Out of Control

Traffic snarl: Rail or rubber?
How far should the subway system be extended? And should transit systems be subsidized? All this week, Ted Balaker and Bart Reed debate traffic, transit, and mobility in Los Angeles.
March 20, 2007

Today, Balaker and Reed look at expanding the L.A. subway system. Yesterday they debated how to reduce congestion Later this week they'll focus on transit funding, building roads and rail, alternatives to traffic, and crazy ideas to fix the traffic problem.

Run away, train!
By Ted Balaker

Bart,

Let's keep in mind that Los Angeles rail riders pay only about 3% of the overall cost of their trips. It would be easier to justify these hefty subsidies if the money were focused on helping the poor and handicapped—improved mobility works wonders for improving one's lot in life. But far too often officials spend lavishly on rail to subsidize those who already have good transportation options. And, as Environmental Defense's Robert Garcia explains, railophilia sucks up huge amounts of funds that could be used for moving people who need the only type of transit that really works for them, buses.

The price tag for the most extravagant bus system (think the Orange Line) is only about a third of what light rail costs, and light rail is downright cheap when compared to subways like the Red Line. (BTW, forget the original 376,000 daily ridership projection that was used to sell the line decades ago—we're still waiting for the Red Line to hit half of its reduced projection of 298,000). Public funds are always limited, and that means the more L.A. commits itself to rail, the less it can provide widespread transit service for those who need it most. Yet L.A. officials still back pricey rail systems because they regard them as the best way to lure relatively well-off suburbanites out of their sedans.

The typical Metrolink rider has an income of $65,000 and owns at least one car. The typical bus rider (chapter 12) has a household income of $15,000 and owns no car. But guess which way the subsidies flow? Each new bus trip costs taxpayers about a buck; each new rail trip costs about $21.

City Hall ribbon-cutters love rail, but it receives a much chillier reception among those with expertise in cost-benefit analysis. I was the lead author of a literature review (pdf) that examines economists' views on rail transit. Economists are fascinated (and annoyed) by how rail remains politically popular despite persistently failing to make good on its core promises, such as cutting congestion, getting motorists out of their cars, and cleaning the environment (operating big, nearly-empty trains wastes plenty of energy).

Economists from the Brookings Institution and UC Berkeley recently published an especially telling paper that helps answer the "how much more rail?" question. The authors examined 25 rail transit systems nationwide to figure out whether rail transit is socially desirable. And they didn't simply point out that all systems lose money; rather, they examined the larger societal benefits that rail backers often tout.

Their conclusion: "We find that with the single exception of BART in the San Francisco Bay area, every U.S. [rail] transit system actually reduces social welfare" (emphasis in original). They find that the yearly drain was typically north of the $100 million mark, and peg L.A.'s train drain (Metro) at $125 million per year. The public mistakenly assumes rail improves social welfare because "supporters have sold [rail systems] as an antidote to the social costs associated with automobile travel, in spite of strong evidence to the contrary." Their suggestion: stop building rail. I'm inclined to agree.

Ted Balaker is a policy analyst at Reason Foundation and co-author of the book The Road More Traveled: Why the Congestion Crisis Matters More Than You Think and What We Can Do About It (Rowman & Littlefield 2006).

Magical thinking theories

By Bart Reed

Ted,

Reading your constant inclusion of pointy-head wonk statistics about how transit is bad and costly but private cars are free is really making my head spin.

And, besides, actually responding to facts can be very tough. So, instead, personal insults. It is often a very effective diversion.

Ideally, any form of public transit, be it rail or buses, should serve where a good percent of the public travels, such as our dispersed areas with high job densities. These systems must transport passengers in a speedy fashion with frequent, punctual service.

True, however, at some point, it is rather important to realize that about the only type of travel where transit can really work is where there are a whole lot of people traveling along the same alignment at the same time – such as to CBD’s – which, unfortunately for transit promoters, has been steadily declining for decades. Rather than trying to respond to these new travel patterns, including the huge suburb-to-suburb commute, the response of rail promoters is to continue to pitch yesterdays solution to tomorrow’s problems – while ignoring the transit needs that can actually be satisfied reasonably well at reasonable public sector cost by non-rail transit modes, particularly if they are well managed and take advantage of proven cost-effectiveness techniques, such as competition in the provision of services.

You insist it's noble to give mobility options to "the poor and handicapped," but these folks are entitled to better-quality transportation just as those "relatively well-off suburbanites" who, you say, love being stuck in "their sedans" because alternative transportation choices don't exist.

The problem here becomes one of, first, what is “better-quality transportation,” and second, what are the costs?

The first attribute of “better-quality transportation” is that it actually works for the people who need it. It needs to have a stop somewhere reasonably close to where these people live and another somewhere reasonably close to where they want to go. It should operate fairly frequently and have good coverage of the hours of the day and the days of the week. It should be safe and secure and people should feel safe and secure.

Note that none of this really has much at all to do with mode – and, in fact, the professional literature is filled with studies that show that mode is simply not very significant in individual decisions to use transit. It is the attributes of the transit trip, the qualities listed above and others, that are key.

But, you say, this list ignores that all-important attribute of rail, SPEED. Yes, speed is important, but, it does not necessarily mean that rail means greater speed. U.S. light rail has an average speed approaching 16 mph and heavy rail about 20.5. Unfortunately, this is not the speed of human travel, it is the speed of the vehicles, and does not include the time to access the stations at trip ends and the wait for transfers – it takes approximately two minutes simply to walk from the sidewalk by a subway station to the station platform, plus another two minutes at the other end of the trip.

Bus transit can be very competitive with rail modes in both transit vehicle speed and particularly in human travel speed. If light rail is a viable option in a transit corridor, bus rapid transit should also be studied. Not only are the transit guideway speeds comparable, the same bus that operates on a dedicated busway can also go through neighborhoods, allowing people to avoid the bus-to-light rail transfer, or the drive to the station. It is almost impossible to spend half as much to build a BRT line as a LRT line, as MTA is proving time and time again.

(Yes, we know that the Los Angeles County Metropolitan Transportation Authority is now trying to get people to call it, “Metro,” but they are not fooling anybody.)

Of course, in many situations, there is not really much need to even go to BRT. After spending years claiming that the Orange Line BRT would be much faster than the logical alternative, Rapid Bus, in its very last revision to the Revised Environmental Impact Report, MTA actually showed Rapid Bus on Victory/Lankershim to be faster than the Orange Line. The wonder of Rapid Bus is, with guideway costs per mile of under 2% of what it costs for BRT and 1% of what it costs for LRT, dozens of Rapid Bus lines can be implemented, providing a faster transit option for dozens of times the number of people who could ever easily access rail transit.

The problem with rail is that it costs so much that there is simply not enough money to build very many stations. After over two decades, and over ten billion dollars of capital expenditures, there are currently 62 light and heavy rail stations (and, in the interests of full disclosure, 17 more currently under construction) in MTA’s 1,224 square mile service area, leaving the vast majority of greater Los Angeles many miles from an urban rail station.

There are over 25,000 bus stops in that same service area.

Quality transit service begins with there being transit service, somewhere transit users can actually get to it, and the huge cost of building rail lines means that the overwhelming majority of Los Angeles residents will never have a rail station that is useful for their daily travel – and every billion dollars spent on building more rail means that transit that could actually be useful to those who need it will not exist for them.

Subways should be built where traffic is too dense to make private vehicles, buses, and at-grade light rail trains reliable. They should also be built when there are no available right-of-ways and development is too dense to acquire property for surface routing.

… and they should only be built where they make sense as transportation options, where there are no superior options – and, at well over $300 million a mile, and climbing, frankly, there aren’t too many of these places in existence.

Admittedly, subways are inherently costly pieces of infrastructure, but you ignore the $40 million-per-lane-mile costs for roads whose capacity is a paltry 1,650 vehicles per hour.

Even in Los Angeles, $40 million per lane mile is an exceptionally expensive roadway – although, with MTA, certainly not unknown.

Your 1,650 vehicles per hour is a low figure (rush hour traffic in LA will often reach 2,200), which appears to imply that you are discussing an high-occupancy vehicle (HOV) lane or an HOV/Busway. Are you aware that the El Monte Busway/HOV actually produces more transportation work (passenger-miles) at peak hour than all four general purpose lanes on the San Bernardino Freeway (I-10) where it runs – combined?

Oh, and an even larger multiple on what the Blue Line, American’s most heavily utilized light rail line, produces?

Or, perhaps you meant an High-Occupancy Toll (HOT) lane – you know, one of those additions that can pay for a substantial share of their own cost of construction, even the entire cost in some cases, through user fees – and can be used by transit buses as well?

Transit buses on the El Monte average over 55 mph – compared to slightly over 40 mph for Metrolink, which is most comparable rail mode.

You make wild claims that Los Angeles can get a more attractive bus system for maybe half the capital cost of rail, …

Oh, God no, the bus capital costs would not be anything remotely close to that – the MTA capital investment in rail to date, to carry about 15% of the total MTA ridership, would fund the entire MTA bus fleet, plus all the support facilities that go with it, almost ten times over.

The necessary capital renewal and replacement costs for the existing MTA rail system would pay for a bus fleet well over double the size of the current bus fleet and all the capital support structures – forever.

… but ignore the huge bus operating costs.

Well, I’d rather not, considering that bus operating costs are so much lower than rail operating costs for comparable types of transit services – although I can see why you would want to.

Oh, I understand, this is the MTA rail investment philosophy – we were placed on Earth to build rail, and, in order to afford to keep doing this, we must cut out unnecessary frills – like carrying passengers on buses.

Therefore, it is vital that we reduce the number of bus passengers that we carry by raising fares, reducing service, and the most unbelievably ill-devised bus restructuring plan – “Metro Connections” – ever devised.

Yes, as long as you keep pumping money into it – MTA’s rail capital renewal and replacement cost projection for the years 2001-2025 is $4.7 billion. This has nothing to do with building any new lines or stations, it is simply to keep what was built operating at an acceptable state of repair and operations.

The existing Red Line was built for $4.7 billion, but more than 304 million boardings have been made in 14 years of operations. 40 million boardings were made in the last fiscal year alone, and the number keeps climbing.

First, the Red Line total cost was more like $5.5 billion, but who’s counting (MTA surely isn’t)?

Let’s take the entire MTA bus fleet – approximately 2,700 vehicles – add the costs of bus maintenance and operation facilities, and cost them all at current market prices (the Red Line costs, by the way, are from the early 1990’s, on average) – and we get nowhere remotely close to $2 billion.

In 2005, MTA buses carried more than ten times the number of Red Line riders.

AND the number keeps climbing, or at least it will until MTA cuts bus service enough so it can build more rail lines.

Because subways are underground, they are not constrained to follow above-ground physical features such as mountains or street grids. Thus, you could build subways in any direction necessary.

Isn’t it amazing what it possible when money is no object? Other people’s money, that is.

Rail in general also promotes better uses of scarce land by bringing various services close to stations and making them accessible to pedestrians. You hate the fact that this discourages auto travel and in turn saves consumers on fuel, parking and maintenance costs. Time on transit can be spent productively, unlike driving time. You should try Metrolink and see how many laptops are in action.

Yep, isn’t it wonderful how many of the fewer than 20,000 daily Metrolink commuters can do stuff like this, while the ten million who have no ability to use Metrolink can’t.

By the way, it is not that anyone “… hate(s) the fact that this discourages auto travel” et al, it is that there are so many far more productive ways to spend taxpayer dollars.

For those who want to live next to rail stations to reduce their driving, well, I say, good for you, and I sincerely hope that you have this option … as long as you pay the costs, and not stick the taxpayers with them, of course.

And, if the vast majority of people world-wide have the American/Australian/French/German/Indian/Turkish/You-Name-Your-Nation Dream of a house in the suburbs, well, if you don’t mind, I’d like for them to also have the same option of being able to fulfill it.

… as long as they pay the costs – and ONLY their costs.

As a sharp divergence between your magical thinking theories and actual practice, Metro spent over $1 billion on the bus system during the Consent Decree years yet has not achieved meaningful ridership increases, hovering just over 1 million boardings a day for several years now.

Odd, Special Master Bliss, the man both MTA and Bus Riders Union picked to oversee CD Compliance, didn’t see it that way.

In fact, here is what he said about it in a published Order to MTA to add bus service to comply with the requirements of the CD:
“… in the six-year post-Consent Decree period, the MTA has gained a total of 81.6 million annual riders. … This in stark contrast to a loss of 133.6 million annual passengers over the eleven year period preceding the Consent Decree.”

By the way, MTA is currently carrying about 1.25 to 1.30 million bus passengers a day, or, a bit over five times the number of rail passengers.

Buses and even those on dedicated busways lack the capacity to handle passengers, as demonstrated by the too-popular Orange Line.

You know, it is interesting that the Orange Line BRT and the Gold Line LRT are both just under 14 miles long, but, even though the Gold Line has been in operation a few years longer, the Gold Line carries about 17,500 daily passengers, while the Orange Line carries 21,400.

You see, before you get into the question of how much a line can carry, you first have to address the problem of, does anyone really want to ride it?

There is no real functional capacity problem on the Orange Line – all MTA has to do is to “platoon” buses, running two of them together when the demand is there.

The stations were specifically designed to be more than long enough for MTA to do this, one just wonders why MTA isn’t.

Also, the cost of moving passengers by rail is historically less than doing the same on buses on a mile-per-mile basis, largely due to constantly increasing labor costs.

Ah, yes, the great, “rail has higher capital costs, but the operating costs are lower” myth.

Here’s the deal – first, when rail capital costs are 70-85% of the total costs, while bus capital costs are more like 15-30% of the total, yeah, I can see why rail proponents want to focus on something other than capital costs.

However, when you compare bus and rail in similar transportation corridors, and get away from all the lesser used lines where bus is utilized because rail is way too expensive to be even considered, you know, guess what?

Bus is generally a lot cheaper.

For example, on Line 720, the Wilshire/Whittier Rapid Bus line, the operating subsidy per passenger is well under half what it is for MTA light rail – and capital cost is chump change for this bus line, compared to any type of rail.

Worse yet, buses must be replaced after 500,000 miles, which is only 5 years of service for Orange Line buses, versus 30-50 years for rail cars.

This is a very misleading – and factually incorrect – comparison.

OK, let’s do the math. (Some of the numbers may be a bit off, but not much.)

On weekdays, MTA operates 171 Orange Line bus trips in each direction, for a total of 342 in both directions. The end-to-end distance is approximately 13.8 miles, so this works out to approximately 4,720 revenue miles each weekday.

MTA utilizes 26 buses at peak, and, if we assume the usual 20% spare factor, that means that there will be 31 buses (I gave you the benefit of the doubt and rounded down), in total operating on the Orange Line.

Divide 4,720 miles per working weekday by 31 buses and you get 152 miles per bus per day, working weekdays.

MTA operates fewer Orange Line trips on Saturdays, Sundays, and Holidays than it does on working weekdays, but, let’s give you the benefit of the doubt and assume that the number of trips is the same each day.

So, multiply 152 miles per bus per day times 365 days per year and you get 56,000 miles per bus per year (I rounded up again).

Let’s add in some additional miles to get from the bus garage to the Orange Line, driving around the bus garage, and whatever, and call it an even 60,000 miles per bus per year.

That’s very high for MTA bus, or any bus, but that’s a lot less than 100,000. If MTA actually took these buses out of service after 500,000 miles – which is highly doubtful – that would be a life of about 8 1/3 years.

MTA paid $633,000 for each bus. Over an 8 1/3 year life, that’s $76,000 per year.

Each bus has 57 seats. That’s about $1,333 per bus seat per year.

Now, let’s look at light rail cars. MTA’s newest light rail car, the Ansaldobreda 2250, cost $2.9 million each. Using the Federal standard 25-year life (yes, it would be possible to operate these cars for 50 years, but only with at least one complete rebuild, which would cost a lot of money – and you can rebuild buses to extend their useful lives as well, so let’s not go there), that’s $116,000 per car year of life.

Each car has 76 seats, so that’s $1,526 per rail seat per year.

Ah, Bart – bus wins.

Now, it is certainly true that you can get a lot more standees on a light rail car than on a bus, so, if your idea of good transit is to stand while you ride, please, go ahead and make the adjustments that you’d like to these figures. Of course, I could come back and talk about the time value of money, where the full $2.9 million for the light rail car has to be paid out all at once, but, for the buses, it is $633,000 spread in three installments over almost 17 years, and there is a time value to money, but, really, do we really need to get into this kind of detail?

After all, when the cost of light rail guideway is well over double the cost of a bus rapid transit guideway, and the cost of bus and rail cars are a pretty small portion of the total cost of the respective guideways, can we just say, we’re done here?

Ted, reality is quite unlike the strange social theories you promote, as you could move more people with less operating and capital costs on rail than on bus.

A fanciful bus system would be more costly and less attractive than a rail system in the long run. But for you, a supporter of the concrete, asphalt and rebar industries, efficiency is not part of the mantra.

You know, you can show people the facts, you can show them again, you can try one more time – and then, you begin to get the idea that it doesn’t matter how many times you present the facts, some people just are not interested, they have made up their minds.

Bart, you can keep repeating that rail is cheaper, but that just doesn’t make it so.

In virtually every case, you compare bus and rail in a similar transit corridor, and the capital and operating costs for bus will be far lower than those for bus in most cases.

I’m sorry you don’t like the way the facts play out, but that doesn’t change the facts – bus is a whole lot less expensive to build and operate in the vast majority of cases, which means that a whole lot more people can be carried on bus than on rail.

Repeating false statements many times, and calling people names, simply does not alter that.