Is congestion pricing — charging more to use a facility when it's in high demand, in order to decongest that facility — an unacceptable burden on the poor? Joshua Arbury of the Auckland Transport Blog asks this on my recent post on congestion pricing terminology:
Fundamentally though, there's a political question to consider. Is it acceptable to have a road pricing scheme that prices the poor off the roads to create more room for the rich? Because, in a nutshell, that's effectively what road pricing/congestion charging/decongestion charging is.
These discussions will never get anywhere until we can separate two completely different questions:
- What is the accurate price for this facility that arises from the relationship of supply and demand?
- For whom, and for what social purposes, should we offer discounts from that price?
Australian and British transit agencies offer a nice example of how to keep this clear. As in North America, transit in those countries routinely offers discounts to senior citizens and the disabled, because, well, we as a society want to honor those people. Those discounts don't serve any particular transit agency mission, but the society judges them to be important. Fine. So in both Australia and Britain, the cost of those discounts is added up and a central government reimburses the transit agency for that amount, as a "shadow fare." The transit operator can then count all riders as equal, and compete for all riders as being of equal value to it, because it experiences them as all paying the same fare.
That's very clean, because it separates transit's real purpose from a separate (perfectly valid) social agenda, rather than just expecting transit operators to pay the cost of the social agenda. Seniors and disabled persons get their discounts, but the transit operator continues to value their patronage as much as they would value that of a full-fare customer.
Meanwhile, back on the roads, our current prevailing road pricing policy is that "when demand for road space exceeds supply, government will subsidize everyone's travel so as to elminate any monetary cost." The effect, of course, is that instead of paying in money we pay in time. That's what congestion is. Something of value is being given away for free, so we have a long queue of people waiting for it. If you want to "fix congestion," you have to change the price.
Is that hard on "poor" people? Yes, it is, like many things. The answer may be to subsidize the price as an expression of a social objective, in exactly the way that British and Australian governments subsidize senior/disabled fares as something entirely separate from other transit subsidies, thus enabling them to more cleanly connect each spending to a public purpose.
I wouldn’t even accept the premise that congestion pricing “prices the poor off the roads to create more room for the rich.” It prices discretionary and marginal trips off the road. People for whom driving is barely better than transit would switch, as would people who can easily take a trip at less congested times, leaving more space for people who have fewer alternatives.
US transit agencies, OTOH, typically have to pay for the social benefits of senior discounts themselves–not to mention the cost of providing paratransit service, an expensive social service operation which is mandated by the feds.
My personal preference for how to handle this sort of question is to take service providers out of the social-service discount business altogether, and provide general social insurance programs that take things like poverty and health into account. Some transit advocates might not like that, on the concern that were the poor given vouchers for travel rather than things like subsidized transit fares, they might turn around and buy automobiles instead; but my preference also assumes that roads are priced appropriately. (There’s also the problem that naked social welfare programs are more likely to be attacked as “welfare”).
The premise is also simplified in a way that make the argument invalid or at least misleading: “a road pricing scheme that prices the poor off the roads” would be correct if everybody now owned a car. This is not the case, and insofar as the government subsidizes automobile travel, any decrease in that subsidy will spread the money elsewhere and thus potentially help the poor who cannot afford to drive. You can even deliberately use some funds to run better transit service, which is not a handout to the poor because anyone is free to take transit, but it does help them.
It’s really easy to use the excuse “well, if we cut this subsidy the ‘poor’ lose out” even though that’s probably only marginally true and only in the short term. Taken in the aggregate and over time, such thinking bankrupts governments.
To expand on Bradley’s point, the congestion pricing opponents pulled this trick in New York. It was a complete lie here because 96% of commuters to the CBD take the train.
In other cities where a larger percentage of the population drives it may be more regressive, but even those cities have 10% of the population that doesn’t drive, and they would benefit from congestion pricing.
I recall reading that in New York, those who drive into Manhattan make, on average, $20,000 a year more than those who take the subway into Manhattan.
I would assume that most congestion pricing schemes would be for transit-rich CBDs, so I would suspect that New York is no different than anywhere else. The people who drive into the CBD can afford it. If people “needed” to drive for their commute, (like many outer-borough commutes in NYC, due to the radial subway and slow buses), they probably wouldnt be going into the CBD.
“As in North America, transit in those countries routinely offers discounts to senior citizens and the disabled, because, well, we as a society want to honor those people. ”
Is that really the case?
Is not because seniors are, theoretically, retired, and thus don’t need transit during the commuting hours? So then in turn offering a cheaper fare is to get them to ride for their other needs, which tend to be when the system has excess capacity.
Think of it like transit systems that offer free rides for kids. They’re not “honoring” kids, they have excess capacity, and by offering free rides for kids, they can attract paying adult fares. 4 full fares might be more expensive than parking and gas for a single car, so by letting the kids ride for free, that equation is reversed.
As for the disabled, I sort of see it as “our system doesnt really work for you, so heres a discount because you wont get full use of the system”
A good argument can be made for congestion pricing of transit fares, to encourage non-peek use; but that’s a different policy than giving cheaper rides based on age. Child fares make more sense, for the reason you state, but again–such fares are typically offered all day.
But giving “senior discounts” is a longstanding practice in the US, and most businesses which do so don’t do it because seniors shop during non-peak hours. The longstanding rationale is that retired people have fixed incomes and thus are entitled to some assistance; unlike other need-based benefits which are stigmatized as “welfare”, most everyone gets old at some point, so there’s little stigma attached to the practice–and in many businesses, it’s practically expected.
At least at most businesses, cheaper prices for old people don’t exist to “honor” them, nor to do they exist out of some sort of charitable desire to help those less able to pay. They are simply price discrimination. If a business can manage to charge price-sensitive customers less and price-insensitive customers more, it can make greater profits than if it had to charge a single price to everyone, so businesses will tend to do this without any subsidy or government involvement. This is the same as the rationale behind discounts for university students. I would think this is the origin of senior discounts with most US transit agencies.
The British system is something different, where seniors don’t pay at all and the government pays a (full?) fare in their place. This can be justified on various social grounds, but one should realise that it is still a market distortion – seniors will take more transit than they would otherwise as a result (Land’s End to John o’ Groats, free on local buses!) and bus operators can respond by raising their fares so the government pays them more and more for each senior who rides while other users are priced off the bus.
Ah interesting that you’ve picked up on this Jarrett.
Of course I was over-simplifying things a bit by saying that congestion charging prices the poor off the road and certainly you can say that the market system prices the poor out of buying Porches, but we can live with that (and fair enough too).
Phone companies (often) charge more for on-peak calling, roses cost more around Valentine’s Day, some foods become extra expensive just before christmas and so forth. We manage demand this way, it’s just how the market works.
I guess the issue with pricing transport in such a way can be related to how fundamentally basic it is to be able to get around the city. If congestion charging prices you off the road, yet the public transport (or walking or cycling) alternative isn’t good enough then your ability to get to work and be economically productive becomes mighty difficult.
My most recent post on congestion charging (http://transportblog.co.nz/2011/03/20/congestion-pricing/) suggests that we need to figure out, based on the characteristics of the city we’re dealing with, two questions:
1) Why are we pricing road space (to raise more money or just as a smarter way of raising the same amount of money)?
2) How should we price road space (CBD cordon, GPS tracking etc.)?
In cities like London and New York, with hugely developed public transport alternatives, but a need to raise more money to keep them going, pricing road space to raise more revenue and to keep down congestion makes good sense. The extra cost is unlikely to fall on the poor – as generally those who continue to drive into cordons are able to easily afford the charge. The vast majority of people catch the bus or train.
Big cities also have the advantage that a CBD cordon approach is unlikely to kill off their city centres. If, for example, we were to put a CBD cordon congestion zone around Auckland’s city centre tomorrow, the biggest result would probably be the death of the city centre with retail and offices racing each other to business parks and shopping malls in the suburbs. So we need to be careful about issues like that.
If we assume that we’re applying a congestion charge simply as a replacement of existing road fees (ie. lower local government taxes and/or lower fuel taxes) then we are likely to improve political support, but we then need to ask questions about how efficient it is to collect the money compared to current measures like fuel taxes. A lot of the money raised from congestion charging in London simply goes towards running the system, the same is true for a number of toll roads (including one north of Auckland). By contrast, the administrative cost of collective fuel tax is negligible.
Ultimately, I think if your PT system is good enough then congestion charging is unlikely to hurt the poor. They’re probably fairly unlikely to be driving anyway. However, if your PT system is pretty rubbish (like Auckland’s) then you do have huge swathes of the population that are highly dependent on their cars to get to work – particularly poorer people as their jobs tend not to be clustered in office blocks easy to serve with public transport. In such situations, congestion charging is likely to have significant social equity issues.
The key is ensuring quality alternatives to driving are available.
I want to start by saying that I’m actually in favor of congestion charges; however, I’m not sure that “you’re paying in time or you’re paying in money” is a good response to arguments that such charges price the poor off the roads. It ignores the fact that poor people don’t have money, but do have time — in fact, people who are un-, under-, or irregularly employed actually have less time constraints than those with full-time jobs. Having, say, faster-moving tolled lanes allows people in a hurry to exchange money for time, but congestion charging essentially asks driving society at large to make this choice.
I got into a bit of an online spat with someone once who believed that if Amtrak’s Northeast Corridor could be improved to the extent that Washington-DC trains took two hours or less and maintain current Amtrak fares, it would bankrupt Chinatown buses and Bolt Bus and the like, simply because it would be such a better deal. I really don’t think this is the case. There will always be people for whom a $150 round-trip ticket isn’t justifiable when a $50 alternative is available, even if the trip is more than three times faster.
I think the argument about congestion charges can fall into this same sort of myopia.
To elaborate further on jfruh’s remarks… one bit of economic doublespeak which has long annoyed me is the concept of “highest, best use” in the context of pricing discussions: the claim, that by accurately pricing some good or service, social utility (of some sort) will be maximized.
What is true is that increasing the price of something will exclude those unwilling or unable to pay for it; and that ideally an equilibrium can be reached where there is neither unmet demand at the price nor excess supply. (In practice, this does always occur). While there is probably some correlation between the willingness of someone to pay a higher price, and the social merit of their use (persons whose need for a good or service is dire will, all else being equal, be willing to pay more than those whose need is frivolous); many other factors unrelated to any concept of social utility come into play. First class airline tickets aren’t sold, generally, to those with physical needs for larger seats; but generally to the well-to-do, who are satisfying other needs (comfort and social status) with their purchase. Yet many economists (and politicians using economic arguments) seem to treat it as axiomatic that those willing to pay more really are the “highest, best” consumers of something.
That said, it always amuses me to hear right-wing politicians, of the sort who generally oppose social safety nets, defending the right of the poor to operate motor vehicles free of any social encumbrances such as tolls or congestion charges. Road socialism, indeed.
Jfruh. The "poor" people I know would bristle at being told that they have surplus time.
Your comment also suggests that we need to think differently about CBD cordon charges vs motorway HOT lanes. I think it's the same principle; in either case there needs to be a transit option, though in the case of the motorway the motorist also has the option of staying in the slow lanes. The main point is that everyone needs to be able to complete the transactions that they can make today, so that the charge doesn't suppress the economy.
The fact is, there's a limit to the ability of government to feel and heal everyone's pain. Congestion charging arises from a fundamentally conservative view of government which says that its role is to promote fairness rather than equity.
@EngineerScotty: Poor people driving to work have jobs that generate less value than those of rich people driving to work – if not, they wouldn’t be poor. Giving the commute-time road space to people whose jobs pay the most really is likely to be the “highest, best use” socially. (Admittedly in some industries salary may be poorly correlated with value created, but that’s an entirely different problem that should be addressed directly.) Likewise it makes more sense to allocate first-class airline seats to business travelers who need to be rested for work/meetings on arrival than to leisure travelers who are planning to rest at their destination anyway, and this is mostly the effect of the current market pricing.
@Jarett: “The main point is that everyone needs to be able to complete the transactions that they can make today, so that the charge doesn’t suppress the economy.” The transactions made under the status quo are not necessarily the economically optimal ones to make. Perhaps most road-space in a congested area today is being used to make low-value transactions (e.g. to drive to jobs that would cost only slightly more to automate), whereas road pricing would make these transactions impossible but thereby free up space for more high-value transactions.
I also don’t see why everyone expects that poor people who have to drive to work would really bear the full cost of congestion pricing. If CBD cordon pricing were implemented somewhere auto-dependent, a low-wage employer like McDonalds in the CBD would have to pay its employees more, because otherwise they wouldn’t be able to afford to get to work/their job would no longer pay more than welfare after accounting for costs. This would in turn be reflected in the cost of eating at McDonalds in the CBD, with the net result being that people eating there would pay more of the social cost of their actions, which seems more optimal economically.
Poor people driving to work have jobs that generate less value than those of rich people driving to work – if not, they wouldn’t be poor.
Less value to who? Certainly, employee compensation is related to the amount of value a worker generates for his/her employer (though other things, such as bargaining power, factor into it as well); but this is a different question as to how much value is generated for society. There’s plenty of junk bond traders, professional athletes, and such who produce millions for their employers (and are compensated appropriately), but whose value to society is nil, other than taxes paid.
This doesn’t mean that congestion pricing of some form isn’t good policy–in general, I think it is–but again, I have objections to the conflation of financial net worth to societal worth. The presumption that the rich are Worthier Than The Rest of us–an idea which seems to dominate a lot of political discourse in the US and other parts of the west–is IMHO a deeply flawed one. As is its inverse–the idea that the rich are all evil and corrupt. (In many ways, as Chait explained, supply-side economics, especially in its more Randian forms, is best understood as an inversion of Marxism).
It’s also worth pointing out that the poor–including those who do not drive–already pay for the roads through general taxation. This is especially true for local street networks, which are generally paid for out of municipality’s general fund (rather than via excise taxes on fuel). In many parts of the US, municipalities derive revenue from property and/or sales taxes, both of which have a tendency to be regressive in nature. A shift of road funding from general taxes to user fees might well benefit the poor, even those who do choose to drive.
@EngineerScotty, would your objections be addressed by scaling congestion charges to income? This is done with traffic fines in some countries. At the same time, I have to wonder: is all this talk of pricing the poor off the roads really a front for rich people worried about dispelling the perception of the “egalitarian” status of driving? If the poor are indeed priced off the roads, then they’ll be much less likely to support huge road subsidies for the rich, once it becomes painfully obvious who’s benefiting from them. Driving cars on the public roads is a privilege, and it would only be fair that those who benefit from this privilege pay for it to compensate those who do not, or cannot.
This is an entry from my website publictransport.about.com:
Los Angeles is in the process of constructing High Occupancy Toll (HOT) lanes along existing carpool lanes on the 10 and 110 freeways with the assistance of $210 million in federal grant money that became available when New York City’s attempt to introduce a Manhattan congestion charge failed. When completed in 2012, these lanes will allow motorists driving alone to pay a toll to access the lanes which are currently restricted to transit and carpoolers. Money generated from the tolls will go towards upkeep of the freeways and improved transit service along the freeway.
Although this project is slated to open next year, it has recently come under attack from both Republican representative Gary Miller, who claims that since taxpayers already have paid to build the freeway they should not have to pay again to drive on it, and Democrat representative Maxine Walters, who claims that since only rich people will be able to afford to drive on it it is unfair to the poor. Both of these arguments are fallacious; in fact, the success of this HOT lane, which, if successful, is likely to be replicated in other American cities, is to a large part dependent on how much political interference is allowed in its operation.
First, Gary Miller’s argument is clearly nonsensical. It is akin to arguing that once you buy a car you should never have to pay for an oil change because you already paid for the car, or that once you buy a house you should never have to pay for a plumber or a house painter since you already paid money for the house. The cost of paying for the lanes only paid for the initial construction, it did not pay for maintenance. While theoretically gas taxes should pay for the maintenance, the fact that the federal gas tax has not been raised since 1993, in conjunction with the increased fuel economy of cars, has meant that for the past few years funding from the general budget has had to be redirected to the transportation trust fund to ensure its solvency. Someone has to pay for the upkeep of the roads, and it seems better that the people using the roads should pay rather than the general automobile driver.
Maxine Walter’s argument is simply not born out by the facts. Studies like this one (http://www.cato.org/pubs/regulation/regv23n1/poole.pdf) have shown that everyone benefits from HOT lanes, not just the wealthy, because their existence offers them a choice to pay extra for a speedier trip when the purpose of the trip is extremely important, like a job interview. I do not usually agree with Robert Poole and the Cato Institute, both of whom are anti-transit, but in this case they make a good point for HOT lanes. In addition, the fact that some of the toll proceeds will go towards the operation of improved bus service along the respective corridors means that low income people who cannot afford cars will still benefit from the lanes. More importantly, the fact that not everyone can benefit equally from something does not mean that we should not do anything at all.
The above two arguments have been repeatedly brought up in regards to HOT lanes and have been able to be overcome. The true obstacle to HOT lane success will be if political interference in operation of the lane means that the tolls can not be optimally set. For example, if the maximum toll, or ceiling, is too low that means people that would be willing to pay more to use the lane will be stuck in traffic; and if the minimum toll, or floor, is set too low not enough vehicles will use the lane which will result in an inefficient volume of traffic. To my knowledge every HOT lane has a toll floor and toll ceiling that vary based on time of the day but true economic efficiency demands no such floor and ceiling. Why should a HOT lane not be free at 3 A.M. or cost $50 to use at 5 P.M.? Let us allow the consumers to decide how much their commute time is worth to them and not politicians.
In fact, the toll lanes on the 91 freeway in Orange County, California are guilty of having too high a price floor. On a recent off peak direction trip (westbound in the P.M. peak hour) the price was $2.05 and the lanes were virtually empty; the price should have gone down – to zero is necessary – to have better usage of the toll lanes.
I’d just like to add that to me one of the keys about congestion pricing is choice: you have the choice to pay more to get where you are going faster. Though the comparison is not necessarily applicable, should we ban the sale of Lexuses because not everybody is able to buy them?
Chris, did you really just quote Cato as a source?
@Chris, I kind of doubt that the HOT lanes will actually lead to any improvements in transit service. Letting more cars onto the El Monte Busway will make it more congested, which will increase travel time, which will require more buses and probably caused reduced patronage, which will probably eat up all the revenue collected from the HOT lanes, after the infrastructure for collecting that revenue is paid for.
Stopped clocks are right twice a day, Alon. And there appear to be a few staffers at Cato who actually bother to apply libertarian dogma to roads and highways–not every “libertarian” writer is a shill for energy interests, or using the philosophy as cover for anti-urban agendas.
@Anonymouse, you could see what L.A. is trying to do with the ExpressLanes. Some of the revenue has to go into improving transit service along the corridor.
Unfortunately, the money is going into the Silver Line, which connects the two busways/soon-to-be-HOT lanes. The Silver Line has also become a “walled garden” service, replacing the express/local buses that ran on the corridors. Now, the single-seat ride has become a forced transfer at both ends.
I believe the HOT lane proposal in LA would ban single occupant vehicles from entering lanes when the HOT lane speed fell below a prescribed minimum, perhaps 45 mph. If followed, that should keep transit speeds up.
In public finance, the term second best refers to a solution that is less than optimal but may apply to (de)congestion pricing. Suppose part of the toll revenue was spent improving transit service in the tolled corridor. Further, suppose the poor were provided subsidies or discounts on the tolls.
If dynamic tolling produces free flow travel all modes and all incomes benefit. The most poor do not even have cars, they ride transit or bikes or do not travel highways at all. improvements in transit flow and frequency (subsidized by toll revenue) would be a second best solution to the optimization problem.
In Orange County, studies have shown that drivers of all income levels use the HOT lanes. all have limited time and time is what free flow provides. Wage earners need to get to work on time; parents picking up kids from day care need to be on time.
congestion is transport market failure due to not pricing scarce lane space.
@Scotty: sure, but you wouldn’t refer to a stopped clock even at the two times a day it’s right. You’d refer to an external source corroborating that the stopped clock is indeed right.
One of the chief proponents of congestion pricing in the Portland area (an idea which doesn’t, at the current time, have much traction but has been discussed) is our local anti-transit libertarian think tank, the Cascade Policy Institute.
While I frequently disagree with them–I’m not a libertarian and thus do not consider “gubmint is bad” to be an axiom–they at least are somewhat consistent in their principles. Unlike a certain nationally prominent ex-Portland libertarian writer I can think of, whose drunk far too much petroleum-laced Kool Aid.