create a transportation system that’s the best possible from a cost/benefit perspective, not just try to make everyone happy regardless of cost?”

            Rail is always the highest-cost provider. You might like the efficiency, but a large part of the cost is always imposed on taxpayers who don’t use the system. I think that part of the cost of the system should be disclosed, but it never is by proponants.

            We have seen here gauzy statements about the desirability of alignments that result in per-trip subsidies lower than the Red Line subsidies of $20.

            I challenge anyone here to justify the subsidy level that is likely for a new alignment, and to accurately estimate what that subsidy might be.

          • Kory Strickland

            I responded to your parking post above. I doubt that outcomes negotiated on an ad-hoc basis between developers and local government types are going to be better than what you’d get if you just let the market allocate parking.

            My impression is that rail is actually unusual in how transparent its costs are. It doesn’t cause congestion externalities which are difficult to price. It causes fewer pollution externalities which are difficult to price. The real estate market is pervasively distorted by parking minimums in a way that’s difficult to price. Finally, I imagine it’s usually relatively easy to determine how much is being spent on any given rail line, while the subsidy for roads drawn from the general fund is then spent on new construction and maintenance for many different roadways. (Note that all of the costs I’ve listed are borne at least in part by non-drivers.)

            In trying to decide whether a rail line is a cost effective transit investment, I imagine that figuring out a likely subsidy is the easy part. The hard part is just figuring out whether that level of subsidy is justifiable, given the extent to which the true costs of driving are hidden and difficult to calculate.

            That being said, I don’t know that anyone has run the numbers on a possible subsidy for the Highland alignment that’ll probably be under consideration in November, since the details of how it’s going to look are still being worked out. I’m expecting it to be quite high; it’s a real dog of an alignment. For that reason I intend to vote against it if it shows up on the ballot in November, but I still don’t understand why you’re so convinced that rail is always and everywhere a bad investment.

          • Davesix

            You wrote, ” I doubt that outcomes negotiated on an ad-hoc basis between developers and local government types are going to be better than what you’d get if you just let the market allocate parking.”

            First, “ad-hoc” is not how it’s done. There are requirements incorporated into the code, based on the occupancy of the development. because the code is an iterative document, it is subject to periodic change based on experience.

            In addition, in most jurisdictions, developments are subject to traffic-impact studies, and mitigation fees are imposed by local authorities based on projected traffic impacts and resulting capital requirements for new facilities.

            Then you wrote, “My impression is that rail is actually unusual in how transparent its costs are.”

            If you mean projected costs, this is a howler.

            Then you wrote, “In trying to decide whether a rail line is a cost effective transit investment, I imagine that figuring out a likely subsidy is the easy part.”

            Are you serious?
            Always and everywhere, rail advocates underestimate costs and overestimate ridership. The result is a rosy projection that is never (!!!) borne out after the line is built.

            Compare the cost of rail with the cost of an additional freeway lane that will carry as many people who are paying for their own labor, and you’ll have a clear analysis, if you do it in an honest way.

            It used to be that transportation systems were designed by civil engineers and traffic planners with the goal of accommodation of demand as the central driver of planning.

            Now that planning has been usurped by planners with an entirely different agenda: they want to drive people out of automobiles and into inflexible mass transit facilities.

            The market doesn’t buy it.

            94.9% of commuters to the CBD commute by car, and no amount of obscenely huge investment in light rail will change that by more than a percentage point or two.

            Don’t spend the money. It’s a loser.

            By the way, your replies to my comments suggest that you’re not reading what I write very carefully.

          • Kory Strickland

            Re: parking, see my post above. I’m still not sure why you think a centrally planned political process will result in better outcomes than a market process.

            Assuming you’re right on construction costs, the relative costs of construction for freeway lanes and rail lines fail to encompass the total costs of driving vs. riding, right? What do we get from a simple comparison of construction costs?

            If you’re concerned about the budgetary impact of trying to improve transportation system efficiency, you should know that for all the market distortions I’ve discussed, there are policies that correct the distortion while actually raising revenue. Increasing user fees (gas taxes, tolls, etc.) could have drivers pay for road construction and maintenance without having to dip into the general fund. Eliminating parking minimums costs the government nothing and increases potential parking meter and property tax revenue. A Pigovian tax on gasoline could internalize pollution externalities. Congestion can be controlled using congestion pricing and variable rate tolling.

            If you implemented all of these policies, the private cost of driving would equal the true cost of driving, and you could eliminate all subsidy for transit and get the optimal level of transit use purely through market action. The other option is to subsidize transit. I prefer the former, but I support making our transportation system more economically efficient than it is today, one way or another.

          • Davesix

            You wrote, “… but surely you know that only some (essentially half) of congestion costs are borne by the driver, …”

            I’m sorry, but I don’t see data that support that argument.

            Parking in the CBD is primarily financed by developers as the result of prescriptive parking requirements. Those developers always argue (plead?) for reductions in required parking because of cost, and very often offer vouchers for the patronization of light rail and transit in exchange for reductions in parking facilities.

            Those arguments ring hollow with me, since the residents or occupants of the developments don’t use transit or rail at the promised level.

            My experience is that the resulting parking loads are transferred to neighboring properties and streets, in a way that truly transfers congestion costs to those who don’t deserve them.

            I know that my arguments have morphed away from rail, but I continue to maintain that the capital costs of light rail are usually ignore or at least minimized in any analysis.

            It is important to remember that developers pay for parking in a properly managed environment, and that surface streets are cheaper to develop than light rail.

          • Kory Strickland

            Data? It’s an accounting identity. If I choose to drive in rush hour traffic and experience 20 minutes of congestion related delays, I’m also causing 20 minutes of congestion related delays to other people, at least on average. So, my choice to drive at rush hour wastes 40 total minutes, but my private cost is only 20 minutes. That’s what I mean when I talk about congestion externalities. Does that make sense?

            I still don’t understand your arguments for government mandated parking minimums. You’re worried that if we got government out of the way and let the market allocate parking, we would see less of a glut of off-street parking and more people would park on the street? All else being equal, that would seem likely, but I’m not seeing the problem. Unused street parking space strikes me as wasted public property. (And, where demand is high enough to support meters, a wasted opportunity for the government to collect revenue in a non-distortionary way for once)

          • Davesix

            Speaking of UT, it’s apparent that the city of Austin has allowed UT to expand its campus without requiring it to construct code-stipulated parking, resulting in a transfer of costs from UT to the taxpayers and inflating the demand for light-rail from students, staff, and faculty, subsidized by the taxpayers to the tune of $20 per trip.

            That, sir, is likely the reason that Demographia left UT out of its analysis.

          • You’re an amazing troll.

            Austin has no control whatsoever over UT, but even if they did, the process you apparently favor – requiring the building of parking – functions as a subsidy to drivers; NOT requiring the building of parking would be the normal baseline.

          • Davesix

            Twenty bucks a ride? The typical transit per-ride subsidy in the US is about $3. I think that’s absurdly high, but that should be achievable, even in Austin. After all, Austin is in Texas, even if it’s uncomfortable in its environs.

          • Davesix

            “I suspect most of us agree that $20-per-ride subsidies are too high.”

            I assume you wrote that with a straight face.

      • Davesix

        You wrote, “Uber is useless for daily commuters.”

        You have no idea how Uber or other ridesharing services work, do you? Uber is a very nice solution to the congestion problem, especially with small subsidies.

    • Capital costs, yes rail lines are more expensive; operating costs, not at all. I’ll just give St. Louis as a common example for farebox recovery: MetroLink 30.5%, MetroBus 21.9%. (data from here 30-40 pages from end)

      And automobiles are 19th century technology, as well.

      • Davesix

        “. I’ll just give St. Louis as a common example for farebox recovery: MetroLink 30.5%, MetroBus 21.9%. (data from here 30-40 pages from end)”

        30.5% farebox recovery is a triumph!

        • Yes it is. And Missouri may vote for a 1% sales tax (also known as a subsidy) in November to fund “transportation” as MoDOT is slowly going broke. I believe that will lower the farebox/gas tax recovery for highways to 30% or lower.

          • Davesix

            “MoDOT is slowly going broke. ”

            With good reason.

            The best of luck to the citizens of Missouri, in the hope that sanity might prevail.

            Rail advocates might refer to the wisdom of the voters in Tigard, Oregon, who just voted against an expansion of the incipiently bankrupt Trimet.

          • That comment would make sense if MoDOT gave Metro more than a $1M per year, which they never have. Nope, it’s all the fault of a very low gas tax that no politician is willing to even peg to the rate of inflation.

          • Davesix

            I have faith in the voters, surprisingly, but I think that they are often misled by people who do not have their interests in mind. I don’t that my opinion should surprise you.

  • Terry Grier

    Nice Post Mike. Thanks for posting.

  • As far as operating subsidies go; let’s use Minneapolis for an example because google works better with them:

    http://www.startribune.com/local/west/123722719.html

    Once derided as a “train to nowhere,” the Hiawatha light-rail line is more popular than predicted. It carries enough riders over 12 miles that it is cheaper to run than metro buses.

    The cost of operating the Hiawatha between Bloom-ington and downtown Minneapolis is about half that of Metro Transit buses in the Twin Cities, according to comparisons by the federal government. The Hiawatha also is less expensive to operate than most other light-rail lines in the nation.

    […]

    Labor, energy and maintenance make up the bulk of the operating costs. Hiawatha’s costs are lower than buses in part because one person can operate a three-car train carrying more people than a single bus.

    The operating cost per passenger mile of the Hiawatha line is the seventh-lowest among 27 light-rail lines in the United States, the NTD reported. Salt Lake City, Houston and Baltimore operate light-rail systems with costs comparable to the Hiawatha. Buffalo, Newark and Pittsburgh light rails are among those costing more than twice as much as the Hiawatha.

    The NTD figures are “the best single source of data …consistent data, that the industry has,” Cherrington said.