Tuesday, February 21, 2006

Car parking levies – do they reduce congestion?

Melbourne has this year introduced a levy on car parking spaces in the city centre and owners of car parks will have to cough up the first fee next week. The fee is a flat fee of $400 per space per year, doubling to $800 next year and then rising with inflation (indexed to CPI).

Essentially it applies to all commercial spaces. Residential spaces are exempt - as are visitor spaces, loading bays, disabled spaces, spaces provided free by bodies such as schools, hospitals and churches, and some others.

The stated aims of the levy include reducing peak hour traffic congestion, pollution and greenhouse gas emissions in Melbourne’s CBD and encouraging the use of public transport.

This got me thinking: Do car parking levies really work? Is this likely to make any real difference to Melbourne’s city traffic congestion?

It’s difficult to find any hard data on the effect of car parking levies – most of the information comes from self-serving submissions by car parking operators and property owners asserting that car parking fees make no difference to congestion and will make all businesses flee the city in terror, taking their jobs and money with them etc etc. I imagine it’s very hard to isolate the effect of car parking levies from the host of other influences on traffic in cities.

The relevant stats that I could uncover suggest the following:

  • There is a response from commuters to higher car parking charges: a Vancouver study (PDF here) suggested that each $3 increase in parking fee reduced the probability of driving to work alone (as opposed to carpooling or using public transport) by about 10%. Another study from the University of California (Berkely) (PDF here) suggested an average price elasticity of demand for parking of –0.32 (which if my maths is right would suggest that a doubling in price would reduce demand by about a quarter). There must be some Australian studies on this surely?
  • In Sydney, according to some figures, 7.7% of vehicles travelling through the City of Sydney each day use off-street parking (which would be affected by a levy - another 7.5% use on-street parking and 85% is through-traffic). This figure is probably higher now that the cross-city tunnel has diverted some of the through-traffic out of the city. Still, if we’re generous and say that parking levies are passed onto parking users and result in 25% of those users switching to other modes of transport, that still only reduces congestion by about 2% (although it might reduce peak hour congestion by substantially more than that because many more peak hour users are commuters who will require parking).
  • Everyone agrees it’s a solid source of revenue. There’s some bickering about how fair it is, but taxing people who drive to work in the CBD of a large city seems intuitively a pretty progressive tax (many of my acquaintances work in the city and few of them drive to work – those who do are certainly the wealthier ones).

Here are my thoughts on its likely effects on congestion:

  • Demand for parking is fairly inelastic – it will take a large increase to induce much change in behavour.
  • A parking levy will mostly be passed through to consumers (because demand is inelastic).
  • This increase in price will have a small impact on demand for parking, encouraging some drivers to use other modes of transport. (A very high fee would have a substaintially larger effect of course).
  • If it is true that about 8% of city traffic uses off-street parking in the city (92% of city traffic is just through-traffic or uses on-street parking), a parking levy will have a very small impact on congestion overall. (We need other measures to catch these other drivers – tolls for driving through the city and obviously fees for on-street parking).
  • However, it might have a higher impact on peak hour congestion, as a greater proportion of peak hour users are commuters who will park their cars all day in off-street parking.
  • It is a solid source of revenue. It is progressive, targeting relatively wealthy individuals, it is relatively non-distorting (because it targets a product the demand for which is inelastic) and any ‘distortions’ probably act to counteract market failure due to the unpriced public good of driving in the city. It’s easy to administer. It also has a ‘polluter pays’ effect which gives me a nice warm feeling (though I acknowledge that there are probably better ways of targeting the polluters here).
  • If car park levy revenues are used to improve public transport, that is something that could have a substantial effect on congestion. Part of the reason for the inelasticity of demand for parking is that public transport is not a perfect substitute for car travel but if its reliability, speed, comfort and convenience are improved, it becomes a better substitute.

In summary, car parking levies seem a good source of revenue to direct towards public transport but on their own are probably a relatively ineffective method of reducing inner city traffic congestion.


David Jeffery said...

By the way, if you're interested in the economics of traffic congestion, the Becker-Posner blog had a great post on it last week: http://www.becker-posner-blog.com/archives/2006/02/the_solution_to.html

hc said...

Its a good post and generally I like your blog and will visit. I am also interested in environmental economics.

Because it only addresses terminating traffic and because people may search for spots and engage in drop offs it might have limited effect. It would be better to peak load price the spots so users get hit with a bigger chjarger if they leave or arise during the peak.

Where are you doing your Masters?

Beth said...

Good analysis, Davo, and I'm inclined to agree. Unfortunately Lord Mayor John So is using the raised revenue to run a free shuttle tram service which duplicates existing PT! Such a crazy idea/vanity project.

Rob Dawg said...

Overcharging for parking is an exact analog for minimum wage laws. Both are disincentives for correct market pricing. Parking or congestion charges ala central London ultimately "work" by displacing and/or reducing economic activity. Reducing physical access or introducing preferential access to some parking only serves to create another class structure amongst roads users. Reducing the amount of parking below demand only serves to increase congestion and reduce respect for the rule of law. Price elasticity is not the correct term as these levies are taxes not true costs. The rule of government applies; tax something, get less of it. Subsidize something get more of it. I guess my biggest complaint concerns what ever happened to government actually responding to societal chages rather than resisting them.

David Jeffery said...

Thanks Harry. I saw your presentation on traffic planning in Melbourne at last year's environment and economics network workshop at the ANU which certainly increased my interest in this area.

I'm doing my masters at UNSW - MComm in Environmental Economics.

Beth I agree that sounds like a wacky idea. That's also what they use the revenue for in Perth but it's probably more justified in Perth than Melbourne where the city already has good coverage.

Russ said...

David, good post. Melbourne, unlike Sydney, is not particularly congested near the CBD except at peak times. So, while reducing traffic in general is possibly worthwhile, measures to reduce through traffic or on-street parking may not affect congestion terribly, and are politically sensitive.

Robert, there is barely an aspect of transport that isn't an implicit or explicit market distortion subsidising one mode of travel or another. The levy is an attempt to try and reduce the biggest negative externality -- congestion in peak hour -- by specifically targetting long-term parking in the CBD. It is probably too small a charge to be effective, but it is not a bad way to do it.

Rob Dawg said...

Russ in a very intelligent reply remarks:
The levy is an attempt to try and reduce the biggest negative externality -- congestion in peak hour

Who says congestion is an externality? People caught in congestion pay in time and energy and personal stress at the time of incurring the congestion. Those are all direct internalized costs. That said, what about the peak hours makes them special? It seems to me that congestion charges and such are attempting to establish different classes of roads users. Preferential access to common resources is never going to achieve global best outcomes.

-- by specifically targetting long-term parking in the CBD.

This is the part that always mystifies me. What about making our pre-existing societal investment in CBDs less useful and more inconvienient is good policy? Driving people away only serves to encourage sprawl right? If cities had embraced the POV insted of "driving" them away we would have half the built environment wars we see now. "Hey, our mall is doing poorly. Let's cut the parking so it is tougher to do business here!"

David Jeffery said...

The externality is this Robert: If I choose to drive to work rather than walk tomorrow, I incur congestion-related costs. But I ALSO impose them on everyone else who's already driving, because there's one more car on the road. So my costs are not fully internalised at all.

It's really simplistic to suggest that congestion charging necessarily reduces economic activity. People who would like to do business in the CBD stay away because of the congestion costs. Access to the city has to be rationed in some way. The question is how do you do it? If you do nothing, as you suggest, you ration it through queues, which means that those whose time is least valuable use the CBD. If you ration it through some form of pricing, you should attract the people who really put an economic value on doing business in the CBD.

The fact is, Robert, you make a choice and set up 'classes' of user either way. You are just suggesting the class of user we should be favouring is those who have nothing better to do that sit in traffic for an hour or more. That's fine, but don't pretend it's natural or 'the market' at work - you're just making a value-laden choice like everyone else.

Russ said...

Robert, fair questions, which I think David answered well. In a sense you are right about the internalised costs. Following David's example, any product or service I might buy today will increase the demand and price, and impose costs on others. But time, unlike money, is not easily tradable, so it is better to have a market in the latter.

On whether people will be driven away, it depends on the value they place on doing business in the CBD and the alternative transport options that are available. If the money is put into other transport modes it will offset some or all of the imposed costs; depending on how efficient the alternatives are.

Rob Dawg said...

David, you cannot impose congestion on others merely driving yourself. It requires the other party to volluntarily participate as well. I do agree with the greater concept that there are externalities associated with inadequate roads capacity. Contrary to poular belief it is possible to build one's way out of congestion and new road capacity does not fill up with induced demand.

Russ, your suggestion to invest in transit is a common one but in the context of congestion and economic vitality transit is a drag on both not a solution. With the promotion of transit dependency you are forced to admit to the costs and pay for it. Then once you've raised taxes and bribed users with subsidies you are then faced with the immutable reality that tranist is slower. This not coincidentally is a lot of the reason you need to bribe people to use transit. The mechanics of transit introduce bias towards public policies that deliberately cause congestion; density and misallocated public investment for but two instances.

Any attempts to internalize the true costs of mobility would be fatal to CBDs so instead we see all these reward/punishment incentives being applied that are fully divorced from the enumerable criteria. Social engineering, nothing more. [N.B. recognise the large devil's advocate content above]

Russ said...

Robert, a few points:

Induced demand does exist, but it is a long term, not a short term phenomenon. Put simply, most people like to live around 30-60min from their workplace, so the effect of faster travel times (be that by car or by rail) is to induce travel from further away (where housing is cheaper and land more plentiful). On a heavily congested road there will be demand for travel that can't be justified by the current travel cost. Therefore, any improvement to that road will change the travel preferences for anybody who can justify that benefit under the new conditions (in relation to their other choices). Maybe the new road will congest, maybe it won't, it depends how strong the latent demand is.

However, regardless of how big you make a road, eventually it needs an exit and entry point, where most congestion takes place. This is the 'bottle-neck' problem, which I posted about on my own blog a few months ago. As often as not you are just shifting congestion elsewhere by building. As a long-term strategy it is not a good idea.

Regarding transit, see my earlier comment regarding subsidies in transport. A free road versus subsidised transport still favours the road user. Whether transit is faster or cheaper depends on the structure of the city and transport system. I don't necessarily advocate new transit systems, but we don't live in a true market, so any funding decisions are a matter of political preference. Moreover, for a fairly significant number of people, walking and cycling would be the most cost-efficient and fastest ways to travel, but people don't for other reasons (mostly crap urban design and a driving mindset).

I don't buy this statement: "Any attempts to internalize the true costs of mobility would be fatal to CBDs". Perhaps they would decline, perhaps they wouldn't. If there was a true market in transport (and infrastructure, and services, and...), and if the CBD did decline, it would probably be for good reasons. This may not be a bad thing.

Rob Dawg said...

"A free road versus subsidised transport still favours the road user."

Gosh, all we need is an example of a free road to test your hypothetical. Tell you what, you just fill your auto with tax exempt gas and refuse to pay for liciense and registration and insurance and make sure you place placards in your windows announcing those facts as you blow through any toll plazas. Get back to us when you get out of jail.

Don't make me post my "Pay only your operating costs day" experiment.

Russ said...

Done. And look, I am still here. Easy enough when you walk everywhere.

Lets not say 'free' then, but they are certainly subsidised. Not least, because petrol taxes create a pool of funding that is heavily misallocated away from where it is most needed in favour of other road projects. You could create the same funding model with public transport if you devoted that much public land to it (not a good idea, but still true).

I am not advocating either road building, nor public transport, nor any other method of transport as an appropriate solution. They all have their merits in different areas. But they are all politically driven, not market driven. And they are all subsidised in different ways.

We've probably exhausted this discussion. But thanks, it was interesting.

Rob Dawg said...

Russ, you didn't do anything close to what I asked. Instead you've just managed to reveal another form of transportation subsidy paid by roads users and given to another class, pedestrians. My point was to absolutely remove the myth of free roads from the discussion. Roads use is heavily taxed and regulated. We've only come part way to the truth when you admit "Lets not say 'free' then, but they are certainly subsidised." I'm sorry but I don't see the subsidy. Where does this subsidy come from and who benefits in what fashion? Maybe it just too hard to track down all those various funding streams. Maybe if you just gave a general figure, you know something like AU$0.01 per kilometre? Or is ten times as much or 10 times less? I think you'll be surprised just how trivial even the most agressive attempts to assign subsides ends up.

Russ said...

Robert, I wasn't going to continue this, however...

Firstly, footpaths, cycling paths and local roads are paid for by local councils through rates. The State government collects license and registration fees. The Federal collects petrol taxes. Both State and Federal fund major highways. State government subsidises public transport. To some extent money is fungible between different government levels, but not really. It is very questionable to say pedestrians are being funded by car drivers.

Secondly, the problem is not assessing funding streams, it is assigning costs. I have read (questionable) figures that show roads are heavily subsidised because they assign the costs of externalities to them (health costs, pollution etc.). I have also read the opposite. The biggest subsidy is the allocation of public space to roads over other uses. What is the opportunity cost of a length of car-only road over say, a car park, or a footpath, or a cycle path, or a set of cafe tables, or housing, or a public park? It will depend on the road, the surrounding uses, the efficiency of different transport methods and public preference for different methods. I don't have an answer. If you do I'd like to read it.

Thirdly, you asked me to stop paying road specific taxes. I don't pay them because I don't drive, my rates still get spent on roads and car parks however. This is the misallocation part. Some roads get very little traffic and are heavily subsidised, some get a lot and aren't. Roads benefit greatly from the fact that they already exist everywhere and operate a virtual monopoly. Funding paid through license registrations and petrol taxes cross-subsidises the building of a road even when that is not the most efficient use, further reinforcing automobile dependency. Very few people are in my position of not having to use a car and there is a substantial misallocation of resources. To draw an analogy, if the government runs a deficit does that mean taxes are too low? Even if roads are more than paying for themselves it doesn't mean we have too few of them.

To conclude, roads are subsidised by government fiat, the same way many television stations are (in Australia), and the airlines were. I can't put a figure on it. Nor would I necessarily trust anybody's figures who did. It is a ridiculously complex problem.

Rob Dawg said...

Firstly, footpaths, cycling paths and local roads are paid for by local councils through rates.

Is that what you think? Even if it were true these paths are a small percentage of all the roads and sidewalks that are "donated" by roads users fees.

(questionable) figures that show roads are heavily subsidised because they assign the costs of externalities to them (health costs, pollution etc.). I have also read the opposite.

It is good that you understand that those enumerated externalities are totally bogus. But you then go on to miss the truth about allocating "public" space to roads. Not only is it a very very small area in practice it is generally not "public" land. Yes for major freeways and such but all those tertiary, secondary and collectors in exurban developments are private property until "donated" to the general public. A "gift" of private land for public use is a subsidy how? And remember the value of the adjacent property and thus taxes goes way up far more than mantaining those roads costs. Again a tax not a subsidy.

Finally, you totally misunderstand the balance of payments with respect to who pays. You most certainly do pay for roads and you get a really good deal in the transaction. The prices you pay for the goods and services that use roads includes those costs. Interesting sidebar; when you shop at a "big box" store and arrive by walking the people in the parking lot are subsidizing your purchases.

Roads are not subsidized, thay are taxed. Government does not spend as much money on raods as it collects nevermind the money they save or the economic benefits they tax as well.

Aaron said...

Do car parking levies apply to ALL commercial spaces? What about the ones that are provided free (or as part of a salary package) to employees on the basement levels of CBD office buildings?

David Jeffery said...

Yep, that's exactly the kind of car spaces it's aimed at. The building owner pays the levy and then presumably passes it on to the tenant.