Tax gas more when it’s low. Less when it’s high.

Gas prices are low.

Really, really low.

But gas prices are also volatile. They can swing—and have swung—by 200% in a few months. What other such omnipresent commodity has such wild price swings? Eggs? Maybe. Gas prices were low in the early 2000s, rose mid-decade and then spiked after Katrina, rose further in 2008 to more than $4 per gallon, plummeted with the economy and bottomed out below $2 in 2009, rose back to $4 by 2011 and have since dropped back under $2.

This is all related to the base cost of oil, variable global demand and long lead times for production. But one thing that doesn’t change with gas prices are gas taxes. If gas is cheap, the tax is the same per gallon. Expensive? Same tax. There’s a pretty good reason that gas is not taxed as a percentage: it would be punitive when prices were high (which can be harmful to the economy) and wouldn’t raise enough money when prices are low. So we have a set tax per gallon.

But it’s not enough. In Massachusetts, the economy is doing well, but is held back by pesky infrastructure. We can’t rest on our laurels, but we seem unwilling to pay for what we need.

The problem is that—especially given the recent volatility in prices—raising the gas tax makes people worry, somewhat rightfully, that the tax will cause high gas prices to go even higher. Sure, gas might only be $1.79 today, but if there were another oil crisis and gas spiked to $4.25, would we really want to tack on another ten or fifteen cents? There’s a very cogent response that, no, we would not: gasoline usage is very inelastic as most travel is non-discretionary. Sure, some people will switch to other modes, and some won’t make trips, but higher gas prices mean that middle-class consumers have less money to spend on other goods. Raising the gas tax when the price is high is akin to hitting consumers while they’re down.

But low prices are just as perverse, for the external effects of gasoline usage. Low gas prices provide little incentive to purchase fuel efficient vehicles, so more consumers buy gas guzzling vehicles, which lead to more pollution, climate change issues, and boneheaded government programs when gas prices rise and everyone cries poor. Gas price volatility is a problem for pretty much everyone.

The gas tax is really different from every other tax. Few other taxes are on a good with a price as volatile as gasoline. Other taxes which we use, in theory, to reduce the consumption of a good (“sin taxes” on cigarettes, for example) can stay high even if prices rise, since there is no economic benefit to easy access to them. Many other taxes are collected as a percentage, but that is anathema for gas taxes as it would only serve to exaggerate price changes. What we really need is a tax that is high when gas prices are low, and low when gas prices are high.

Paul Krugman has cogently argued for a price floor for gas, but this has a couple of issues. It completely disrupts the supply and demand curves with a flat floor, so there is no natural variability in the gasoline markets. Second, if the price of gas rises above the “floor”, gas tax revenues cease; it it rises high enough, a major funding source for transportation dries up. (But, I must point out, one that does not come close to covering the full cost of our roadways; in fact, roadways in most states, including Massachusetts, have a “gas tax recovery” ratio of about 40%, similar to the MBTA’s farebox recovery, so subsidies are about the same.)

Another issue with a floor is that with volatile gas prices, it would mean that the floor price and the actual price would often be quite disparate, and it would have to be implemented on a national scale to keep motorists from having a high incentive to cross state lines. There’s no incentive built in, so it’s harder to sell as having any benefit when prices are high. It also requires all gas stations to charge the same price even though non-gas expenses—real estate, operations—often result in prices which differ by area and business. It’s a blunt instrument.

But what if you split the difference and had a gas tax which rose and fell inversely to the price of gas. When gas is cheap, the tax would be high, providing a sort of moving floor. When gas prices spike, some of the tax revenues could be saved and actually help to mute some of the spike with a rebate: so if the base price of a gallon of gas climbed to $4.50, consumers might only pay $4. I would contend that there is not much difference between gas prices of $1.75 and $2.50—both are considered cheap. Cheap gas is bad, and expensive gas is bad, and volatility is worse. With this sort of plan, when gas prices are low—which often correlate to a more sluggish economy—you get extra money for infrastructure improvements which provide jobs and economic stimulus. When they are high, the effect of the high price is muted somewhat, dampening the effect on the economy of high prices.

Let me put it another way. In the 2015, gas prices ranged from $1.97 to $2.71. Was gas expensive in the last year? No, it was cheap. What if there had been a tax which meant that prices ranged from $2.66 to $2.98? Would gas have been expensive? No, because under-$3 gas is still cheaper than any time since 2011. $3 gas doesn’t kill the economy. $5 might.

Looking back at the last twelve years of gas price data, I’ve come up with an idea which would raise the average gas tax—thus raising revenues for road and transit projects—but would also reduce the volatility of gas prices and, when prices are very high, actually reduce prices at the pump. Note that this is in relation to the Massachusetts state tax of 24¢, but would be just as applicable to the federal 18.4¢ per gallon tax. If the price of gas per gallon is G, then the tax would be:


Note that these numbers can be moved around to increase or decrease the taxes paid, or to change the variability of the tax. For instance, $3/G-$0.70 would yield a tax with more variability; a lower floor but fewer savings when prices are high. But this seems to be a happy medium which matches the experience of the past 12 years well.

As for the costs—going off a baseline of the before-state tax price of gas (take the current price at the pump and subtract 24¢ in Massachusetts):

At $2, this would yield a tax of $1 per gallon and a price of $3
At $3, the tax would be 33¢ per gallon, for a price of $3.33
At $4, the tax would be 0¢ per gallon—a savings of 24¢ per gallon compared to the current tax rate—and the price would be just $4.00
At $5, the tax would be -20¢ per gallon, meaning that the price would be a quarter lower than it otherwise would have been; the price would be $4.80

If gas prices were to drop to $1—and this is unlikely, as it would equate to basically free oil—the system would break down a bit, as the tax would be $3.00 and the price $4.00, so some baseline could be built in (maybe a maximum tax of $1.25). If gas were to spike to $6, the tax would fall to -33¢, so the rainy day fund would be able to pay in to that for a while before becoming exhausted (and the tax could have a provision to readjust when a certain portion of the rainy day fund was paid out). 

Remember when we all had graphing calculators
in high school? Now, theres an app for that.
This graph shows the 4/G-1 (in red) and the 3/G-0.7
(in blue). Note that below $1.50, both increase
exponentially, so a cap would be needed.

Had this sort of tax been in place since 2004, prices would have ranged from $2.76 to $4.01 over that time, rather than the actual range from $1.82 to $4.38. (The 3G-$0.70 would yield a slightly larger range of $2.55 to $4.15.) It would mean a goodbye to very, very cheap gas. But we seem to have survived somewhat higher prices. We even celebrated when gas prices dropped below $3 for “the first time in, like, forever” back in the heady days of—let me check the dateline—October 2014. Having gas prices stay in the $2.50 range, even when the price of a barrel of oil plummets, is not necessarily a bad thing.

There would be plenty of benefits. One would be predictability for auto purchasers. Cars are durable goods, and a single vehicle can last through several oil price cycles. Consumers buying less-efficient cars would have a better idea of what gas prices would be, and they would no longer drop so low that inefficient cars would become (relatively) cheap to operate. And when prices spiked, it would be less of an impact for these consumers. High efficiency buyers would be incentivized to buy cars without the worry that recouping the extra cost for a hybrid or alternate fuel vehicle would be negated by a drop in prices. It would even benefit automakers, who would be able to better plan vehicle models and production schedules knowing that the volatility of gas prices would have less of an effect on consumer choice.

While this tax could be implemented as a revenue-neutral scheme, it would also make a revenue-positive gas tax more palatable. For instance, the tax described above would have averaged 37¢ from 2004 to 2015, during which time the gas tax in the Commonwealth was 21.5¢ (21¢ for 10 years, 22¢ for two). The gas tax in Massachusetts raises about 38 million dollars for every penny it goes up (and an average increase of 10¢—just 5% of the volatility in prices in the past decade—would be very unlikely to bend the demand curve significantly) so raising the tax from 21.5¢ to 37¢ would have raised a net of $7 billion dollars. That includes money for the rebate drivers would have received in 2008, when the tax would have been -9¢. That certainly would have been beneficial to the economy, which was already teetering on the brink for a variety of reasons, and certainly wasn’t helped by $4.50 gas.

There are some issues to this scheme, but they would be surmountable. First is that if the price of gas was consistently high, the tax—and the necessary revenue—would disappear. So any such tax would need a set amount that the rainy day fund would pay out and a specific criteria to recalibrate the tax towards a new baseline. If there were a few years of cheap gas, a months-long spike wouldn’t be an issue. If gas prices spiked for a long time, the tax would have to rise, but could keep the same adjustability provisions. (Why not set it to inflation? There are two issues. The first is that gas and inflation correlate very little. The second is that gas prices are part of the inflation, so if gas prices go up, inflation goes up, and this, if anything, causes a feedback loop.)

Another issue is that if prices are a lot higher in one state than in another, people are going to cross borders to buy cheaper gas. This would be an issue, but not a huge one. There are already disparities in gas taxes between states, although Massachuetts’ population is generally concentrated away from borders, and the distance that it’s worth traveling to save money on gas is relatively small. People are also more likely to feel the pinch of prices and try to avoid them when overall prices are high, and when prices are relatively high—above $3—the price disparity would be relatively small. And when gas prices are high, this gives the state an economic advantage, as residents would have more disposable income for other purchases. (You could even allow some gas stations very near the border to opt out of some of the increase in the tax when the prices were low, but they’d also have to give up the tax rebates when prices were high.)

Why not use a Vehicle Miles Traveled tax, or a VMT? There are variety of reasons. First of all, a VMT does not have the benefit of encouraging people who do need to drive to drive smaller cars, which this would. Second, it would be harder to sell a VMT than a gas tax, as there is a vocal minority of citizens very worried about the government tracking their movements (note to the tin foil hat crowd: that ship sailed a long time ago). The gas tax works. The main issue is that since it is so rarely raised, it has lost the purchasing power it needs

The idea of this plan is to pair a gas tax increase with the sweetener that, when the gas tax is most perverse and regressive, the pain would be muted. This has the potential to appeal to a wider portion of the population, because everyone remembers high gas prices, and it would actually help to reduce those spikes. We certainly need to increase transportation funding. The gas tax is a good tool, but we need to think outside the box and figure out how to make it work better, no matter the price of gas.

Would this work politically? There’s no way to know. But it would diffuse the argument that if we raise gas taxes and then prices go up, it hits low- and middle-income families hardest. This would actually give a tax break to motorists at times of high taxes. Maybe that would be a rallying point and a compromise to get the funding that we need while reducing the adverse impacts of the volatility of gas prices.

Uber surge pricing: anti-congestion pricing and economic externalities

I’m a devotee of NPR’s Planet Money podcast, but listening to a recent episode had me screaming at my—well, at my earbuds—when they described how Uber’s surge pricing and business model was good for everyone. The episode is moderately informative, but I take issue with a major premise of their reporting: the argument that charging a lot for service and matching supply and demand is a benefit for all users. You can listen to the entirety of the episode here; the segment I take issue with starts around 11:00.

Here’s a quick breakdown of the conversation:

  • Uber benefits car drivers, because they can make more money when prices “surge.”
  • Uber benefits people with high disposable income ($150 for a trip from Manhattan to Brooklyn), because they can choose to pay a lot of money to get a car when demand is high.
  • Uber benefits people waiting for cabs because Uber gets more cars on the road, so the wait time for other cars is less.

The last piece is where I take issue. In a closed system, this would be the case. But it’s not. Here’s a quote from the show: “when drivers see an area of the city where fares are high … they all go there.” Remember, we’re talking about New York City. When there’s a congested area, Uber provides an incentive for more drivers to go to this congested area. This creates more congestion, which actually results in slower speeds for taxicabs, and—I would posit—fewer available rides for people not willing to pay a lot of money, not more, as they suggest.

Which is fine, in economic theory. However, when looking at traffic, “congestion” is shorthand for “demand.” So by attempting to provide efficient transportation options, it is a transfer of utility from the poor to the rich. Taxicabs are strictly regulated because they serve as a part of the transportation network, both in fare and in number. By sending a fleet of expensive black cars to a congested area, Uber squeezes out the more affordable taxis from the street. Someone willing to pay triple the fare for a ride might call and Uber, but someone who can’t afford that service now has a potentially longer wait for a cab (which can’t get to the area due to congestion) and then everyone has a longer ride due to the congestion. Part of the taxicab regulations—at least in theory—provide enough cabs to provide service without gridlocking the city (obviously, here too theory and practice are almost mutually exclusive). Adding more and more cars to already congested streets creates crippling congestion. That’s—to use economist speak—a major negative externality for everyone.

Now the argument could be made that people waiting for cabs should just take the subway. It’s a good argument; in the podcast they interviewed someone waiting in a 30 minute cab line, in the snow, at Penn Station for a ride “downtown,” an area well-served by the nearby subway (for lower prices, too). And for some of these riders, they might make that choice. But others probably have rational reasons to look for a cab: perhaps they have a bulky item. Perhaps they’re trying to get somewhere not well-served by transit. Plus, the added street-level congestion might drive more people to use the subway, which will create more congestion below-ground too. Adding more vehicles above-ground doesn’t seem like a great solution.

Uber’s surge pricing, in my view, is the opposite of congestion pricing. What works in London and Singapore has faced stiff opposition in New York City (especially from interests outside the city). With congestion pricing, Uber (in theory) would make more sense, since the overall traffic would (in theory) be mitigated enough that surge pricing wouldn’t send more cars in to already-congested areas. Until then, Uber’s surge pricing is part of the problem, not part of the solution. And it certainly doesn’t benefit the vast majority of travelers.

MBTA ridership and demand elasticity

This spring, I wrote about how the MBTA’s fares are really not that high. Fares went up on the order of 20 to 25%, and the T apparently expected ridership to fall by 5% or so. The numbers are in for the first month and while ridership did fall (the first decline, year-over-year, in 14 months), it was off by less than 1/10 of 1%.

There’s a local company which builds pricing software which is not surprised. They studied the base transit fare and calculated as long as it stayed under about $2.75 there would not be a major impact on ridership. Demand for transit is rather inelastic—people have to get to work—and they’re willing to pay an extra 60¢ a day (or $11 per month) in order to get to a job which pays many dollars per hour. For the average subway commuter, it’s still far cheaper to pay $70 for a monthly pass than to pay gas, tolls and (especially downtown, or in Kendall or Harvard) parking.

What’s interesting is where ridership fell. I expected that the increases in the cost of Commuter Rail to be a detriment to ridership. While transit and bus rose by 17% and 20%, commuter rail fares rose by 25-30%. And with higher fares to start, the nominal increase was $1.25 to $2.25. For monthly passes, which now range from $173 to $314, the increase in fares ranges from $38 to $64, which is not chump change. Commuter Rail customers are more likely to own cars than bus and transit users, and there was a lot of hand-wringing that commuters would abandon the rails and drive instead.

But that hasn’t happened. Even with lighter summer traffic, commuter rail posted gains. So did buses and boat traffic. The only declines were on The Ride (where fares doubled*) and on the Subway. Commuter Rail ridership is up. But the subway, where fares rose by a quarter and a nickel, is down.

However, the more I thought about it, the more it makes sense. While the subway prices stayed well within a range where they won’t have a major impact on demand, there is enough elasticity in supply to allow people to utilize other options. For most Commuter Rail riders, the trip is more than 10 miles and their only other option is driving, which is more expensive and subject to the whims of traffic. The cost of driving might be somewhat less marginally more, but it’s still more. So the supply is quite inelastic—even with fare increases the train is still cheaper than driving.

For subway riders, it’s a different story. Except for the Riverside and Quincy lines, most of the T’s ridership is concentrated within about 6 miles of downtown. Many subway riders have a commute which is only three or four miles long. The supply here is not constrained to transit or driving, but adds walking and cycling to the mix. With a mild summer (June was 1 degree below average, July and August 2 above) and expansion of bike sharing, as well as at-capacity rush-hour trains, it’s quite possible that many commuters looked at the fare hike and tested the elasticity of the short-distance travel supply. And that people looked at a mile-and-a-half ride on the train and decided to save $2 and take a half-hour walk. The price for walking or biking is essentially zero, so commuters were able to overlook the inconveniences of these modes due to the price savings. In other words, they pumped up their tires or put on their shoes.

Come winter, when it’s cold and rainy (or, like in 2011, snowy) these new riders may stream back towards the turnstiles (or, uh, Charlie Card machines). In any case, I doubt the drop in subway ridership is due solely to the rise in prices as much as it is due to riders exploring other options.

* I am torn how to feel about The Ride fares. On the one hand, paratransit is a lifeline for many disabled and disadvantaged groups who would otherwise not be able to get around without it. On the other hand, it commands a subsidy on the order of $40 per ride (if that was the rate of subsidy for all MBTA riders, the annual subsidy would be on the order of $15 billion per year—the cost of one Big Dig), and its ridership is growing exponentially. Paratransit is an important government service, but because it is shouldered by the T, a service for a few thousand riders a day is subsidized heavily by over a million other riders. A good solution, perhaps, would be sequestering funding for The Ride and funding it separately from the T at large.

I love car sharing. I just wouldn’t invest in it.

[They say that hindsight is 20/20. So I’ll preface this post with the fact that I was making this argument years ago, as in 2010, well before ZIP ever went public. (A quick search of my Gmail for Zipcar IPO did the trick.)]

Car sharing is great. It’s not applicable to every market (i.e. it works better in Boston than Houston) and in the short term, outside of dense cities with pricey parking it will be a niche product. But its users dramatically reduce the number of cars on the road, and they also reduce the amount that they drive when they pay the full cost of driving, not just the marginal cost after the static costs of insurance, depreciation and parking. From its infancy less than 15 years ago (most large city car sharing organizations—CSOs—launched between 1998 and 2002) it has grown to an almost-mainstream product with close to 10,000 cars and a million drivers in North America.

But I’m not about to invest in it. And Zipcar (ZIP), which went public last year, has promptly lost two thirds of its value.

Zipcar’s aim in an IPO was to both pay back their initial investors and raise money for expansion. However, as they expand, they are not going to be able to exploit economies of scale in the way many growing companies can. They have entered their best markets, and further expansion will be in to markets with lower margins and which require more marketing and long term demographic change (Atlanta instead of Boston). The heady days of the early ’90s, when car sharing organizations in dense cities threw cars on to the streets as fast as they could, have come to a close. The top neighborhoods are well-served. Future expansion will be slow and methodical, and not particularly profitable.

In fact, several successful car sharing organizations don’t really ever intend to turn a profit. The following are major car sharing organizations in the US and Canada:

  • Zipcar, for-profit start-up, founded 2000 in Boston, main vendor in Boston, New York, DC, Portland and Seattle, with a presence in all large markets except Montreal. Bought Flexcar (for-profit) in 2007; Flexcar had originally been formed as a public-private partnership in Seattle in 2000, and bought a Portland-based not-for-profit (Car Sharing Portland) in 2002 before expanding and being bought by Zipcar, apparently close to bankruptcy.
  • Hertz on Demand, for-profit spinoff of car rental agency, founded 2008, New York
  • Mint, for-profit spinoff of parking garage manager, founded 2008, New York
  • City CarShare, independent not-for-profit, founded 2000 in San Francisco (for a couple of years, San Francisco had competition between CCS, Zipcar and Flexcar)
  • I-Go, not-for-profit division of a larger organization, founded 2002, Chicago
  • PhillyCarShare, not-for-profit start-up in 2002 sold to Enterprise in 2011
  • CommunAuto, for-profit start-up in Montreal in 1994
  • Autoshare, for-profit start-up in Toronto, 1998
  • Modo, founded as The Car Coop, cooperative founded in Vancouver in 2000
  • Most smaller CSOs are non-profits (HOURCAR, Ego, Ithaca Carshare, Carshare VT) with a few for-profits (CommunityCar, OccasionalCar) and coops as well.
Interestingly, every city with even a medium-sized car sharing scheme, except New York and Washington D.C., has had, at one point, it’s own homegrown car sharing service. Zipcar only succeeded in using it’s size to run Flexcar in to merging (and this may have been more due to management policies on Flexcar’s part) and PhillyCarShare to a sale (again, quite possibly due to management issues). Washington, D.C. is the only city in which competition has markedly decreased (it and San Francisco had Zipcar-Flexcar competition before that merger; San Francisco retains competition with CityCarShare); the rest of the markets Zipcar has entered it has failed to dislodge the existing player in town. (That being said, with the exception of New York no one has entered a Zipcar market after they had established a presence.)
But look at how the major players were founded:
  • 4 as for-profit start-ups
  • 2 as other for-profits (Both in the New York market)
  • 4 as not-for-profits
  • 1 as a coop
About half of the players in car sharing started off as non-profits, and two (I-Go and CityCarShare) remain non-profit today. I’d be reticent to invest in a business model where nearly half of the original founders didn’t see the opportunity or necessity for profits.

There are three main issues which will keep car sharing from ever being a terribly profitable business. (Yes, it likely will achieve a steady level of profitability in some markets and may pay small returns to investors, but won’t grow dramatically.) One is the issue of car sharing being a capital-intensive business. The second is that, charging by time, CSOs are hamstrung by there being only 24 hours in a day. Finally, for both mission and practical reasons, car sharing organizations are one of very few businesses which often actively encourage their customers to use their product as little as possible. None of these helps the bottom line.

As far as capital goes, car sharing is expensive. CSOs have to buy (generally) new cars, install a couple thousand dollars of hardware, and continually maintain them. Unlike rental car companies, which can sell their fleet seasonally or to adjust to market conditions, CSOs decal their vehicles and install locking mechanisms, relays and mileage tracking devices. These investments raise the barrier to selling capital quickly, and most CSOs keep their vehicles for several years, not several months. Having a lot of capital tied up in depreciating assets encourages stability and slow growth, but not necessarily high profits.

With millions of dollars of capital tied up in vehicles, car sharing organizations are hamstrung by the number 24: the number of hours in a day. This, however, is misleading, because very little car sharing takes place overnight, when many CSOs have reduced rates. In addition, most car sharing organizations have daily rates, which allow consumers to cap the cost of a shared car at around the cost of an 8 hour rental, but this also caps the potential revenue at 8 hours. Since car sharing breaks even with between 5 and 6 hours of use per day, this leaves a small margin for profitability. And even without daily rates, customers become agitated when availability dips, generally around 9 hours a day (50%, assuming the hours of 12 to 6 a.m. rarely see car usage).

Is there room for profitability between the break-even point and the daily rate barrier? Yes, but not much. Usage is lower on weekdays, and if cars saw maximum profitability on weekdays they’d be swamped on weekends. Thus, most CSOs have higher rates on weekends (this is what we call supply and demand). And since the fixed costs of buying, parking and insuring a vehicle is quite high, there is little room to reduce the break-even point. Some cars in high-use areas will always profit, but a CSO will generally have fringe cars in neighborhoods which see less use (but absorb extra demand and expand the geographical market). And cars which do very well in the summer often see usage decline in the winter, but with annual parking space contracts and investments, CSOs can’t change their fleet size based on the season. So making a profit in car sharing is akin to threading the needle of a balance sheet.

Finally, once a member joins a car sharing organization, they aren’t generally encouraged to drive as much as possible. For one thing, some members might find that, by driving a lot, they would be financially better off buying a car, depriving a CSO of a highly-valued member. In addition, non-profit CSOs, and many of their for-profit brethren, are motivated by environmental goals, and find it hard to encourage their members to “drive as much as you can!” So car sharing is one of very few businesses where a consumer is told to buy a product and then use it as little as possible. (Imagine, for instance, McDonalds running ads saying “cheeseburgers are a great alternative to eating at home, but really you should eat our cheeseburgers only when you’re really in a pinch for time.”) CSOs will encourage their members to use the cars, but have to tread a very fine rhetorical line to encourage “exploration” and “new experiences” rather than just driving more.

Put these together, and car sharing is a maturing service which will continue to grow. However, it’s not something that’s ever going to turn a huge profit.

On the other end of the spectrum, current investors, like the esteemed (at least, when he’s getting yelled at by Jon Stewart) Jim Cramer, also have things wrong. They argue against buying Zipcar shares because they fear big car rental agencies are going to push in to Zipcar’s turf and take over its business. If this was the case, it would have happened by now. But car rental agencies have a different business model, and they’re not about to change their stable, successful businesses to jump in to something new, and something which barely makes money.

In other words, Hertz is not about to move their airport-based fleet of rental cars in to the middle of the city and take over Zipcar’s business. Their business model is to buy new cars, drive them for 25,000 miles, park them in big lots at an airport, and sell them within a year. They operate with very low overhead, comparatively few locations, and, outside of airports, limited hours. Some have made feints in to the car sharing market from static locations with limited pick-up times and key boxes; these have been unsuccessful. The only money they put in to their cars are 25¢ key rings, so they can sell off capital at a moment’s notice. They don’t have to invest in parking spaces, signs and space leasing agreements. They play the car market, and, from time to time they rent them out.

The one opportunity for car rental agencies is that the demand for rental cars and shared cars is flipped. Rental cars see low use on weekends, shared cars see low use on weekdays. If Hertz could magically move half their fleet from the airport to the city every Friday and back on Sunday night, they’d have something going. This, however, would mean driving their cars at high revenue and high traffic times, from one central location to dispersed spaces in residential areas. They’d have to install car sharing hardware which would only be used a few days a week, disable and reenable it to meet the whims of the rental car market, and convince traditional renters to drive a moving billboard. The logistical and personnel cost to do this would negate any potential savings. 

There are also reverse opportunities for car sharing organizations: encouraging travelers to rent shared cars during the week—especially in cities with good transit links to airports. Imagine, for example, a business traveler going to the East Bay from SFO. She could get on the BART and ride it through San Francisco and under the Bay, avoid the usual traffic on the bridge and 101, and pick up a car in Berkeley from near a BART station. By the time a traveler bound for a location west of Boston was at the desk after a rental car shuttle trip at Logan, they could take the Silver Line to the Seaport, grab a Zipcar, and jump on the Pike. If anything, I think there’s more of a market for traditional car renters to use shared cars than the other way around, especially as business travelers become more likely to be city-adept car sharing members and less likely to own a car. (Plus, from an expense-reporting standpoint, it’s easier to expense one item, a rental car, than a rental car and gas purchases.) But for the most part I expect car sharing and car rental to remain separate business segments as this would only be feasible in markets with sizable car sharing markets, good airport transit and cross-city reciprocity (right now, only Zipcar).

So, car sharing will continue to be successful, but never wildly profitable. If shares were to fall far enough and Zipcar shows continued profitability, I could see it being a decent investment, but not one that would ever run up Apple- or Google-like returns. For now, I’m staying away.

Why the 2010 census does not paint a full picture

In a recent post, Andrew Sullivan linked an article by Joel Kotkin arguing that the trends shown by the 2010 census show a continuing flight both from cities to suburbs and from denser cities to less dense ones. This page has never been a fan of Kotkin, who is continually trying to reprove his thesis from 2005 that the suburbs were the way of the future and that the cities of the future were “Orlando, Fla., San Bernardino-Riverside, Calif., Phoenix and Las Vegas.” (Seriously, this 2005 piece gets just about everything wrong.)

I am not going to waste time arguing several of the points Kotkin raises (for instance, he extols falling condo prices without mentioning that suburban house prices in many markets have plummeted just as much) or go in to the politics of subsidized home ownership (Sullivan uses another blogger to point out that high urban housing prices stem from high demand and policies which restrict growth in supply, further subsidizing suburban housing). Instead, I am going to look at recent, post-housing bubble data that shows that Mr. Kotkin’s thesis may be based mostly on the first half of the decade, not recent years. (He goes on to compare the 2009 American Community Survey and the 2010 Census, which are different data sets—so he’s comparing apples to oranges.)

To do so, I am going to use the census data beautifully displayed by Forbes (and commented on by several, including the Kotkin). I’ll concur with one of the other commentaries (all are worth reading, especially interesting is one point out the income difference between in- and out-migration) that there are large swings between 2005 and 2009, but point out that they are not only between certain cities but within them as well.
For each, I’ll show the migration chart from 2005 to 2009, and then the local migration maps from 2005 and 2009. And, yes, it would be great to have data on a more granular level (i.e., by zip code) or more aggregated level (by metropolitan area) but these doen’t exist, as far as I can tell.
1. Suffolk County, Mass. (Boston)

In the second half of the decade, out-migration from Suffolk Country went from well outpacing in-migration to falling behind it, and it fell while in-migration slowly rose. Locally, adjacent counties in eastern Massachusetts were recipients from Suffolk County each year, but further-flung counties, including sprawl-heavy Worcester County and those in Southern New Hampshire, flipped from receiving many migrants to giving them back.

Further afield, in 2005 Boston lost migrants to most every sprawling metropolis, including Dallas, Las Vegas, Atlanta, Charlotte, Jacksonville, Orlando and Miami. It gained population from all these centers in 2009. Middlesex County, including the very dense cities of Cambridge and Somerville (but large exurban areas as well adjacent to New Hampshire) shows similar, if less pronounced, trends.

2. Philadelphia County, Penna.

Philadelphia’s migration trends are not as pronounced as Boston’s—there is no switch from net out-migration to net in-migration—but it mirrors those of its northeastern neighbor. In the region, we can note a pronounced shift away from migration to far-flung suburbs, especially the second ring in Pennsylvania and New Jersey. There are similar shifts to the nearby suburbs in both years (and a similar draw from North Jersey) but the out-migration to the exurbs disappears. On the national front, there were fewer major shifts (like in Boston) but still several trends away from less-dense areas.

3. Washington, D.C.

The demographic trends for Washington show similar patterns, especially on the national scale. It’s out and in migration flipped, like in Boston, although the changes were less pronounced, as it never had the same mid-decade baseline losses as Boston. Locally, there was less movement towards some outer suburbs, although this can be confounded by the presence of Baltimore to the north, which doesn’t really qualify as a D.C. suburb. Still, there was less movement to counties far off in Northern Virginia in 2009 than 2005. Nationally, DC lost residents to Charlotte, Atlanta, Phoenix, Las Vegas and Los Angeles. By 2009, it was gaining residents from all of these.

While I wish I could go on this thread forever, I can’t, because the data (here, at least) is at the wrong scale. New York City is broken in to counties, and it’s so big that each county has different trends since there is so much migration within the city, as well as in to the city from the rest of the country. San Francisco doesn’t show such trends within the metropolitan area because the population is so spread across counties, and except for SF itself, most of the counties have significant urban and rural populations. But, further afield, San Francisco saw dramatic changes: in 2005 it lost significantly to LA, Las Vegas and Phoenix, in 2009 it gained population from these three cities. And in Atlanta and Chicago (Fulton and Cook counties), there weren’t major changes in national out-migration, but dramatic reductions in moves to the outer suburbs.

In any case, go and explore the Forbes map. It’s a lot of fun, both from a geography-nerd level and an exercise in proving Joel Kotkin wrong.

Re-imagining a gas tax chart

There was an interesting article on The Infrastructurist a few days ago about the gas tax. It poses the question of whether user fees might be a better idea (my thought: way to complicated to get the same result). And they show a chart of the gas tax since its inception in 1932:

That’s interesting, but, well, wrong. If you look at this chart, it sure looks like the gas tax keeps on rising. Look, it’s gone up 450 percent since 1982! The government just wants our money! We can’t raise the gas tax (seriously, we can’t, it’s a third rail). But it’s not really rising.
Once in a while, we need money. And we raise the gas tax. On the chart below, the blue line is the same as the chart above—see how it rises? Now look at the red line. That’s the same value—except adjusted for inflation. Once adjusted, the gas tax has varied, from about 9 cents to about 29 cents, in the past 80 years. 
It’s interesting when it was raised: first in the 1950s (when the Interstate system was funded) and then in the 1980s, after the oil scare. And if it seems like it’s taken a while since it was last raised, look at 1959 to 1982, when it went from an inflation-adjusted 29 cents to an inflation-adjusted nine. If we wait that long, it will take until the end of this decade to raise the gas tax—and it still won’t fall as far as it did in the late 1970s unless we have dramatic inflation again. So, yes, the gas tax should go up. But, no, it’s not at a historically low level. However, we haven’t really raised the gas tax, well, ever. We only raise it as a reaction to it being too low. (Gas tax data from here, inflation data from BLS)
Quickly, why is the gas tax good? Well, first, why is it bad? It’s regressive. Everyone pays the same. But why is it good? Well, in addition to raising revenue, it has tons of positive externalities. It taxes heavy users more than light ones (and people without cars get off scot-free). It encourages people who need to drive to buy smaller, more fuel-efficient automobiles. It encourages people to move to areas where they don’t need a vehicle, which are inherently more efficient. Less petrol consumption means less pressure on us to buy oil from unstable, foreign nations. It’s very economically sound: you’re not forcing anyone to do anything, but you are able to affect change simply through taxation. And, finally, it’s really, really hard to get around. Smuggling gasoline is hard, and gasoline is bulky and dangerous to transport. Drugs and cigarettes are easy to sell on the black market. Gasoline? Not so much.
I’m sure we’ll get to the gas tax more in the future. But, for now, remember this chart.
(Yeah, I know I haven’t been posting here in a while. Skiing has gotten in the way.)