Phoenix Light Rail Fail -- Half My Light Rail Bet Settled

When Phoenix was building its light rail system, I made the following two-part bet:

  1. I could take all the money spent on construction and easily buy a Prius for every single daily rider, with money to spare
  2. I could take the operating deficits for light rail and buy everyone gas to run their Prius 10,000 miles per year and still have money left over.

This bet has been tested in a number of cities, including LA and Albuquerque, and I have not lost yet.  Now the numbers are in for Phoenix initial ridership, and I am winning the first half of my bet in a landslide.

The other day, Phoenix trumpeted that its daily ridership had reached 37,000 boardings per weekday.  Since most of those people have two boardings per day (one each direction) we can think of this as 18,500 people making a round trip each day.

Well, if we bought each of these folks a brand new Prius III for $23,000 it would cost us just over $425 million.  This is WAY less than the $1.4 billion we pay to move them by rail instead.   We could have bought every regular rider a Prius and still have a billion dollars left over!  And, having a Prius, they would be able to commute and get good gas mileage anywhere they wanted to go in Phoenix, rather than just a maximum of 20 miles on just one line.  Sure, I suppose one could argue that light rail is still relatively new and will grow, but even if ridership triples, I still win the fist half of my bet.  And as the system expands, my bet just looks better, as every single expansion proposal has been at a cost of $100 million a mile or more, more expensive than the first 20 miles.

So now, all we have to do is wait to see the operating results to settle the second half of my bet.  If common practice is followed from other metro areas, this will be extremely difficult to prove because the authority will do everything it can to hide the huge operating dollar hole light rail is creating.

But Coyote, what about congestion?

I am glad you asked.  Folks will argue that rail reduces congestion.  Normally, I would agree but argue that it reduces congestion at way too high of a price.  But for Phoenix light rail, it may even be that rail makes congestion worse.

Here is why:  In building Phoenix light rail, the city along most of the line had to remove two lanes of traffic (one each way) to build the line.  So here is the comparison:

  • Light rail carries 37,000 trips per day or about 2,000 per hour  (1,000 each way) through its 18-hour operating day, though certainly there are peaks and valleys around this average
  • A typical lane of road has a capacity of 2000 cars per hour, so light rail removed 4,000 cars per hour of road capacity (2,000 each way).  Its unclear how many riders this equates to, but the average car in the city has 1.5 passengers, so we will call this a road capacity of 6,000 trips per hour (3,000 each way).

So, we have replaced roads that can carry 6,000 trips per hour with train tracks carrying 2,000 trips per hour.  Sure, the train carries more than 2,000 in some peak periods, but probably not more than the road it replaced was capable of carrying.  Further, I can attest from personal experience that the complexity of trains on the road and passing through intersections screws up the timing of lights and results in lost capacity on the roads in the area that remain.

A Thought on Sports Team Subsidies

I would love to see the ridership of the light rail on days with and without a baseball game or basketball game downtown.  My sense is that a significant portion of the ridership is from game attendees (its the only time I have found it useful to ride the train).  If this is the case, then this massive overspending for light rail represents yet another subsidy for professional sports teams.

By the way, I just realized that I am underestimating the financial cost to the city of the train.  Most sports fans ride it not as a transit substitute per se but as a parking substitute -- the train allows one to park cheaply away from downtown and ride the the game without traffic hassles.  I wonder how much the lost $15 a pop parking in city lots by the stadium due to the train is costing the city?

Light Rail Hurts the Working Poor

I think it is always important to reiterate why light rail is such a threat to the working poor who depend on transit.  As I wrote the other day:

"¦light rail is simply not transit for the working poor. It is transit for yuppies that happens to be used by some working poor.  They are built for white collar workers commuting to town who are too high and mighty to be caught dead in a "grubby" bus.  But since light rail is orders of magnitude more expensive than buses, two things happen in every city that ever builds light rail.

1) Light rail fares skyrocket to cover their immense operating deficits and capital costs, giving the lie to politicians that sold these systems as helping working poor.

2) Bus service, the form of transit that serves most of the working poor even today in the Bay Area, is cut back to help pay for rail.

Light rail is the worst enemy of providing transit services to the working poor ever devised in this country.


  1. Doug:

    You're missing one other factor in this "less traffic" argument. I live near Mountain View, CA. These guys are the epitome of political correctness, so they lobbied to have the San Jose Light Railway brought up north to a dead-end station at the foot of downtown Mountain View. At times, particularly on weekends, I have had occasion to travel down Central Expressway, which is crossed over by this train track.

    More often than not, particularly on weekends, the rail crossing lights go on, the barriers drop to stop all traffic on the Expressway, and we wait. And wait. And wait. Eventually, along comes a 4-car train, always carrying no more than 10 people total, and going no more than 5 miles per hour. I make a point of counting cars that have had to stop and wait for this little show, and often see 20 cars *just in my direction* stopped and waiting for this albatross to cross the road. I would say it's a good 4-5 minute delay.

    How much CO2 do you suppose the waiting cars had to put into the atmosphere waiting in line for this "zero pollution train"? Bear in mind, Central Expressway is a major thoroughfare around here, so weekdays cause a whole lot more cars to wait than they do on weekends. I should add, that I have NEVER seen more than 10 people on board on the weekends in several years of observation.

    Train advocates would be good if they could stop ALL cars running dead in their tracks at such crossings, and still argue that they create no pollution.

  2. Bob Smith:

    Folks will argue that rail reduces congestion

    Such arguments assume that every rail passenger would have been driving a car but for the train, and that the train is always at peak capacity. Both are demonstrably false.

    Doug, why didn't those bozos build a bridge over Central Expressway? Blocking a major artery like Central is crazy.

  3. Allen:

    LRT does not reduce congestion. The problem with your calculations is that you need to factor in :

    a) Number of trips (aka ridership) that were already occurring via transit that got shifted to light rail. That is, trips that weren't being made by car. These don't reduce congestion because they weren't driving in the first place. In the US about 2/3 - 3/4th of the initial trips (aka ridership) was already taking transit. If initial daily trips for the first year was 30,000, it's most likely that 20k-23k of those trips were already occurring without autos. Well... shouldn't frame it that way since a large number of the LRT trips are only part of a commute that often involves driving to the train station (or driving to the express bus stop previously). So you're looking at 7k-10k for new trips thanks to LRT.

    b) Over time it stands to reason that not all new light rail riders are only taking transit because of light rail. A certain portion of them would be taking transit no matter what. What is a good rule of them? I've never seen anyone take this one on. Given that most riders take transit to avoid paying for parking (going downtown to work; attending sporting events) or because they don't have a vehicle, let's just go with 50% of new riders would've taken transit no matter the form. Personally I think it's more like 70%, just like initial ridership. But let's be under counting this one just for the sake of putting on the kiddy gloves. So in the case of Phoenix's LRT, we're looking at 7k increase over the first year (I might have these number of trips - aka ridership - wrong, just took a quick minute to peak at them). So about 3500 of those trips wouldn't have been via car no matter you had buses, built bike pathes, PRT, LRT, BRT, heavy rail, monorail, subway or carts pulled by shaven yaks.

    c) How accurate are the ridership numbers? From what I've some agencies means of counting riders are questionable. First off, I have yet to see a single one release numbers based on revenues? Could that be under counting trips? Sure but doesn't that mean there's a big problem with people not paying if they're off by a lot? As for how they count, I've seen two forms. Initially MTC, Minneapolis/St.Paul's transit agency, counted ridership via "random" head counts on trains. They then moved to what RTD in Denver does, which is have about 1/3 of their rail vehicles have the steps designed to count people getting on and off. Why 1/3? It's hard not to be cynical as to why. From what I've seen, for 2 and 3 car trains, riders tend to head to the middle and back car more than the front. Gee, if you only have 1 of those 3 cars with step counters, how many do you think are going to end up front? Anyway, if possible I'd take a hard look as to how Phoenix's transit agency is coming up with that 37,000 daily trip number.

    To get back to A and B, in this situation you're looking at maybe 13k or 14k of those 37k trips actually being trips that used to be made by car. So what? 6,000 cars off the road? 7,000? Another way of calculating new trips would be to compare the no build alternative's transit numbers to those of the build lrt numbers.

    That mentioned, I'd love to know how those trips on LRT compare to other routes near that same corridor. For example, the SE line in Denver does chest thumping over it's 38,000 trips / day during the week. Yet just part of the freeway that it parallels, the tech center portion (the portion near downtown has even higher "ridership"), has a "ridership" of 425,000 a day.

    Keep in mind though that those more cars will need road capacity to get around. But 6 or 7k could easily be handled by a HOT land for far, far, far less of a cost.

  4. William Newman:

    "Light rail is the worst enemy of providing transit services to the working poor ever devised in this country." Really, the *worst* enemy?

    It seems to me that making transit services a city-imposed monopoly is a pretty ferocious enemy. If private companies were allowed to operate buses and jitneys under traffic rules comparable to those for delivery trucks now, and if people were free to advertise carpooling arrangements involving fees, would low-cost non-personal-auto transport be worse or better than it is now?

    Also, it might be interesting to run a back-of-the-envelope calculation on the impact of limiting imports of relatively economical Japanese cars, too. How many marginal buyers became unable to afford their own car? I have no idea, but it might be large. Possibly the number compares to the number who ride buses every day?

  5. Matt:

    It is posts like this that really worry me about what kind of crap they'll be putting in over here in Tampa. Great idea on the bet though.

  6. DaveK:

    Now here's a couple of questions: Just how many new, fuel-efficient buses could you buy for the cost of the light rail system? And what would be the operating cost of those extra buses compared to that of the light rail system? How much of that operating cost could have been paid by the operating deficit of the light-rail system?

    Just askin'

  7. Dan Smith:

    Why don't you add the light rail in Minneapolis to your wager? Built at a cost of $715 million (estimate was for $400M)it serves 10,000 riders/ day. That will buy the Prius and leave $47,500 per person for incidentals. By the way, construction is starting on an extension into St.Paul as we speak.

  8. Rail Life:

    "the train allows one to park cheaply away from downtown and ride the the game without traffic hassles. I wonder how much the lost $15 a pop parking in city lots by the stadium due to the train is costing the city?"

    Really great point. The studies that have shown $6 billion in economic development along the line probably didn't figure in this huge loss in parking revenue.

  9. Gordon:

    About lost revenue from light rail. Question? How many parking garages are needed downtown for all the driving commuters? What is the total revenue collected from parking fees and taxes (property taxes)?
    Parking garages are not considered the highest and best use for downtown real estate. So if a parking garage is torn down and replaced with a high rise office tower, what is the tax revenue the city and county can collect vs the parking structure it replaced? How does this economic development figure into your figures?

  10. Randall Scott:

    Well thought out. LRT is almost always a mistake. For the same cost, instead 2 freeway lanes & 3 bus routes could be paid for, roughly.

    There are two extra-worse mistakes done in the planning of the PHX LRT. The crossing that was picked for the Salt Wash is about the widest & the only place that has water. And the route goes right by Sky Harbor, instead of going through. That could have doubled ridership. Now there's another transit system (all vehicles are people movers) being developed for a $billion for a few miles.

    I come across your wisdom via various , unrelated, items that I read. Particularly in my searching for the doom-&-gloom of the AGW fakery, I've seen some videos & comments posted by you. I would like to visit your blog more often.

    Randal O'Toole even has a link for your blog on his Thoreau Institute site. Are you aware of Wendell Cox? His site is; he analyzes all things urban & then some. Also involved in urban & econ is a prof at USC, Peter Gordon.

    I see you have some links here for great sites, such as Mises, Club for Growth & Cato.

    I'm sure you also visit Heritage, Heartland, Reason, Pacific Research Institute, Hoover & so on. I could list dozens of great groups.

    Classical Liberalism rules.

    Do no harm.

  11. William Newman::

    William Newman ---> Excellent point. I'd add to that in branching outside of what we normally label "transit" and throw taxi licensing into the mix as affecting the poor.

  12. Αμάτι νώνυμος:

    $23,000 Cars For Commuters in Phoenix is Cheaper Than Light Rail for $1.4B

    Breaking down the 1.4 * 10^9 dollars you then find that large part of the mix is the destruction of complex things above the proposed rail. There are physical structures, business good will, interruption of traffic between distant businesses, and the list goes on. By contrast the completion of such a project in the middle of a Wyoming cattle range will be a minute fraction of the present quote. Would you guess that there are thousands of cities in our enormous nation? Net worth of all those cities would overshadow the expense of building just one totally new metropolis within our most remote county of some centrally located state.

    Start from a scratch into the hard soil. From the ground up you build first the trenches to become future tunnels for sewers, pipes, wiring, and trains. Then you build culverts, bridges, and foundations for buildings. Then highways and railways. Then secondary streets, landing strips, heliports, emergency services buildings, and continue until basic things are provided. Then sell land to entrepreneurs, private citizens, etc.

    Entire project will be our biggest bargain since self-control.

    Now that we have torn the &#!- out of Iraq, perhaps we should build our first bargain city within their country. That brand of kindness will create more Iraqian Stability than all the backwater troops in the jungle.

    With bargain priced spiffy new light rail transport and bike sharing system, people of all ages can enjoy the outdoor freedom, exercise and natural beauty of their country.

    Avant Guard Suggestions are quickly swept under the carpet, of course. But when the time comes to renovate the decay of inner city, or rebuild a bombed out crater,

    Renovate Not

  13. larry s:

    This notion of buying a car for every commuter at the expense of light rail is getting to be old and, as usual, is still misleading. What about the cost saved on road wear & tear, car expenses, gas, human fatigue and frustration, pollution, and on and on. Why can't people see the big picture or are our minds too little for such.

  14. joshv:

    Coyote, how then do you explain the continuing success of light rail in Chicago and NYC? I think the failure of light rail in places like Phoenix is more an indication of a failure to create a vital central business district. Simply put, 'there is no there there'. One might go downtown for a game, but few people actually work there. There is no workable hub for such a system.

    So please, qualify you objections, in sprawling cities that were built mostly after the advent of the car, light rail systems will probably never work. In older cities with dense central business districts, light rail works just fine.

  15. SteveH:

    Some clarifications:

    The $1.4B expense includes about $400M of street and infrastructure (sewer, water, electrical, gas) replacement and repair that was going to have to be done anyway. Also in that number are all the one-time start-up expenses associated with planning, design, right-of-way purchases and environmental clearances.

    The useful life (GASB / FTA) of a car is typically 5 years; a city bus, 12 years; and a light train, 30 years. Ignoring operating costs, you’d need 6 cars over the 30 years so that $425M is now a little over $2.5B for the cars v $1.4B for the train (which is inflated, see above).

    Not to mention that of the 18,500 people who received the cars, based on deaths per passenger mile between car and rail, ~90 of those people would still be alive if they’d stayed with the rail. Not sure how to put a price on that.

    The Central corridor and downtown Phoenix are the largest commercial centers in the state and the seat of many city, county and state government agencies. I’m sure a large number of government workers take the rail as it goes to where they work. I’m also sure that when Prop. 400 passed it wasn’t a bunch of government workers looking for a nice ride. It was reasonable people understanding that if you’re going to have a serious and attractive city, you need a comprehensive, multi-modal and forward looking transportation plan.

    The rail is an investment in the future and we will thank ourselves tremendously as the city grows and congestion continues its unrelenting pace.

    And I will bet on that!

  16. SWB:

    To Gordon,

    Any new 'down-town' high rise building, which does not include adequate parking for it's patrons in its basement; and or, first few floors is idiotic anyways! Go figure that into to mix!

  17. James:

    a boarding is one person stepping onto a train. anyone can step onto a train fifty times a day that is fifty boardings
    most people board a train twice a day. So yes there is about 18,500 riders but there is about 37,000 boardings a day.

  18. Losers:

    Are you guys enjoying your little automobile-loving orgy? "2 more lanes on he freeway"...jizz! Adding two lanes to a 10 lane freeway does jack shit to reduce congestion. Maybe light rail would be used more if your dumphole of a city weren't designed for all you car crazy morons.

  19. Neville:

    Light rail is best understood as the modern equivalent of the medieval cathedral. A large scale devotional object, with few practical uses and substantial operating costs, which may well be around for centuries.

  20. 5chw4r7z:

    1 lane/mile of expressway costs $100 million so why not a mile of track?
    Sure only looking at upfront costs a Prius is cheaper, never mind Toyota is selling them at a huge loss for PR.
    And that roads are subsidized, so no one sees the true cost of them either, the highway fund is $7 billion in the hole.
    What about all the people who don't want a car? its not all about cars, lets start building our cities around people.
    Instead of spending $400+ a month on a car, what if everyone could dump it and spend that money at a local business?
    Kind of changes the dynamics there doesn't it?

  21. Setenta y seis:

    So, you could buy each rider a Prius, huh? Maybe even pay for their gas? What about paying for the additional wear and tear on roads caused by all of these extra cars? Or build the parking garages that would be necessary? What about the hidden insurance and maintenance costs that always seem to be left out of any cars-vs-rail equation?

    And you say light rail hurts the working poor? Hasn't the reorientation of our entire American society to be car-centric hurt the working poor more than anything else, by requiring them to take on the additional expense of car ownership? Light rail isn't just about the transit itself, it's also about allowing for more walkable urban development, where people can live, work, and play without the required burdensome costs of a car, gas, insurance, and parking.

    I don't understand people like you who are convinced that we can keep adding more and more highway lanes for the rest of time. It's not a sustainable method of development, and the costs -- both economic and societal -- are very high.

  22. Stephen Karlson:

    Light rail in Chicago and New York? Not really. The Chicago and New York rail transportation uses either heavy rail (rapid transit in trains of up to 8 cars in Chicago and longer in New York) or commuter rail (using conventional railroad cars). You've got very thickly settled areas and high land values leading to office tower development and expensive parking, and resistance to widening the expressways or the arterial streets. The systems work, although they are tax sinks (although perhaps not as much as the roads are.)

  23. Daniel:

    I want to point out a lesser known pieces of information. The Phoenix Valley Metro light rail was placed along a very popular bus route called the red-line which had been transporting thousands of people daily between Phoenix and Tempe. This meant that when they put the light rail in they knew the ridership would be there right way. Additionally since the establishment of the light rail along the old red-line ridership has tripled. Sure probably thanks to yuppies, but also thanks to ASU students between the downtown and Tempe campus and also thanks to commuters from Mesa who utilize the park and rides to cut down on their driving commute. Point being light rail users are more than just yuppies or working poor.

    Second, the light rail is still in its infancy... ok maybe its not a baby but it is still relatively new. Ridership has been steadily growing each year. The primary reason being that the freeway commute is way more stressful than the light rail commute. I live in east Phoenix (close to Tempe) commuting downtown on the light rail takes about 40 min total. Trying to make the drive during rush hour traffic along the 202/i10 takes me about 35 min, assuming there are no accidents or lane closure issues and then I still usually have to find parking or pay for parking. The big difference is, when I take the light rail I can relax, read a book, play on my phone, check emails, have a phone conversation, etc. When I drive I have to be on pins and needles to make sure I don't get an accident, I have to speed up and slow down every 30 seconds, I can't do anything except drive and curse at the other morons on the road. Riding the light rail is not only more pleasurable and productive but it probably keeps me from developing an ulcer!

    Also, I'd be willing to be you don't live or work in Phoenix, or maybe you do and just haven't noticed. The light rail has been slowly but surely revitalizing downtown Phoenix. Three years ago at 5pm downtown Phoenix was practically a ghost town aside from basketball and baseball game nights when there would be a crazy packed influx of cars for a few hours. But rarely did anyone venture outside the sports arenas for food or entertainment. Now, a majority of the traffic is foot traffic. More importantly a number of small business such as local restaurants and bars have sprung up because people go to them and fill them out. Phoenix is starting to look like a real metropolitan capital hot spot once more. Phoenix in conjunction with the sports arenas, the ASU campus, the UA medical school, the concert venues, and the light rail have revitalized what was one a dying area because people can easily and inexpensively travel to and from downtown Phoenix.

    Finally, I realize I probably won't convince you to the benefits of the light outweighing the "costly" cons. But I will make one final point, a majority of the start up funds for the light rail came from a federal government grant for developing public transit. That money was specifically designated for such projects, if we had not taken advantage of it there would be no 1.4 billion dollars to buy people a Prius. While I understand your argument that its a waste of money (which I obviously don't agree with), its not like the money could have been used for anything else.