Transportation microeconomics 101

Reformatted in August 2012 to fix some dead links and make it look slightly nicer. Keep in mind this was written back in 2005; I have changed no content beyond this first paragraph.

I talk about this enough that it might should be its own category.
Problem: Bozoes in government, in the media and elsewhere think about transportation at only the highest level – where you’re moving thousands of people around the city. This usually ends up producing plans which fail spectacularly at serving their intended constituents. Since this often boils down to money, I’ll call this “transportation macroeconomics” even though most of the people who do it aren’t thinking about economics. (Hint: they should be).

Solution: Transportation microeconomics. Whenever evaluating some transportation plan or change in economic conditions, take a couple of representative ‘use-cases’ and analyze the economics of their decision-making at their local (individual) level.

Example 1: Toll Roads. Local activist Roger Baker has been on my case on the austin-bikes email list for talking favorably about toll roads (as the least noxious of the two realistic possible outcomes – the other one being that all of those toll roads are built anyways, but as free roads). I’m going to be more favorable to him than he is to me, and construct an argument based on his stated motivations (he likes to accuse me of being a toll-loving road warrior). Roger’s point is, basically, that the toll roads won’t have enough traffic to pay off the bonds once the “oil peak” causes gasoline to get even more expensive than it is now. He’s definitely one of the SOS-bloc (don’t build these roads at all because they promote sprawl and hurt the aquifer) rather than the free-roads-bloc (“double taxation!”) best exemplified by Brewster McCracken and Gerald Daugherty, who will end up getting central Austin to pay for these roads via property and sales tax kick-ins.

So, is Roger right? Would expensive gasoline lead to an exodus from the suburbs and a default on the bonds which back the toll roads? Or am I right – that the traffic which today would fill the toll roads in a second isn’t going anywhere even as gasoline gets more expensive. Let’s look at a use-case.

Joe Suburban drives his Suburban on a 30-mile round-trip every day from western Travis County to his job in one of the southern suburban office parks. He gets roughly 15 mpg on this commute and pays $2.00/gallon for gas today. By some calculations, which include depreciation, he pays a hefty price for his commute even today, but I categorically reject the idea that suburbanites will reduce the number of vehicles they own (barring catastrophically high gas prices), so depreciation should not honestly be part of the cost equation. Using my handy depreciation-free cost estimator, Joe’s daily commute cost is $2.79 today (remember, no tolls yet). Is that enough to convince Joe to carpool? Not today it isn’t. Is it enough to convince him to use transit? Even at the discounted rate, the bus trip from the park-and-ride at 290/71 costs him probably an hour extra time per day, and still a buck ($1.79 savings at the cost of an hour). This assumes he even HAS a transit option, of course. Most suburbanites don’t.
Suppose gasoline DOUBLES in price – to $4.00 a gallon. Joe’s daily commute cost (with new tolls of, let’s say, $1.50/day) is now: $6.91/day. His “transit cost” is now $5.91 for an hour of time, assuming no rise in bus fares (unlikely). Still not very attractive, I hate to say.

All right, suppose gasoline TRIPLES in price – to $6.00 a gallon. Joe’s cost is $9.58/day. Transit option would save $8.58 a day at the price of an hour. I hate to break it to you, but most suburbanites would still drive at this cost. Bad news for Roger: $6.00/gallon gas is roughly equivalent to $160/barrel (working backwards from this logic which is admittedly crude). That’s quite a bit further down the “oil peak” road than most people think we’ll hit anytime ‘soon’. In other words, it will take such huge increases in the cost of gasoline to get suburbanites to stop driving to work alone that it’s not even a factor for the foreseeable future. Even then, one would assume that rather than abandoning their stake in the ‘burbs, some large percentage of suburban drivers would just get more fuel-efficient cars. At $6.00/gallon, driving a Toyota Prius, Joe Suburban’s daily commute cost drops back to 2.48 without tolls and 3.98 with. Oops.

See my previous article on my ‘week without a car’ — even for me, who is the only guy at my 60-person office who could possibly take the bus to work without transfers, it’s not cost-and-time-effective to use transit until gasoline is really REALLY expensive. It costs me about 30 extra minutes per day and saves me pocket change.

When does transit make sense? When the time penalty is minimal and/or the cost savings are comparatively large. Two obvious (much shorter) use-cases:

  1. If I worked downtown, I could take the #5 bus straight there at a time penalty of perhaps 5 minutes. This time penalty is so small as to be not worth counting, and I could actually get rid of a car, thus moving us into the realm of the traditional commute calculators – a huge economic win for the transit alternative. Unfortunately, the current economic regime penalizes businesses who locate downtown rather than in the ‘burbs (far higher property taxes) even though they generate far less demand on city services.
  2. Lucy Leander works at the University of Texas and has to pay roughly $5/day for parking. She lives close to a park-and-ride where she can pick up a good express bus to work which isn’t much slower than her car would be. Here’s her comparison. Even at $2/gallon, she saves $7.36 a day (without getting rid of a car) and only spends a few more minutes. Note that having to pay for parking makes this comparison far more favorable for transit.

So my lesson is: Major employers should be downtown (where transit can serve them), and parking shouldn’t be free. Until either one of these is fixed, however, you’re going to get nowhere with me by claiming that a plan is economically viable (or not) based on gasoline prices.

Unfortunately, current conventional wisdom is still that spreading jobs through the suburbs reduces average driving (absolutely false). The facts have an anti-suburban bias, I guess.

Game On

Today’s Statesman (registration required) contains the first non-gushing comment about Capital Metro’s plan to screw the center city in favor of Cedar Park and Round Rock (who don’t even pay Capital Metro taxes) in order to curry favor with Mike Krusee.

But the agency will have to win over some lukewarm Austinites.
“I absolutely reject it on its own merits because of the benefits for people who don’t pay and the lack of benefits for people who do pay, said Mike Dahmus, a member of the Urban Transportation Commission, an advisory board for the Austin City Council.

He said the plan would shortchange the large number of city residents who provide the agency’s tax base in order to serve residents of the suburbs. Plus, he added, “the commuter rail doesn’t go anywhere near the University of Texas or the densest urban core.”

The bulk of Capital Metro’s budget comes from a 1-cent sales tax levied in Austin and a few surrounding communities that are part of the agency’s service area.

News 8, on the other hand, interviewed current bus passengers. Even Capital Metro isn’t quite stupid enough now to think that the opinions of current bus users should shape a rapid transit line, although they’re still attacking the issue from the angle of cost, which is not a winner with rail or bus.

Today during lunch, I hope to get the first fact page up (this one about the proposed rapid bus line). This will be an uphill struggle at best.

Cost of driving to driver

So in Tuesday’s Cap Metro briefing, one of the points I made is that an attempt to encourage people to use transit based on cost savings is doomed to failure, because the bus really isn’t any cheaper than the car for most people. Assumption here is that you won’t be able to completely get rid of a car, i.e., you ride the bus 4 days a week, or even 5, but can’t reduce your family’s number of cars.

The two downtown lawyers looked at me as if I was crazy. Well, I’m used to it.

Here’s the problem: Most of the people who pay a lot of money to park work downtown. Almost none of the new buildings there are underserved with parking, though; so the average cost per employee to park is dropping, even in the one place in town where it isn’t free. Free is a good assumption to work on (I suspect that most employees in those new buildings are getting free parking from their employers).

Then, we hit the “well, the IRS claims 27.5 cents per mile”, or whatever they’re saying now.

Yes, the IRS does in fact allow you to deduct business-related driving at that level in most cases. A big chunk of that is not gas, or tires, or maintenance – it’s depreciation, which makes sense for a business (which usually must depreciate assets like that as a matter of accounting principle).

But I went over this with my bicycle cost comparator. The fact is that unless you can get rid of a car completely, this depreciation number is not applicable to using your car for personal use (and yes, commuting to work is personal use).
I have never gotten one more dollar for a car on a trade-in for having disproportionately low mileage. Anectodal evidence exists of a few people who got an extra hundred bucks or two on a ten-year-old car for low mileage, but even that figure is trivial compared to how much of the original value of the car depreciated as a function of time, not mileage.

So, if you’re talking about taking the bus to work even every day but you live in the suburbs, you ain’t getting rid of that car, and thus, you ain’t saving 27.5 cents per mile. Gas and tires are about all the consumables you can treat as a mile-based expense; most maintenance is necessary every N months even if you drive the car a tenth as much as the typical user. Insurance is not mile-based (even though there were a flurry of press-releases about it supposedly being offered in Texas, it hasn’t materialized). Neither is registration.

So, a comparison for me:

I drive my wife’s old Honda Civic to work (when I drive). I take my bike on the other days, using the express bus for a boost in the morning. Let’s suppose I took that bus both ways.
From my calculator on my trip:

  • Car cost: $1.20, of which $1.10 is gas.
  • Bus cost: $2.00 ($1.00 each way).
  • Note that the following bus savings can be used:
    1. You can buy pre-paid tickets at half price, thus bringing the bus cost down to $1.00.
    2. You can buy a monthly express bus pass for $17 ($0.84 per day if you used it 25 days a month).

    Even in the most optimistic scenario, I’d only save $0.16 per day by taking the bus. That’s never going to be compelling enough to get me to vote for any transit proposal whatsoever, which was the point to begin with.

    For comparison, Cap Metro’s calculator says it costs me $184 a week if I drive all 5 days.

    Cap Metro doesn’t understand “choice commuters”. The things that could get them to vote for more money for transit are:

    1. Reliability – my trip down Mopac takes 20 minutes to 1 hour depending on traffic. A guaranteed trip time of 45 minutes on which I could read would be worth something.
    2. Performance – 45 minutes, OK. 1 hour, no way.

    Unfortunately, their rapid bus proposal does next to nothing on either metric above.