Monday, August 18, 2008

Miles per Gallon on the Bus

A week or three ago I read that if you divide the number of passenger miles ridden by the number of gallons of fuel used by the public transportation systems of many cities, you end up with a pretty lousy number (see here for the full scoop). That makes sense if the public transportation system runs lots of mostly-empty, big, heavy, diesel-hungry buses or trains or ferry boats.

Which made me wonder: how does our local public transportation system fare?

The National Transit Database gave me the number of passenger miles ridden in 2006 for the entire PVTA system: about 31 million miles.

And an email to the nice folks at the PVTA got the number of gallons of fuel they used last year (2007): about 1.5 million gallons.

Divide miles by gallons and you get about 20 MPG (that's not quite right-- I should divide 2007 miles by 2007 gallons; I'll update this post when the 2007 ridership miles are available).

20 MPG is lousy. The CAFE mandated fuel efficiency for single-passenger vehicles is 27.5 MPG. Unless you own a gas-guzzler, and especially if you're going somewhere with a friend, then the raw data says driving will use less gas than riding the bus.

But that's not really right. The "marginal cost" of riding the bus is probably pretty close to zero; add another person to the bus and it's not going to use a whole lot more fuel. At least, until the bus gets too crowded and you need to buy and run another bus to fit everybody. Then the marginal cost of that one extra person is ginormous.

I think the lesson is that it is not as simple as "bus good, car bad." Running lots of empty or mostly empty buses is worse for the environment (and our economy) than running smaller, more fuel efficient private cars or taxis. If we really care about the environment, then the fuel efficiency standards that we set for cars should also be applied to public transportation. Why not tell PVTA: "You must be at least as fuel efficient, on a passenger-mile-per-gallon basis, as the CAFE standard for single passenger automobiles (27.5MPG)."


Anonymous said...

I'm utterly stunned by these numbers, so much so that I think that there must be some flaw in your calculations. Perhaps the discrepancy between 2006 passenger miles and 2007 gallons explains it, although I don't know why that would be.

I'm hoping that someone with knowledge will see your post and have something worthwhile to add.

20 passenger miles per gallon is pathetic. I recognize that much of life is counterintuitive, but that's too counterintuitive for me to accept without more.

Thank you for raising the issue, however. It does raise the spectre of lots of empty busses running around waiting for riders.

Rich Morse

Gavin Andresen said...

My initial reaction to the numbers quoted for other public transit systems was disbelief, too. That's why I decided to derive the numbers for PVTA myself.

I'll keep on it, and will see if I can get the 2007 numbers and then double-check them with the PVTA. I'll also see if I can find out how they calculate passenger miles (maybe lots more people are riding than they think, or are riding farther, although that seems unlikely).

It shouldn't be too surprising that PVTA gets lousy gas mileage, because I don't think they're judged on that metric. Incentives matter.

robert said...

These numbers are neither surprising nor unreasonable. Gavin's marginal-cost analysis is compelling, and it argues for more folks to use the public transit service we already have.

The "big" versus "small" bus issue is more complicated: it's a defensible decision by the operator to have "interchangeble units" that are well-built and easy-to-repair, rather than a panoply of models that pose major operational and maintenance problems. I haven't always agreed with the decision to use big buses and have argued for the development and utilization of smaller buses. This may be institutional inertia more than a reasonable business practice. It's an issue worth pursuing further, especially if we want to see any transformation of the equipment to, say, hybrid-electric buses with regenerative braking.

But, getting back to the gas arithmetic, I am not sure the comparison with a 27.5 mpg car is as meaningful as you believe. First of all, only a fraction of the miles driven in a private car are "revenue" miles (I can't think of a better term, but you should be able to figure out what I mean):
the miles going to and from the mechanic, the car wash, the oil-change place, the gas station, etc. should be discounted.

Would you be willing to do the same experiment with school transportation, in all its various forms? For example, there are parents who drop off and pick up their child daily: do you count the round-trip miles or the the one-way miles? If the latter,
your 27.5 mpg car is getting less than 14 "revenue" miles per gallon. (Anticipating some
counter-arguments here, such as the trip to the school being part of another trip, one counter-counter-argument is that almost all car trips under 10 miles can be replaced by walking or biking anyway.) A better comparison might
not be to the hypothetical 27.5 mpg private car, but to one of the alternatives Gavin suggested: what are the actual passenger mile per gallon for
a typical taxi company around here? Anybody wish to wager under 10 mpg?

Of course there are equally-compelling "social justice" reasons for having a well-functioning public transit system. We don't force individuals to hack their own trails through the woods and fields and swamps every time they wish to travel. Instead, for reasons of political economy (and social justice), we instead provide a publicly accessible and publicly maintained system of roads and sidewalks and paths. That hasn't always been the case.

Indeed, it's tempting to argue* along the same lines as Gavin that we should revert to what we had in many places, including Amherst, up until the middle of the last century: a network of railroad tracks and some narrow dirt/sand/gravel roads feeding them on which folks could walk or bike or ride a horse. These would all much cheaper to (re)build and maintain than the vast sprawl of asphalt roads we have now....

*perhaps _reductio ad absurdum_ ;-)

Gavin Andresen said...


Excellent question about taxi gas mileage; I'd just note that taxi drivers have a very direct incentive to save fuel (bus drivers don't), so I predict we'll see a switch to fuel efficient taxis a lot quicker than fuel efficient buses.

And yeah, I think it would make sense to hold the school transportation system to the same fuel economy standards as private cars. I don't think it makes sense to say HOW that should be accomplished-- maybe more, smaller schools within walking/biking distance of the students makes more sense than hybrid school buses-- but I think policy consistency and fairness is important.

I'm sympathetic to the social justice argument, although I'm also very wary of unintended consequences. My gut reaction is "Keep it Simple, Stupid" -- if we want to subsidize the poor/disadvantaged, then I'd rather we gave them some kind of Transportation Vouchers that they could use to pay for whatever type of transportation works for them-- be that a bicycle, bus, taxi, train, jitney or (I can hear you cringe) to pay for gas for their very own car.

In fact, if we REALLY want to have a big impact, let them decide to move downtown and walk or bicycle everywhere and use the Tranportation Voucher money to buy themselves an iPod and some Milli Vanilli MP3's.

As for going back to gravel paths: it WOULD be fun to have more places to ride my mountain bike...

LarryK4 said...

Yeah,and Amherst taxpayers spent around $120-K this year to fund outreach buses with less than average ridership. Yikes!

PS: one of those stops is directly in front of my business

robert said...

Hey Larry (and anyone else paying attention),

When I chaired the PTBC, you and I chatted quite a bit
about how to encourage folks to use the bus to access your business and others nearby. We talked about advertising your business on the buses, and putting a bench out at the bus stop with advertising as well. I thought you were genuinely interested, and don't know if you ever followed up wiith that, but I am sure
that if you gave Jane Ashby a call, she'd be willing to
help too (though I am aware of what Ronald Reagan
said about being wary of "help" from the government ;-).

The Amity Shuttle (the bus route which I designed
design and helped implement during that time) has become
one of the most popular (and therefore cost effective)
in the entire PVTA system. It would be nice if we could decipher what feature of the route, schedule, amenities along the route, interlining possibilities and so forth have contributed to that.

Ultimately, I am scientist and a pragmatist, very open to new ideas and trying new things. I am also willing to discard things that don't work (with a few exceptions, one of which you already photographed ;-)))!

For example, I think Gavin's idea of transportation vouchers is a good one, as long as there is some taxi service out there (rumors abound that what we have is unreliable, and I can vouch that it's not cheap). That makes sense for outlying towns like Shutebury and Leverett, but I don't think those can replace fixed route bus service in Amherst.

Realistically, I think we should be pressing for PVTA to buy somewhat smaller, more fuel-efficient, hybrid-electric buses. The USPS is supposed to be experimenting with these in their
fleet: since they stop and start often, the regenerative braking features have similar value to their operations as they would to transit (and school) buses which also make frequent starts and stops.

I hope that these are common in the next decade. But I also hope that the bus system is discarded
in the meantime - it's very hard to resurrect once it's gone (see, streetcars/trolleys).


P.S. It is too bad more folks aren't like you and Gavin (and Denise and me) and willing to ride a bike for a lot of short trips, but we can't expect everyone to do that. We need to help folks without cars get around inexpensively, and so far,
buses are the best option for that (IMHO :-).

robert said...

Of course, there was a missing "NOT" before "discarded" in the last paragraph!

LarryK4 said...

I liked the original closing paragraph (minus the "not")

His Lordship Gerry Weiss has used that argument to maintain the Black Hole, Money Pit out at the Cherry Hill Golf Course.

Anonymous said...

In reply to CAFE standards for school busses and having smaller "closer" schools. In my neighborhood I see parents driving thier precious HIGHSCHOOLERS to the Busstop to be picked up. Now tell me that makes sense. These kids won't even walk 2-3 blocks to a bus stop!!!!!

jte said...

Gavin, though 20mpg might look worse than what people can get in personal vehicles according to the sticker mpg ratings, actual fuel efficiency of personal vehicles, on average, don't turn out to look so great. If I understand this Bureau of Transportation Statistics table 4-23 correctly, the top two rows show mileage efficiency achieved in the auto fleet, not just the rated efficiency of new cars. For 2006, passenger cars averaged 22.4mpg and "Other 2-axle 4-tire vehicles" averaged 18.0mpg. Put together, that looks like a statistical dead heat with the public transit's 20mpg.

This could well be the average rated fuel efficiency, rather than a measure of actual efficiency achieved by cars on the road. I'm going to do some number hunting for that statistic. If true (that it ignores actual achieved efficiency on the road), then it ignores the advantages of carpooling, as you noted yourself. Still, it does put into some question the degree of efficiency disadvantage you may have found for public transit.

jte said...

PS: this Table 4-24 gives the following calculation for "transit motor buses" in 2005:

21,825 million passenger miles
534 million gallons diesel
40.87 passenger miles per gallon

Can PVTA be that much less efficient than the average? Strange. Even if the BTS table is basically right (and my interpretation of it), remember, this is diesel we're talking about, so more carbon-polluting per gallon than gasoline. A rough equivalency in terms of carbon emissions would translate the above number to about 33 passenger mpg.

dale said...

This is all very interesting. WRT the comment about "car wash, oil-change" all adding up, this can't be more than 1% of miles driven. And in the same spirit, not all "public transit" is revenue generating either. The "outreach busses" probably likely are 50% or more "non-revenue generating. Given that drivers don't take their busses home at night, and municipalities don't have dozens of bus yards, it seems likely that core public transportation's "non-revenue" load is not any better than the private car, so what's the point of adding it to the debate?

But more fundamentally, why has nobody mentioned the marginal cost of peoples' time? There is a reason why collective transportation concepts of a hundred years ago don't work today: the marginal cost (either lost productivity or lost leisure time) is far more valuable today then it was in the good ol' days. This seem to me to be a frequent and fundamental blind spot in the eye and brains of those who champion public transportation as the solution to global warming.

If you want what worked 100 years ago, please tell me how you plan to loose the ~5 billion people on God's Green Earth that depend on the technology and productively that public transportation can't support. Face it, modern energy technologies, transportation systems and agriculture methods are what let the earth's population grow above 1 billion. Take it way, and you've got to explain how we get back to that number. In a thinking democracy, that seems like a harder proposition to sell than what McCain's gamely sloggin' for :-)

Gavin Andresen said...

It's odd, I can never predict beforehand which blog entries will generate the most comments...

A quick update: I haven't heard back from PVTA; I need to send them another email or maybe call to try to confirm the numbers.

And dale: I think you've hit the nail on the head; public transportation isn't more popular because people value their time (they also value flexibility, predictability and convenience). I looked at the cost for me to commute to work (it's only 2 miles) in a 'Bus Compute: Does Not Compute" post last year, and figure it's somewhere between $8 and $20 per day MORE expensive for me to ride the bus, depending on how I value my time.

That said, cars and driving are subsidized in lots of big and little ways, and alternatives are often regulated to death. I'd like to see a more level playing field for ALL of the alternatives...

Anonymous said...

Well the obvious response to this post to me is Ride the Bus! Obviously efficiency increases greatly with more riders. Also rarely is another bus added for one more rider. And if that happens two buses with a total of fifty people is a hell of a lot better than one bus with 7 people.

I agree we need more fuel efficient public transportation this is clear. But is it easier to change our own lifestyles of convenience or is it easier to change public policy. One of these is clearly more in our domain of influence than the other, I will leave it to you to figure which one that is.

Also I see some sense in dale's statements about modernization, but if you understand that when Oil production peaks as it may well have already done then all of this productively (for what its worth) is going to be in crisis. Underwear from China??? I don't know, just thought I'd chip in.