Why can’t we internalize the cost of commuting?

Yesterday I was biking from Saint Paul to Minneapolis at rush hour. In to a stiff headwind, which shaved four or five miles per hour off my speed—more on downhills. However, when I crossed over Interstate 94 (*) just east of the Mississippi, I smiled. Both sides of the highway were parking lots. I probably wouldn’t have gone any faster in a car.

I hate traffic. Actually, I should rephrase that. I hate being stuck in traffic. I love the concept of traffic, as long as it is not “solved” by building more roads. Livable cities have traffic. (They have transit as well. The ones with traffic and no transit, well …) I continued on a few blocks to my destination and, while I was not particularly happy fighting against the gale, I was glad I wasn’t giving myself ulcers in a traffic jam.

There have been several recent articles about the fact that people are unable to properly calculate the cost—both economically, the time cost and the emotional distress—of a long commute. It’s been called the commuter’s paradox. The long and short of it comes down to the fact that people, when making an important decision (where to live) will worry about rare, very inconvenient occasions more than frequent and not-quite-so-inconvenient-but-still-bothersome occasions. In other words (mostly those of Ap Dijksterhuis of Radboud University in the Netherlands):

Consider two housing options: a three bedroom apartment that is located in the middle of a city, with a ten minute commute time, or a five bedroom McMansion on the urban outskirts, with a forty-five minute commute. “People will think about this trade-off for a long time,” Dijksterhuis says. “And most them will eventually choose the large house. After all, a third bathroom or extra bedroom is very important for when grandma and grandpa come over for Christmas, whereas driving two hours each day is really not that bad.” What’s interesting, Dijksterhuis says, is that the more time people spend deliberating, the more important that extra space becomes. They’ll imagine all sorts of scenarios (a big birthday party, Thanksgiving dinner, another child) that will turn the suburban house into an absolute necessity. The pain of a lengthy commute, meanwhile, will seem less and less significant, at least when compared to the allure of an extra bathroom. But, as Dijksterhuis points out, that reasoning process is exactly backwards: “The additional bathroom is a completely superfluous asset for at least 362 or 363 days each year, whereas a long commute does become a burden after a while.”

In addition to the environmental catastrophe of building and heating (or air conditioning) superfluous, unused rooms, people have convinced themselves that they need this extra space, and pay dearly for it: first when they buy something larger than necessary, and then when they spend hours a day in the car because they can’t get anywhere without turning the ignition key. (Has no one heard of a fold-out bed or a hotel room? It’s a lot cheaper than a bigger house.)

The article also goes in to the issue that the worst thing about traffic isn’t that it’s bad, but that it’s unpredictable. With all sorts of technological advancements, the best we can do now are sporadic signs above the Interstate telling us how long we have left in this particular hell. With a little money, we could make the trains and buses run on time (the worst thing about a poorly functioning transit system is, generally, its unpredictability). Making traffic predictable, on the other hand, is all but impossible.

There’s another piece up recently regarding a book about car dependence. It goes in to more of the economics, that we don’t internalize the costs of driving because they are so ingrained in the American psyche. It’s really a problem, and one that will take decades to fix. Whether the current economic status (and, yes, economics, not environmentalism, will be the driver of less automobile use) or higher gas prices will make a change is yet to be seen. However, it may be a generational change, and amongst a generation of always-plugged-in folks who see time in a car for what it is—almost completely wasted—we may see more people who are less interested in driving the latest, greatest shiny new automobile.

[ * Why is I-94 bad going west? Because in the course of about three miles there is a lane drop on the right (Riverside), a lane drop on the left (35W), two horrible merges on the right (from 35W and 11th), and then a sharp curve in to the Lowry Tunnel. There are no lanes which don’t disappear or have just brutal merges. You can add all the suburban lane miles you want, but it won’t address these bottlenecks.]

Cash for clunkers: proof that a gas tax would work?

There has been a lot of debate as to the overall efficacy of the Car Allowance Rebate System, (legislators love acronyms) colloquially known as “Cash for Clunkers.” On a few subjects there isn’t much contention: it has been “successful” in getting people to buy new, and generally more efficient, cars. In other words, if people have a financial incentive to trade up to a more efficient car, they will do so. Especially if the incentive is (probably) set too high.

So, I’m not down on Cash for Clunkers. First of all, it’s proof that a government program can work. It was quick and effective and probably stimulative (more so than environmental)–most of the cars in the program were made in the United States. That’s good in that it may help convince some anti-government types that government is not always the problem. Second, it is not increasing the number of cars on the road. While it is certainly not perfect, a far more worrisome development would have been a program that mailed out checks to people to buy new cars; a program which I could see government embracing. Third, it can’t be debated that the new cars on the road are, in fact, less polluting than the current ones. While not everyone went out and bought the newest Prius (although many are), a 60% gain in efficiency is nothing to scoff at. Even if these cars may be driven more than their predecessors (since they’ll be new and reliable and, well, not clunkers) there will likely be an overall decrease in emissions.

On the other hand, the program could have, obviously, been better administered. First of all, $3500 to $4500 is a lot of money. I thought about buying a clunker, trading it in, buying a new car and turning around and selling that–even with the title transfers, time involved and money lost to depreciation, I’d probably clear a couple grand. (I’m not sure, however, if I could have qualified with a new-to-me clunker.) In any case, smaller sums–$1000 to $2000–would have likely resulted in many sales but not the veritable run on the bank that car dealers have recently seen. In addition, there was no provision for people with clunkers who wanted to get out of car ownership completely. The only way they could do so would be to trade in the clunker, buy a new car, and turn around and sell it. Maybe the next program should be that if you bring in an old car, the government will give you a year-long transit pass for the agency of your choice and a $1000 credit for your local car sharing agency. This, too, would cost less than $3500, and dramatically reduce emissions and the number of cars in the road. (Yes, I have a bit of a vested interest in the second half of this proposal.)

While the transit-car sharing idea is a bit of a pipe dream, politically, one which is less of one would be a better-graduated system. The CARS program had hard cutoffs. If you car gets tenth of a mile per gallon over the limit, you get nothing. A tenth less and nearly $5000 can be in your pocket. Furthermore, you get this money whether you upgrade to a still-overpowered sedan or SUV getting in the low 20s or a Prius (or similar) getting twice that. So what would make more sense would be a graduated system. Trade in an 18 mpg car and go to a 22 mpg and we’ll give you a few hundred dollars for your trouble. Go from a 14 mpg SUV to a Prius (or a similarly “clean” car), and you can cash in on the full $4500. Or more.

That’s all well and good and probably won’t happen. Nor will credits for transit commuters, cyclists and others who choose not to drive. It costs too much money and isn’t terribly stimulative and probably doesn’t have the votes. Furthermore, the CARS program was very simple. Your vehicle either does or does not qualify, and you can get either $3500 or $4500. For these others, we’d need charts. And if you put mathematics in between an American consumer and a deal, they’re far less likely to do it. In other words, if you make it as confusing as doing your taxes, people are going to like it about as much.

There is a relatively simple way to achieve nearly all of these objectives. It would require little administration, since the methods of collection and distribution are already in place (and have been for years, and work fine). Yet, for a variety of reasons, it is a political third rail. It is, of course, the gas tax.

The federal gas tax is 18.4 cents per gallon. That’s right. 18.4 cents. Most states have their own taxes on top (Alaska is the only holdout) raising the total tax as high as 60¢, in New York State. The federal portion was last raised in 1991. Yup, 18 years ago. Since then, prices have increased 58%. Had the gas tax kept up, it would be 29¢ today. The gas tax in 1991, however, accounted for about 17% of the cost of a gallon of gas (at that time, gas, with the tax, cost about $1.20). If gas taxes were based on percentages, they would be about 43¢, and last summer would have crept to nearly 70¢.

So, it’s obvious that gas taxes are low. And it’s also pretty obvious that there is some climate stuff going on, and that having people use less gas would be beneficial. In addition, using less gas would keep prices lower and supplies more stable, as well as encouraging energy independence. These are all good externalities, but, perhaps most importantly, the gas tax, if it is adjusted for some rural populations and low income communities, is a very efficient way to raise tax revenues.

Mention raising the gas tax and you’ll hear two responses. One is “it’s not politically possible.” The other is “it’s regressive.” The first is, sadly, perhaps true. The second is not, and, particularly when it is offset with some sort of tax credit, potentially a straw man. When the tax was last raised, 18 years ago, this was debunked. In several manners, it has to do with how you look at gasoline: whether it is a necessity or a luxury. If it is a necessity, then, yes, the tax is likely somewhat regressive. This is the reason we don’t place punitive taxes on clothing and food: you need both to survive. Gasoline, however, is a different story. In New York City, 55% of the residents do without a car. Yes, it’s a special case. But is there anywhere where more than half the residents do without food or clothing? In several other major cities, more than a quarter of households don’t have cars. For some it is an economic decision. For others, it is about lifestyle. But it is rather obvious that, especially in areas with decent public transport, owning a car is not a necessity.

And for these people, which number in the millions, a gas tax is not regressive at all. Many of them are the same people who the highway lobby defends; the people for whom a gas tax will be painfully regressive. However, as long as they aren’t driving a gas tax will have no effect, although it might drive more people towards transit use and increase service levels.

The other worrisome issue are those people who live in rural areas. For them, higher gas taxes will result in higher costs, because living at a low density tends to require a lot of driving. And for farmers, a rise in gas prices will create a rise in production costs, for both mechanized agriculture and transportation. There are two ways of dealing with this issue. One is direct subsidies to growers to buy cheaper fuel, although such a system would be fraught with fraud and inefficiencies. (If we’ll sell you 10 gallons of cheap gas, is there much of an incentive to economize and only use nine?) A simpler way, of course, is to pass the costs along: food prices might rise a bit, but everyone would have increased costs, and everyone would pay. In addition, there would be a fine incentive to save fuel, which would both reduce costs and be more environmental. For those who live rurally for the lifestyle, they’ve made a choice to live a car-dependent (and fuel-dependent) lifestyle. It’s only fair that they pay more.

Finally, there is a way to make sure that a gas tax would both not hit the poor especially hard and be stimulative as well: return the extra money spent on gas, in advance, as a tax credit. Estimate the amount of gas used per year (recently about 140 billion gallons) and the amount of money that, say, a $1 gas tax increase would raise (with less use, about $120 billion). Knowing that that revenue increase was in store, the government could turn around and write a $500 check to every tax payer in the country at the beginning of the year. A nice letter could be enclosed:

We know that we’re increasing your gas tax. Here’s $500. If you need it for gas, use it for gas. If you want to buy a more efficient car, here’s some help to buy a new car. If you are interested in local transit service, here’s a website where you can find out more. Here’s information about car sharing, car pooling and other fuel saving techniques, too.

Oh, and enjoy the $500!

People worried about fuel costs could save the money for the year. Many others would spend the money in ways that would stimulate the economy. Others would, in the face of higher gas prices, use it for transit passes. And it would be a very progressive tax rebate: it would benefit those at lower income levels far more than those at the top.

In the long run we might, as a society, want to use this money to fund more effective transportation policies. Maybe the amount would decrease by $50 a year as people got more used to higher taxes, by driving more efficient vehicles or driving less. Any extra money could be put towards funding expansion and operation of transit agencies, and building new energy and transportation networks (as the current gas tax is earmarked for transportation). In the short run, as has been discussed in several places (including liberal blogs), consensus is that we can’t get everyone out of their cars tomorrow. But instead of expanding the Cash for Clunkers program, and making it more top-heavy and unwieldy, a gas tax would likely give us better results with easier implementation (since it’s already implemented).

And if everyone were promised a $500 check from the government, it just might be possible.

Peak Car? Where next?

There was some mention on the blogs this weekend about the United States having reached “peak car.” It is a similar idea to peak oil except that instead of being supply driven (the idea of peak oil is that the available and economically accessible supply will begin to drop), “peak car” is driven by demand.

One of the interesting things about the last year is the extreme drop in car sales. For a while, car sales had been humming along, at between 15m and 20m per year. Before the current recession got in to full swing last September, gas prices hit car sales. All of the sudden, the cars which were readily available—generally gas-thirsty, larger vehicles—were out of vogue, or at least unaffordable. And while everyone wanted a Prius, supply was so short that wait lists grew to several months long and used Priuses actually appreciated—which is almost unprecedented.

That followed through the summer of 2008—a.k.a. “The Summer of Four Dollar Gas.” People drove less, rode transit more and generally showed that high enough gas prices would begin to change behavior (although since demand for gas is so inelastic, even a doubling of its price only diminished demand by a few percent). And then the economy crashed in September, and credit markets tanked. Foreclosures skyrocketed, and many more people made due with older cars. Thus, car sales, which had not fallen below a seasonally-adjusted annualized rate (SAAR) of 16m for more than a month for nearly a decade (1999 through 2007) crashed. By January, the SAAR was under 10m, a rate last seen in 1981. Chrysler and GM have been forced in to bankruptcy. (It is easy to sarcastically remark that it was because they made cars that no one wanted, and surely that is part of the problem. But for ten years they were also feeding a buying frenzy which created more cars than necessary, so that when the bottom fell out, they were not in a position to scale back.) Ford is cutting back drastically, and foreign automakers have seen sales plummet as well.

In 1978, car sales briefly flirted with the 16m mark, but then fell back towards 8m by 1982, sending Chrysler to the brink. It’s an interesting parallel—the peak and nadir numbers are similar—but likely not apt for two broad reasons. One, the 1976 peak at 16m was a singular peak—sales had hovered around 12m for a few years (the Times’ data only goes back to 1976, but we can assume that sales were slow in 1974 and 1975 during the gas shortages), jumped to 16m, and then fell back. But for nine years, from 1998 to 2007, through thick and thin, vehicle sales plateaued at 16m per year. Starting in 2005, the plateau began to slip, until it decreased, parabolically, in late 2008 and early 2009. Car sales, which had been remarkably non-volatile for a long period, fell at a rate never seen before.

The other reason this historical comparison is sketchy is that the early 1980s and the late 2000s are very different times. In the late 1970s, car ownership was still the wave of the future. The Interstate Highway System had been largely completed in the couple years before, and suburbanization, which had drawn out millions from non-car-dependent neighborhoods, seemed to be accelerating. 1980 was, for several major American cities, the low point of their population. Racial strife was simmering down, but still quite recent, and “white flight” was prevalent. By 2000, however, many cities had gained population off their 1980 lows, and the 2008-9 recession may have put the stake in the heart of the ever-expanding suburbs and their three-car garages, something the stagflation days of the late ’70s did not accomplish. Additionally, sentiments towards automobiles seems somewhat different. While I have no anecdotal evidence from the early 1980s (uh, I wasn’t born yet), I can only assume that people rarely said things like “when my car dies, I’m not buying a new one.” I’ve heard this a lot recently.

Finally, wide-scale demographics are different now than they were in the early ’80s. The large generation in 1980 were baby boomers. This generation had grown up largely in the idyllic suburbs of the 1950s and 1960s, or at least aspiring to (or moving to) them. Roads were new and wide, traffic was minimal, and the new American ideal was seen to be two cars in every garage. The large generation today, the millenials (generally the offspring of the boomers), grew up in the same communities, but times had changed. The suburbs were older, sterile and boring. The city was no longer seen as anathema to a healthy lifestyle. More younger folks are moving to cities, where they are less likely to need to own one (or more) cars.

So, given this data and these very anecdotal trends, let’s hypothesize that we have reached “peak car.” Let’s assume that, while auto sales may recover to 10m or even 12m, the number of cars per person in this country will decrease, and the number of cars overall may even do so as well. The country has been adding cars at a rate of about four million per year (almost exclusively the growth was “light trucks”) for quite some time—if sales all-of-the-sudden drop by half, the number of cars may begin to plateau as cars, inevitably, die. What happens if we have, indeed, reached peak car?

Well, first of all, what will the car production capacity be? With recent cuts and both Chrysler and GM going through bankruptcy, it is quite possible that auto production will fall towards current demand. 12m cars per year is, very roughly replacement rate. (A long article in New York Magazine—this is the second page—quotes it as 15m but, as car sales drop and people keep cars longer, I’d expect the replacement rate to drop as well.) And with the return to savings and changes in both demographics and spending patterns, this might be the rate in coming years. If nothing else, if less than 10m cars are sold this year, it is quite likely that, having sold six or seven million fewer cars than normal, the number of cars on the road could, for a year, anyway, do something quite unusual: drop.

But—and this is a pretty big but—this does not take in to account peak oil, or at least oil prices skyrocketing. This could put the damper on car sales, and push them below replacement rate. If gas were to go to $6 or $8 per gallon, vehicle miles traveled would decrease (as we have seen), land use would probably change, and demand for alternate-fuel cars would go through the roof. This would create two problems. First, the current availability and charging capacity for these cars is in its infancy, and will likely take a decade or more to develop. Second, while battery technology is advancing, ramping up to the scale of millions of car-sized batteries would take time, energy, and cost.

For many years, the American car companies built at capacity, and the American consumer bought. When there were issues with American cars (quality, fuel consumption), foreign automakers moved in. The problem is that the auto industry is built on having a very non-volatile demand, as its supply is very inelastic—at least if demand outstrips supply. A normal car takes several years to design, even when it is a derivative of something on the road (a SUV chassis on a car body with an overpower engine, for example.) For significantly different cars—the Chevy Volt or Toyota Prius, the lead time is much, much longer—the Prius was in development for seven years before sales in the US started, the Volt has been rumored for years and we’ve seen only a concept car.

Thus, if the total capacity of the system is decreased and, for whatever reason, the market dictates that consumers want a type of car not being made, the demand for cars will outstrip supply. This will leave consumers in a situation where they can buy a cheap, fuel-inefficient vehicle for cheap (as we saw last summer, when big SUVs had their prices slashed by half) or wait for months, if not years, for a more fuel efficient model—or pay premiums of thousands of dollars for one. If the number of cars per person begins to stabilize or fall—which is unprecedented in the last century of American history—it’s possible that it could create a paradigm shift for the American car industry, and American development in general. When gas was near $4, as it was last summer, we saw unprecedented rises in transit and bicycle use. While the housing market’s completion of its crash has dictated development since then (or lack thereof), a prolonged period of high gas prices without efficient or alternative automobiles could drastically effect patterns of settlement, a recentralization of jobs (since employers would be incentivized to have jobs in places accessible by non-car transportation—especially if a stronger economy created more supply for jobs, as right now most job-seekers will take whatever they can get), and densification of areas served well by transit which, in may cities, consist of surface parking and recent single-story development. And this, of course, is a good thing.