« Dead End For Viaduct? | Main | The Scrap Heap Is History? »

September 19, 2005

Sic Transit

As big-time blogger Duncan Black noted over the weekend, high gasoline prices seem to have boosted ridership on some of the the nation's transit systems -- which led big-time blogger Matthew Yglesias to speculate that gas consumption may be more sensitive to price than economists have predicted.

Yglesias' take seems mistaken to me.  Nationwide, less than 5 percent of all commuting trips are taken on transit; and commutes represent a minority of all trips that people make, but a fairly large share of all transit trips taken.  So even if transit ridership were boosted by, say, 20% -- which is a huge spike indeed -- that might represent a decrease in vehicle trips of, oh, a half a percent or so at most.

In fact, it seems to me that any recent increases in transit ridership are pretty much in line with what economists would predict from recent gas price increases.  (See here, especially table 8, for a summary of economic predictions for the relationship between fuel prices and demand.) Of course, that doesn't necessarily undermine Yglesias' main point, which is that higher gas taxes would decrease fuel consumption. 

I've also got to second this part of Duncan Black's response to Yglesias:

[L]and use patterns in much of the country have made it quite difficult for any proposed transit systems to really improve the lot of most existing homeowners/commuters. Expanding rail systems into existing suburban areas really only makes...sense if its accompanied by some land use changes in those areas (there at least needs to be higher density development around the stations themselves).

As far as I can tell, this is just about right: transit is rarely cost-competitive in low-density, sprawling neighborhoods, since both riders and destinations are simply too spread out.  (Color me extremely skeptical about about supply-side theories of transit ridership.)  Which means that, since the shape of our cities changes much slower than gas prices, a lot of us may be stuck in our cars no matter what gas prices do in the short term.

Posted by ClarkWD | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Sic Transit:


I'm not sure what the point is that you're trying to make here. The boost in ridership is pretty big news, I think, because everything up to now has suggested a lot of elasticity in demand for oil and this is showing that there might be a limit to that starting at $3-$4 per gallon. The fact that we have hardly any mass transit options now only means we aren't planning ahead, not that what's coming doesn't involve a lot of mass transit.

Keep in mind that our entire expansionist economy is built on the premise of cheap oil. Post-war white flight occurred not just because of racism but because of subsidized mortgages and highway construction paired with cheap oil. Take away the cheap oil and the appeal of suburbs is diminished significantly and exurbs may cease to exist.

But since so many people are deeply invested in the status quo, you can't expect ridership to double overnight. The fact that it did in MN should be significant if you're paying attention. Think it's high at $3 per gallon? When people realize that's a bargain, things will change fast and it could be wrenching for more than a few. Mass transit will popular out of necessity, not do-gooderism.

Posted by: Jason | Sep 19, 2005 7:36:04 PM

Jason -
I largely agree with what you're saying. My point, though, is that *even if* the recent spike in transit ridership is real, it doesn't reduce total gas consumption by all that much. The numbers just don't add up -- sadly, in terms of oil saved, the transit spike just a drop in the bucket, or...er...barrel.

I think that there *are* ample signs that higher gas prices are beginning to have the expected effect: when gas prices go from $2 to $3, within a year total gas consumption goes down about, oh, 5 percent. Maybe a bit more, maybe a bit less. And it looks like we're starting to see signs of that decline in higher transit ridership, lower gas station sales, etc.

Now, if price hikes are sustained for a while, then consumption goes down even more: people can switch cars, change jobs, move to shorten commutes, etc. But with US (and Canadian) cities as dispersed as they are, you just can't expect that to happen overnight.

Don't get me wrong--a jump in transit ridership is encouraging news. But it's worth thinking about any gas conservation gains as a multi-year process, not something that will kick in over the course of a few weeks or months. And it's probably not the case that the transit jump is indicative of higher-than-expected gas price elasticity, as Matt Y. mentioned.

Posted by: Clark Williams-Derry | Sep 19, 2005 10:41:05 PM

OK, I guess my meta-point, the one I forgot to make, was that you shouldn't assume the demand curve is linear, or simple, or even the same curve at different points in time. There may well be an inflection point (somewhere in time and price scale) where people really decide their current behavior will burn them in the future and permanently change it.

Fear can do weird things to people, like make them completely contradict normal behavior. I think the spike from last week was just that, but it could indicate the presence of a point where traditional assumptions about price elacticity start becoming outdated.

I'm not saying we're there-- I'm just saying it's not valid to conclude just because the last doubling of the price did little to stem demand that the next doubling would not affect the status quo.

Keep in mind that we are only now approaching previous record oil prices in real dollars because inflation has been so low for so long. It's when we hit records in real dollars that we are in uncharted territory where anything can happen.

Posted by: Jason | Sep 20, 2005 11:43:35 AM

That may be so. It might be interesting to see if gas price elasticities are any different here than in Europe or Japan, where gas is much more expensive. That is, if you see different estimates of what a 10% rise in prices will do in Europe (where gas is $5, and 10% = 50 cents) vs. in the US (where much of the research was done with gas at ~= $1.20 or so, and 10% ~= 12 cents). You might see very different demand responses, or very similar ones -- it's hard to know in the abstract.

If you see anything along these lines, please, by all meanse, send it along. And thanks for your interest!

Posted by: Clark Williams-Derry | Sep 20, 2005 12:39:58 PM

Transit seems to be designed as poorly as it possibly can be - frequent stops, walk-up to seating, driver acting as fare taker, lack of priority lanes and signaling, transfer mismatching, and especially charging extra during peak travel times (thereby discouraging using transit when it's most useful with respect to congestion and air quality).

Few people realize how little fares contribute to the operating budget of a transit system. In the Twin Cities, Metro is estimating expenditures of $307 million (which is probably low, since the estimate was made before the huge increases in fuel costs) with fares and similar revenue of $75 million. So, really only 24% of expenses are covered by fares, at least with our system.

With a metro area population of 3 million, fares come out to about $25/person.

By contrast, the new football stadium being proposed is looking for roughly $500 million in public subsidies, or about the same as 7 years of fare revenue for the transit system.

An intelligent approach would be to have free transit service (like Whidbey Island Transit does). That would speed up service in terms of entry speed, as well as substantially increase ridership. More ridership and more routes/more frequency could be added.

Seems like $25 is a small price to pay for something that would do as much good for congestion, safety, ease, and air quality.

The only real drawback would be increased costs that would go with improved service, though over time, once a transition to a more balanced transportation mix was achieved, some alternative funding approaches could be tried (eg, employer-based taxes).

Posted by: Joseph Willemssen | Sep 22, 2005 12:27:20 AM