I'm guessing you don't think much about parking spaces except when you are searching for one. But indulge me by thinking about them a bit now.

For starters, most economic transactions you make depend in part on the ability of someone (you, a long-haul trucker, the UPS delivery person, an ambulance driver, and so on) to park. And stationary vehicles occupy vast amounts of land everywhere in the world. Indeed, a simple back-of-the-envelope calculation suggests that, in the United States, parking takes up more space than the whole state of Massachusetts. In Europe, where cars are smaller and fewer in number, parking still takes up an area half the size of Belgium.

Much of the recent interest among city planners and long-suffering urban drivers has focused on the potential for digital technology, market-driven pricing and wireless communications to reduce search times for parking. In San Francisco, a substantial portion of downtown parking meters respond directly to the availability of spaces by changing meter rates. The goal is to ensure there is always some parking available for drivers willing to pay enough and in the process reduce congestion by sharply cutting the number of drivers on the street who are searching for spots at any one time. But San Francisco's parking experiment is facing serious opposition, especially in residential neighborhoods yet to have meters. Many people, after all, resent the very notion of paying for parking.

One way or another, though, someone always pays, often indirectly, in the form of higher prices for something else. Moreover, the enormous amounts of land and structures needed for parking almost guarantee that mispricing parking spaces will have substantial consequences on economic efficiency and societal welfare. Here, I examine the implications of two kinds of parking in which costs are routinely paid indirectly.

Start with shopping-mall parking. Shoppers may think they don't pay for mall parking since, with relatively rare exceptions, nobody charges them directly. In fact, the cost of parking is reflected in store rents, and higher rents are reflected in the prices of the goods and services the stores sell. By the same token, city dwellers may think that curbside parking is free in front of their houses. But the value of the parking is capitalized in housing prices.

© David R. Frazier Photolibrary, Inc./Alamy

Mall Parking

According to a survey by the International Council of Shopping Centers and Urban Land Institute, a typical mall in the United States creates four to six parking spaces per 1,000 square feet of gross leasable area. And since those spaces use up more than 1,000 square feet, a typical mall allocates more footage for parking than for actual shopping. Yet, the same survey nonetheless found that parking is free at 94 percent of shopping malls.

This latter number may not surprise you because free parking has almost come to be seen as a right. But it was not so obvious to Kevin Hasker of Bilkent University in Turkey, who has been exploring the economics of parking with me because, at first blush, it would seem to make sense to charge parkers for the cost of the service rather than to find another pocket.

One can always come up with behavioral explanations for free mall parking, based on the assumption that shoppers find direct parking fees unacceptable the way people balk at paying for information from the Internet. Alternatively, we might hypothesize that, as long as one mall seeks to lure shoppers with free parking, competing malls must match them in order to attract patrons.

Both explanations have some bite to them, but both are also problematic. The first seems tied to the tautology that people don't like it because they don't like it. In any event, behavioral theories raise more questions than they answer. The analogy to free parking in the mall business is no separate table fees at restaurants. In fact, some do charge table fees. A cover charge at a bar is nothing but a fee for occupying space in the bar. And many pasticcerias in Italy charge higher prices to patrons who occupy tables than those who choose to gorge on zeppole at the counter.

The second explanation is no more satisfying. It begs the question of why inter-mall competition in this "two-sided market" – the mall is selling services to both shoppers and shops, which are interdependent – is reflected in store rents and prices of goods and services, rather than in parking fees.

Our own theory approaches the question very differently. One can think of shopping as participation in a lottery of sorts in which shoppers either win (find what they're looking for) or lose (don't find it). Shoppers who make the purchases leave the mall satisfied. But, of course, not all mall trips have happy endings: sometimes shoppers leave empty-handed. Since they "pay" for parking only by making purchases at prices that include an implicit charge for parking, those who buy nothing don't pay for parking. Thus, in a very real sense, bundling the cost of parking with the price of goods is a form of insurance for shoppers: if they don't buy, they pay nothing; if they do buy, they effectively pay for the parking of unsuccessful shoppers as well as their own.

Our insurance-based theory is surprisingly robust. First, experimental work in economics shows that almost all people are risk-averse even for lotteries over small items, implying they'd rather participate in the parking lottery, in which they either pay a lot for parking or nothing at all, than pay less upfront. Second, if a monopolist mall decides to charge nothing, the addition of a competing mall would hardly increase the mall's incentive to raise the fee above zero.

Consider, too, that free parking remains the rule in spite of the option of "validated" parking in which merchants pick up the parking costs only of those who do buy something. The aforementioned survey shows that 86 percent of the malls in the United States do not use validated parking, suggesting that mall managers and merchants mostly buy into the insurance approach by charging nothing to shoppers who leave empty-handed.

What Parking Fee Do We Want Malls to Charge?

The insurance approach explains why malls want to provide free parking. But what's good for the decision maker (the mall management) isn't necessarily good for society as a whole. When the price of one service (parking) is embedded in the price of another good (stuff sold at the mall), economists worry that inefficiencies will arise – that is, the marginal cost of providing the parking won't equal the marginal benefit to society as a whole. In this case, though, embedding parking costs seems to be good for society, too.

The reason is simple. Consider a thought experiment in which malls decide to generate more income to cover the costs of parking. Adding a direct parking fee distorts more than raising rents (and thus mall prices) because it would drive away both kinds of shoppers – those who will end up buying goods (the winners of the lottery) and those who will not (the losers of the lottery). By contrast, raising mall rents (and thus mall shop prices) tends to drive away only the former group.

Curbside Parking

The second example of free parking that isn't truly free is curbside parking. Although there has been a trend away from free to paid curbside parking almost everywhere as municipalities attempt to generate revenue and ration scarce space, there is still an enormous amount of free curbside parking available. But if you live in a place with free parking, don't be so sure you're the lucky one. Our research suggests that at least some of the value of that free parking is reflected in your rent or in the market value of your home. So, the issue is not whether you pay, but how you pay – directly or bundled in the cost of housing.

This will all be clearer if we first look at on-site parking that is sold as a bundle with housing. Most cities impose minimum parking requirements that determine how many parking spaces each new land use must include. In most American cities, developers must provide at least one and, in many cases, two spaces for each housing unit. This is no small deal: after taking into account the space allocated for ramps and maneuvering, the area occupied by two parking spaces is usually larger than a two-bedroom apartment.

My aunt recently bought an apartment that comes with a private parking space in the basement garage. But it is safe to say she'll never use it since she is 77 years old and has never driven a car. If the land-use parking requirement were dropped, the savings would be reflected in the competitive market price of housing. In fact, Michael Manville of Cornell University found that in San Francisco bundled parking increases the average asking price for an apartment by $22 per square foot.

Now back to curbside parking. When the parking space in front of your apartment building is free (and reliably available), you are, in essence, using that parking space as your own parking garage. Hence, one would suspect that at least some of the value of that parking space will be included in the price/rent of the apartment.

Istanbul's Crusade Against Free Curbside Parking

A natural experiment of sorts that took place in Istanbul gave us the opportunity to test the hypothesis. In the past, curbside parking was either free or operated by self-appointed "parking attendants." The city decided to end free/informal parking at the curb, and for that purpose created a parking company at the end of 2005 that took over the job of managing curbside parking spaces. The company expanded neighborhood by neighborhood, and thus there has been a gradual transition from free and informal parking to paid and formal. This allowed me and my co-authors Ozan Bakis (Sabanci University) and Rifat Ozan Senturk (University of Texas at Austin) to estimate the impact of unbundling curbside parking spaces on housing prices and rents.

The Three Effects

After the transition from free parking to paid parking, nearby residents are no longer able to use the curbside as their own private garages. This should decrease housing prices in those neighborhoods by unbundling. That is, residents used to pay upfront for parking that was effectively bundled with the house price. Now, they must pay separately, which should reduce the market value of the house. We call this the unbundling effect.

But, complicating matters, there are two countervailing effects. First, the transition from free to paid parking should reduce the demand for parking, which should mean less cruising for parking and thus less traffic congestion. This effect should make life more pleasant in the neighborhood, which in turn should lead to an appreciation of housing. We call this the reduced-cruising effect. One has to be careful here, though. Higher parking fees do not always reduce traffic congestion; in fact, they may increase congestion by increasing parking turnover – which would tend to reduce housing values.

The second countervailing effect is caused by the transition from informal parking to formal parking. In Istanbul, self-appointed parking attendants stood by the road and demanded money to "protect" parked cars. So, the market was really an informal market – one in which contracts between car owner and parking attendant could not be enforced. The parking company established by the city was held to a higher standard, increasing trust in the parking market and tending to increase housing prices. We call this the trust-enhancing effect.

The net impact on housing prices depends on which effects dominate, which one would expect to vary by city and perhaps by neighborhood. How rents react to the transition depends on whether landlords have market power (or renters do, thanks to government rent controls), which is also city-specific. If, for example, landlords can exercise market power, they might feel no obligation to reflect the depreciation in property values in lower rents or to limit rent increases to reflect increases in property values.

© Duane Duholke/Alamy

So, What Happened in Istanbul?

Before 2005 the city did not actively enforce the laws against informal parking attendants. This should not be terribly surprising in light of the reality that Turkey can't seem to prevent a lot of spontaneous appropriation of public property. A prime example is the theft of electricity, which may run as high as 70 percent of power generated in some cities.

The informal parking attendants had strong incentives to favor residents over non-residents. If a resident insisted on complaining to the authorities about "extortion" by a parking attendant, the attendant could easily get into trouble. So, it was wise for them to keep residents pleased. Often, the parking attendants were the janitors, who reserved parking spaces for the residents of the buildings in which they worked by putting barrels or rocks next to the curb. A non-resident, on the other hand, was virtually without legal recourse since he did not even know the name of the guy "selling" the curb space. So, the residents were really favored in pricing over non-residents.

How do the three effects play out in Istanbul? When we look at the city as a whole, we find that housing prices decreased by more than $6 per square foot in the neighborhoods where the city started operating curbside parking. This corresponds to a 9 percent decrease in housing prices in these neighborhoods, which is significant. By contrast, our estimates show that rents in these neighborhoods are not statistically different from those in other neighborhoods. This suggests that landlords have market power in the sense that competition did not drive down rents to compensate renters for the loss of their free parking spaces.

There is also some evidence that people were sufficiently mobile to seek cheaper parking. Car ownership rates increased in all neighborhoods across time throughout our data. However, the increase was greater in neighborhoods where the parking company entered later, and even more in neighborhoods where the parking company had yet to arrive.

Europe vs. Asia

Istanbul straddles the Asian and European sides of the Bosporus strait, and the housing market dynamics are quite different on each side. The Asian side is mostly residential and was developed later. Many buildings have on-site parking spaces sufficient to satisfy residents' demand. These spaces are mostly in the form of open surface lots located around the apartment buildings. Curbside parking is really for visitors. Thus, the ability to use curbside as one's own parking garage is not as important on the Asian side. The European side, on the other hand, has many districts with narrow streets and few attached garages. Thus, the unbundling effect is extremely important on this side of the Bosporus.

When we estimate the impact of the parking change for each side of Istanbul, we find opposite results. On the European side, the market price of housing decreased in the neighborhoods where the city took over curbside parking, but rents remained the same as in neighborhoods with free curbside parking. On the Asian side, however, we find that housing prices increased (by 4 percent) rather than decreased, while rents increased (by 6 percent) rather than remaining the same. Landlords there, it appears, have market power, as rents rose more than property values.

Why the increase in housing prices on the Asian side? Because the unbundling effect is less important on the Asian side (remember that the Asian-side residents usually have their own on-site parking spaces), while the reduced-cruising and trust-enhancing effects combined outweigh it.

The overall analysis shows clearly that at least some of the costs of curbside parking are embedded in housing prices, although these parking spaces are not formally bundled with the housing units. So, we get the same result again: The cost of parking is embedded in the price of other things. As a matter of fact, the impact on housing prices and rents is large enough to significantly change the real income of property owners and tenants. On the European side of Istanbul, the transition made housing more affordable while it made housing more expensive on the Asian side.

Takeaway

A parking space is a temporary home for your car, but a permanent commitment of land. When unpriced, its costs do not go away; rather they are hidden in the price of everything else. If you live in an urban area, you may think that finding a parking space is difficult, and more often than not, it is. However, this is only the tip of the iceberg. A parking space has more far-reaching effects on your welfare than you may assume.

Here, I've offered two tales of free parking leading to different endings. Shopping-mall parking may appear to be free, but in fact you pay for it every time you buy something at the mall. Happily, though, what looks like distorted pricing serves the broader interests of society as well as those of the mall owners. Curbside parking in front of your house may also appear to be free, but in fact its costs are already capitalized in housing prices and rents. Although I can't claim the last word on the subject, our estimate hints that free curbside parking produces negative welfare consequences.