Sunday, September 20, 2009

Is this really a success?

Jennifer Steinhauer thinks that light rail in Phoenix is a success. Her evidence is that while projected daily ridership was 26,000, it has clocked in at 33,000. The reason that ridership is better than forecast is because of weekend riders--people are using the line for pub crawls, among other things.

But ridership of 33,000 per day translates to about 12 million per year. The system cost $1.4 billion to build, not including lost revenue to businesses located along the line (which was probably largely displaced to other business). Let's assume that the cost of capital to Phoenix is 6 percent and that the system depreciates at 2 percent per year. Therefore, the capital costs of the system are about $110 million per year.

Assuming fares cover operating costs (and they almost certainly don't), this means that each ride is subsidized to the tune of more than $9, and according to the article much of the subsidy is going to entertainment.

I think it is great that some people in Phoenix are leaving their cars at home when they go out drinking. But I would guess that a free shuttle service going from bar to bar would have cost the taxpayers of Phoenix a lot less money.


Richard H. Serlin said...

Ok, but living in Tucson, what do you think of a high speed train connecting Tucson and Pheonix, or for that matter L.A. and San Francisco? Including all benefits, like the externalities.

Bob Sharak said...

First, 7,000 daily rides out of 33,000 is 21% and 66,000 weekend rides out of 231,000 weekly rides is 29%. Even if all the weekend riders are pub crawling it's not fair to attribute the full subsidy to "entertainment."

Second, your cost of capital is too high. Munis and Treasuries are at most 4.5% - even if one goes out 30 years.

Third, the American Lung Association conducted a study on the 25 most ozone-polluted cities in America. Phoenix ranks 9th. Your analysis does not take into consideration any health or environmental benefits from taking up to 16,500 cars per day off the road (33,000/2-way trip).

Lastly, your analysis rests on the assumption that the substitute (road driving) is subsidy free - which it decidedly is not. Urban roadway expansion cost - let alone maintenance - is extremely expensive. Moreover, during periods of high traffic, taking even a small percentage of autos off the road can yield outsized benefits in terms of decreased congestion, fewer accidents and shorter commute times.


Jeff Adams said...

Hello Richard

Your Blog on Real Estate and Urban Economics attracted me very much.

I could able to see your comment what u gave on my reply that you are agreed to add my website link in your blog

Jeff Adams said...

Hi Richard

View my real estate news

Building business is booming in certain markets

Please post your comments

Jeff Adams said...

Dear Sir,

I’ve been looking through your blog over the past few days and I found your blog to be very informative.

I was able to find everything I need from real estate to accident information from links in this site as well as by following external links from this blog.

I can see you have a list of real estate sites useful both for your users as well as end users who visit I have a site related to real estate with useful information to our visitors.


It would be great, if you could spare a few minutes visiting my site. If you find my site to be informative, it would be nice if you could provide us a link.

It would be great pleasure if you can add this link in your blog so that it can benefit our visitors.

Thanks for your help and consideration,

Laguna Beach Realtors said...

I still believe that the entertainment has to be subsidized because this is something that provides refreshment to the lives after people have been fed up of the tight work schedule and more so after the prolonged recession.

MJ said...

You can add about $2.00 per trip to the subsidy for operating costs. The 2009/10 budget forecast for Metro indicates projected revenues of just under $9 million and operating costs of about $33.7 million.

I always find it odd that light rail advocates consider a project a success if actual ridership exceeds forecast ridership. Demand forecasts are usually revised multiple times between when the decision is made to go forth with a project and when it is open for revenue service. The revision is usually downward, making actual demand look comparatively good.

It is worth mentioning that the travel demand forecasting models that produce ridership forecasts are subject to a large error tolerance, usually on the order of 25-35%. The 33,000 boardings per weekday are within about 20% of the forecast demand following the opening of the line, so it is not all that surprising.

The costs are quite troubling, though, as Richard outlines. He estimates (charitably, I believe) a subsidy per boarding. Of course, only a fraction of the light rail riders are former auto users (hence it makes no sense to assume that 33,000 -- or 16,500 as Bob mentions -- cars have been "removed from the road"). The article is silent about how large this fraction actually is.

I happen to think these little accounting exercises are worthwhile -- they provide some needed perspective on how effectively we are spending public resources during times of large deficits.

MJ said...

I also don't believe that yields on treasury bonds are an appropriate measure of the opportunity cost of capital. For one thing, the current absurdly low interest rates cannot last, especially when economies start to rebound and the demand for private capital reemerges.

Secondly, these types of projects are inherently risky (in terms of demand and cost forecasts), and the social discount rate chosen for them should reflect this risk.

Thirdly, if government has nothing better to do with people's money than invest it in 30-year Treasury bills, it should leave it in the hands of private citizens.

Treasury yields provide at best a lower bound for the opportunity cost of capital.

Bob Sharak said...


A few points in rebuttal. If you have a URL for the operating cost I'd like to see it. Does your number include debt service? One thing many anti- rail folks ignore is the enormous capital cost for road building, rebuilding and enhancements. Unless the freeway is flipped to a toll road these are costs are not included in the comparison between roads and train transit.

RE: Cost of capital. If not bond rates - then what? Calculating the social discount rate gets squishy pretty fast. And thanks for helping make my case by noting the low bond rates. If the project is bond financed to any degree (which most are - including the federal subsidy) locking in today's low bond rates for 30 years means the time has never been better to finance the project.

Lastly, you note that "only a fraction of the light rail riders are former auto users." Unless you know something I don't about Phoenix (I've visited but never lived there) - from where did these people come? Phoenix is a car town - American suburban sprawl at its best. Did these people walk to work? Even if they're new commuters, which is a possibility, their alternative is taking a car or not taking the job. They might take a bus; however, outside a few densely populated metros bus service isn't very good. Where I live in SE Virginia it can take 2 hours versus a 30 minute car ride. plus buses too have a subsidy. And, take care to factor (somehow) increased access into your social discount rate.

Gopal ch. Sahoo said...

actually this is realy success

visit us

MJ said...


The data on the operating costs can be found in this board presentation (page 62):

And no, this number does not include debt service.

Nowhere in the article did I read that a freeway link was being proposed and evaluated as an alternative.

But since you mentioned it, capital costs of roads are typically paid for through a combination of federal and state fuel taxes, along with some combination of local matching revenues, usually in the form of vehicle sales or registration taxes, toll revenue or general sales taxes. These costs are typically not that large when they are spread across the much larger base of users typical of highway networks.

As for the appropriate measure of capital cost, I maintain that a social discount rate indicative of the opportunity cost of capital is appropriate. You can decide for yourself whether to assume the displacement of private consumption or investment, but that is what sensitivity analysis is for. Simply using bond yields because it is easier is not defensible.

As I mentioned before, these projects have risks that are non-trivial and should be priced accordingly. In this case, it probably wouldn't matter whether the interest rate were 4.5 percent or zero percent -- you just can't make it pencil out.

To your last point, most of the riders probably were former bus riders. This is not so hard to believe. In fact, it has been the case with most of the new light rail systems in the US.

Nor are bus users hard to find in Phoenix. Valley Metro recorded about 57 million passenger boardings in 2007. The City of Tempe (Tempe in Motion, or TIM) recorded another 7 million or so. Like most cities, Phoenix has a relatively dense core, where bus service is feasible. The region also has a fairly continuous grid street network, which is nice for providing bus service. Lastly, the city also has a large, low-income immigrant population, which is typically a natural market for public transit.

Yes, their bus service is subsidized as well, but at a rate that is much more palatable than the $1.4 billion light rail line. The operating subsidy for Valley Metro in the City of Phoenix was about $1.80 per boarding in '07. Bus systems have some capital costs as well, but they are modest relative to new rail links.

Colombia Commercial said...

Colombia Commercial is an independent real estate brokerage, committed to providing outstanding service and value to buyers and sellers.

Anonymous said...

yo... bookmarked thread ))

zhuzhulee said...

PDF Créateur
TIFF en PDF Convertisseur
Texte en PDF Convertisseur
PowerPoint en PDF Convertisseur
image en PDF convertisseur
Excel en PDF Convertisseur
CHM en PDF Convertisseur
Word en PDF Convertisseur
djvu to pdf
pdf to png