Parking Market Failures

Does any market failure argument justify parking minima? Many cities require that residential developments include some minimum number of off-street carparks. Can these be justified?

Well, I'm glad you asked, Agnito.

The best argument I've heard for these regulations is that on-street parking is rivalrous and non-excludable. If you have an apartment building with too few carparks, residents will park on the street or might keep flitting between a few different no-parking zones, staying one step ahead of the enforcement officers. A lot of people see paying for parking a bit like hiring other kinds of personal services: why pay for it when you can apply yourself and perhaps get it for free? And if you find the perfect parking spot, there's then a huge opportunity cost of leaving the space: people with really good parking spaces never give them up. Finding the perfect space can be a joy to be shared with others. Further, where excess congestion from everyone's free-riding on on-street parking gives incentive for the able-bodied to park in handicapped spaces (or in front of fire hydrants), disastrous consequences can ensue.

Having mandatory parking minima then avoids the free-riding on the on-street parking that could best be left for higher-turnover use. And where long-term on-street parking is discouraged only by time limits, it also encourages other socially wasteful forms of entrepreneurship.

The best response to this kind of argument is that it's massively second-best relative to the more efficient solution: price on-street parking. Do that properly and everything else sorts itself out.

Of course, in that world, Seinfeld wouldn't have been nearly as good. And where Gerry Brownlee wants to ban Auckland from using congestion charging, we might not be able to get first-best parking charges anyway.

For the record though: there is no real-world market failure sufficiently large to justify mandatory parking minimums. At least not in Auckland, best I can tell.

Bland by design

I could grok changed building rules in Christchurch post-quake. Earthquake and liquifaction changed what we might want from foundations.

But the percentage of building frontage that must be in windows, no matter what? That car parking be hidden?

For example, the rules requiring buildings facing a road or public space to be between 60 and 90 per cent windows would not suit many businesses.

"Such blanket provisioning ignores that such a percentage of glazing may be inappropriate for the retailer, who may need more security, such as a jeweller or a bank; whose security requirements must take a higher priority than urban design,'' the submission said.

"Furthermore, it ignores the needs of department stores or larger stores who may need to place stock on shelving attached to solid walls around the perimeter of building."

The Property Council said that, because glazing was so expensive, the rule would boost the cost of new shops.

"We are strongly opposed to any provisions in the plan change that call for an increase in development costs without sufficient justification."

The submission criticised the requirement for car parking to be hidden from view, saying visible parking was a principle marketing attraction for retailers.

"This is completely impractical for many retail activities, which rely on visible parking to attract sufficient customers in order to remain viable,'' it said.

Prescriptive zoning rules are what deliver boring, expensive cities. Get a long enough list of "every building must", and you'll get a pretty short menu of options that can fit the bill.

I just don't get why we have to be so prescriptive about things that are orthogonal to "risk this building causes to others that are avoidable at reasonable cost." Minimum engineering standards that keep buildings from falling onto passers-by make sense. Council failed to do anything about this prior to the quakes, and even hindered owners who had wanted to tear down buildings that wound up falling down and killing people in February's quake. Even if we take a hard econ line on that individuals should be free to live or work in a dodgy building and trade safety for money, risk imposed on passers by seem sufficient to require either strictly enforced minimum standards or liability rules with compulsory insurance.

I wonder how much intersection there is between the kinds of people who think prescriptive town planning rules are great things and the kinds of people who don't like the tilt-slab construction that's been the consequence of trying to tick all the planning boxes on a budget.

Meanwhile, in America, Matt Yglesias has taken up Donald Schoup's banner on the high cost of free parking. Parking minimums are pretty common in the States: developers then have to put in more parking than they'd like to. Other places have parking maxima, preventing developers from providing as much parking as they think appropriate.

What happens when you stop being so prescriptive around parking?

Michael Manville of UCLA studied a liberalization of parking regulations in one section of Los Angeles and found that deregulation leads to the construction of more housing units and fewer parking spaces. Conversely, tighter regulation leads to a lack of affordable housing and a surplus of parking spaces. That might make sense if parking spaces were a public good, like clean air. But they’re closer to being a public bad. When Chicago mandates the creation of ahigh number of parking spaces per square foot of downtown office building, it reduces the price of parking, but it has a number of negative consequences. Cheaper parking means more traffic congestion on the streets. It also means lower ridership for Chicago mass transit. Perversely, cheaper parking offers a subsidy to commuters from outside the city limits at the expense of Chicago residents living within walking or biking distance of the central business district. And, of course, it leads to dirtier air, not cleaner.

Yglesias recommends abolishing requirements that buildings have parking spaces; I'll also recommend abolishing requirements that they have maximum numbers of parking spaces. If the highest valued use of a piece of land, as seen by the person with skin in the game, is a parking space, why need Council get involved?

Maybe, just maybe, if Christchurch Council focused really hard on a small number of rules around building safety, and dropped the other stuff, they'd be able to competently administer a set of useful rules instead of, well, what we have instead.

Increasing consumer surplus through price increases

As I noted last year, the University of Canterbury administration has this year increased the price of an annual parking permit threefold from (roughly) $100 to $300. This raises the price from what was a subsidised rate to something they calculate as being approximately marginal cost. Needless to say, this is something that the Economics department had been advising for a long time, given our propensity to value efficiency even at the expense of our own direct wellbeing. After a few months of experience with the new policy, it has become clear, though, that it is not only efficiency enhancing, but it has also increased consumer surplus even without consideration of what use the university makes of the increased revenue.

How can that be? It is an application of how the deadweight loss triangle in a standard S&D diagram understates the cost of a price floor or ceiling. Previously a parking permit at Canterbury did not confer a right to park; it conferred the right to hunt for a park. Many of us wasted a lot of time searching for a park before giving up and parking on the street several blocks away. The problem was particularly acute on wet days. Some of those who successfully found parks had a low willingness to pay, others who missed out valued the parks much more highly. How do we know this? Well this year, as a result of a trivial price change from next-to-nothing to three times next-to-nothing, the carparks are never full.* Even on the wettest days, one can come in late and always be guaranteed a park. Those cluttering up the parks last year but not this clearly didn’t value the parks highly; this year, it is only those put a high value on parking who get the parks. And how high can that value be. Well we don’t know for sure, but I am sure this story could be replicated here.

So there we have it. The price went up, and so did consumer surplus. Could the same happen in reverse. Well imagine if you were to impose average cost pricing in the retail electricity market despite it being an industry with sharply increasing marginal cost. Everyone would get a lower price for power, but with no guarantee that the lights would come on on demand. Consumer surplus might well go down. And that is without even considering the lost government revenue from publicly owned power companies….

* I find it difficult to comprehend the size of the demand response; think of the Slutzky equation: there is a huge shortage of on-street parking around the university, so there are no close substitutes for on-campus parking; $300/annum is hardly a large fraction of anyone’s expenditure, student or lecturer. Can the income elasticity really be that high? 

Popcorn, Parking, Exercise and the Theory of the Second Best

The University of Canterbury recently announced that it will be charging faculty and students alike something close to marginal cost for parking. The stated reason was that they saw no reason why scarce resources should be allocated to providing an amenity benefit that is of value to some staff and students but not to others.

The economists, of course, cheered. A parking permit at Canterbury is currently very cheap, but it is only a licence to hunt, not a guarantee of a park. We have been saying for years that parking should be charged at a market-clearing price. After all, charging parking at less than marginal cost generates a deadweight loss in which the marginal units consumed are valued at less than cost, and so reduces the total value that the university can deliver to staff and students. If there was a reason for thinking that the UoC-specific elasticities of demand of students or of labour supply of staff are higher for those who want parking than those who do not, you could make a Landsburg popcorn price-discrimination argument for subsidising parking, but that assumes that the University is a profit-maximising monopolist, rather than a not-for-profit utility maximising institution.

Then today, I heard from students that in another move, the Univeristy is increasing the student levy next year but including free gym membership as an offset. My first reaction was that this is the exact opposite of the move to increase the price of parking, and is silly for exactly the reasons that the former policy makes sense. But then it became clear. The students explained that a student levy can be added on to a student's interest-free student loan, whereas the cost of gym membership as an optional extra cannot. Given that every dollar of student loan is roughly equivalent to a straight out grant from the government of 33c if it is paid off at the slowest possible rate (from memory of a back-of-the-envelope calculation I made some time ago, so don't take this number to seriously), the new policy might make sense as a way of increasing the total government contribution to the University, as long as that benefit outweights the effect of the created deadweight loss, including, possibly, the cost to some students of having to queue for gym services.

But then I reaslied, we have to pay for parking permits out of after-tax income, but the University's expenditure on parking net of parking fees is not subject to fringe-benefit tax. So subsidised parking is also a way of increasing the government's total contribution to the University......   Isn't second-best analysis fun!