Note: this is the HTML (Hypertext Markup Language) version of this document. It does not have most page formatting features of the original document, such as citations and notes, and tables may be difficult to read. Use Word or PDF formats for more accurate and complete documents.

 

The Trouble With Minimum Parking Requirements

 

Donald C. Shoup

Department of Urban Planning

School of Public Policy and Social Research

University of California, Los Angeles (UCLA)

Los Angeles, California 90095-1656

shoup@ucla.edu

 

 

 

9 December, 1999

 

 

Originally Published in

Transportation Research Part A

Vol. 33 (1999), pp. 549-574

 

Posted at the Victoria Transport Policy Institute website with author’s permission.

 

Abstract

Urban planners typically set the minimum parking requirements for every land use to satisfy the peak demand for free parking. As a result, parking is free for 99 percent of automobile trips in the United States. Minimum parking requirements increase the supply and reduce the price–but not the cost–of parking. They bundle the cost of parking spaces into the cost of development, and thereby increase the prices of all the goods and services sold at the sites that offer free parking. Cars have many external costs, but the external cost of parking in cities may be greater than all the other external costs combined. To prevent spillover, cities could price on-street parking rather than require off-street parking. Compared with minimum parking requirements, market prices can allocate parking spaces fairly and efficiently.

 

How can a conceptual scheme that one generation admiringly describes as subtle, flexible, and complex become for a later generation merely obscure, ambiguous, and cumbersome?

Thomas Kuhn

 

 

Urban planners set minimum parking requirements for every land use. These requirements typically ensure that developers will provide enough spaces to satisfy the peak demand for free parking. This article examines (1) how urban planners set parking requirements, (2) how much the required parking costs, and (3) how parking requirements distort the markets for transportation and land. As a way to eliminate this distortion, I will propose that cities should price on-street parking rather than require off-street parking.

 

 

The Shaky Foundation of Minimum Parking Requirements

Where do minimum parking requirements come from? No one knows. The "bible" of land use planning, F. Stuart Chapin’s Urban Land Use Planning, does not mention parking requirements in any of its four editions. The leading textbooks on urban transportation planning also do not mention parking requirements. This silence suggests that planning academics have not seriously considered–or even noticed–the topic.

 

This academic neglect has not prevented practicing planners from setting parking requirements for every conceivable land use. Figure 1 shows a small selection of the myriad land uses for which planners have set specific parking requirements. Without training or research, urban planners know exactly how many parking spaces to require for bingo parlors, junkyards, pet cemeteries, rifle ranges, slaughterhouses, and every other land use.

 

Figure 1 Selected Land Uses With Minimum Parking Requirements

Asylum

Indoor Soccer Facility

Rifle Range

Bingo Parlor

Junkyard

Slaughterhouse

Convent

Kennel

Taxi Stand

Diet Clinic

Landfill

Ultra-Light Flight Park

Exterminator

Massage Parlor

Veterinarian

Fraternity

Night Club

Wastewater Treatment

Gunsmith

Oil Change Shop

Zoo

Horse Stable

Pet Cemetery

 

Source: Selected from the minimum parking requirements for 179 land uses in Planning Advisory Service (1991, 3)

 

 

Richard Willson (1996) surveyed planning directors in 144 cities to learn how they set parking requirements. The two most frequently cited methods were "survey nearby cities" and "consult Institute of Transportation Engineers (ITE) handbooks." Both strategies cause serious problems.

 

Survey Nearby Cities

Although surveying nearby cities seems a sensible way to set parking requirements, the Planning Advisory Service (1971, 1-3) explains a serious problem with this approach:

Since the establishment of the principle that zoning ordinances may legally require the provision of off-street parking, ordinance drafters have been asking questions like: "How many spaces should be provided for a drive-in restaurant?"–or any other land use for that matter. The question is typically answered by relying upon what ordinances for other jurisdictions require… The implicit assumption is that other areas must know what they are doing (the ordinances were adopted, after all) and so it is a relatively safe bet to adopt a parking standard "close to the average." This may simply result in a repetition of someone else's mistakes. Nevertheless, the planner who needs to present a numerical standard by the next planning commission meeting can't answer the original question by saying, "I don't really know." (italics added)

 

Setting parking requirements by relying on what other cities require not only risks repeating someone else’s mistakes, but also fails to reveal where the requirements came from in the first place.

 

Consult ITE Handbooks

To base parking requirements on more objective data, planners consult Parking Generation, published by the Institute of Transportation Engineers. For each land use, this publication reports the "parking generation rate," defined as the peak parking occupancy observed in surveys by transportation engineers.

A vast majority of the data . . . is derived from suburban developments with little or no significant transit ridership… The ideal site for obtaining reliable parking generation data would . . . contain ample, convenient parking facilities for the exclusive use of the traffic generated by the site… The objective of the survey is to count the number of vehicles parked at the time of peak parking demand (ITE 1987a, vii-xv, italics added).

 

 

The ITE summarizes the survey results and reports the average peak parking occupancy observed at each land use as the parking generation rate for that land use. Half of the 101 reported parking generation rates are based on four or fewer surveys of parking occupancy, and 22 percent of the parking generation rates are based on a single survey.

 

Because parking is free for 99 percent of all automobile trips in the United States, parking must be free at most of the ITE survey sites. Parking generation rates therefore typically measure the peak demand for parking observed in a few surveys conducted at suburban sites that offer ample free parking and lack public transit. Urban planners who use these parking generation rates to set minimum parking requirements are making a big mistake.

 

 

Parking Generation is a questionable resource for several reasons. First, parking generation rates are inflated by the ample free parking. Second, no information is provided on several key issues. Why and where were the surveys were conducted? How long did the surveys last? How long did the peak parking occupancy last? Finally, nothing is said about off-peak parking occupancy. Parking Generation raises more questions than it answers.

 

Figure 2 shows Parking Generation’s report for one land use, fast-food restaurants. At the 18 survey sites parking generation ranges from 3.55 to 15.92 parking spaces per 1,000 square feet of floor area. The R2 of 0.038 shows that the variation in floor area accounts for less than 4 percent of the variation in peak parking occupancy. Parking generation is essentially unrelated to floor area in the sample. Nevertheless, the average parking generation rate–normally interpreted as the relationship between parking demand and floor area for a land use–is reported as precisely 9.95 parking spaces per 1,000 square feet of floor area.

 

Figure 2 Parking Generation at Fast Food Restaurants with Drive-In Windows

Source: Institute of Transportation Engineers (1987a., p. 146)

 

When urban planners consult ITE publications they behave like frightened natives before a powerful totem. For example, the median parking requirement for fast-food restaurants in the United States is 10 spaces per 1,000 square feet of floor area, the same as the ITE’s average parking generation rate. Beyond the ITE’s impressive professional reputation, the ITE data appeal to urban planners because minimum parking requirements are intended to meet the peak parking demand, and no one else provides systematic data that relate peak parking demand to land use.

 

 

Minimum Parking Requirements Inflate Trip Generation Rates

How do minimum parking requirements affect the demand for vehicle trips? The ITE publishes Trip Generation to show the demand for vehicle trips associated with various land uses. For each land use, this publication reports the "trip generation rate," defined as the number of vehicle trips that begin or end at a land use during a given period. In choosing a survey site the ITE (1987b, 23) recommends, "the site should be self-contained with adequate parking not shared by other activities."

 

Half of the 1,533 reported trip generation rates are based on four or fewer surveys, and 26 percent of the trip generation rates are based on a single survey. As with Parking Generation, the survey sites probably offer free parking. The trip generation rates therefore typically measure the number of automobile trips observed in a few surveys conducted at sites with free parking. Free parking inflates the trip generation rates because vehicle trip demand is higher where the price of parking is lower.

 

Figure 3 shows Trip Generation’s report for fast-food restaurants. It shows the total number of vehicle trips to and from each survey site during a 24-hour period from Monday through Friday. Trip generation ranges from 284 to 1,359.5 vehicle trips per day per 1,000 square feet of floor area among the eight survey sites. The R2 of 0.069 shows that the variation in floor area accounts for less than 7 percent of the variation in vehicle trips. Trip generation is essentially unrelated to floor area in the sample. Nevertheless, the average trip generation rate–normally interpreted as the relationship between vehicle trips and floor area for a land use–is reported as precisely 632.125 vehicle trips per day per 1,000 square feet of floor area.

 

 

Figure 3 Trip Generation At Fast Food Restaurants With Drive-Through Windows

Source: Institute of Transportation Engineers (1987a., p. 1119)

 

 

Parking Generation Compared with Trip Generation

To test the reliability of parking and trip generation rates, we can compare the number of vehicle trips per day to fast-food restaurants with the peak parking demand at fast-food restaurants. The number of daily round trips to a site divided by the number of parking spaces at the site can be interpreted as the parking turnover rate, which is the number of different cars that occupy a parking space during the day. Table 1 shows both the trip generation rates (expressed in round trips, or half the number of trip ends) and parking generation rates per 1,000 square feet of floor area for all the land uses that are common between the Trip Generation and Parking Generation editions published in 1987 (the most recent edition of Parking Generation).

 

The final column of Table 1 shows the parking turnover rate. For example, on an average weekday a fast-food restaurant generates 316.1 vehicle-round-trips and a peak parking occupancy of 10 spaces per 1,000 square feet of floor area. Therefore, 32 different cars occupy each parking space during an average day (316.1 ¸ 10).

 

Table 1 Trip Generation Rates Compared With Parking Generation Rates

 

Land Use

Trip Generation

(round trips/day)

Parking Generation (parking spaces)

Trips Per Parking Space

(round trips/space/day)

Manufacturing

1.9

1.6

1.2

Furniture store

2.2

1.2

1.8

Industrial park

3.5

1.5

2.4

Residential Condominium

2.9

1.1

2.6

Quality restaurant

47.8

12.5

3.8

Warehousing

2.4

0.5

4.9

Motel

5.1

0.9

5.7

Retirement community

1.7

0.3

6.1

Church

3.8

0.4

9.0

Government office

34.5

3.8

9.0

Discount store

35.6

3.6

10.0

Hardware Store

25.6

2.4

10.6

Supermarket

62.8

2.9

21.9

Tennis courts/club

16.5

0.7

23.2

Fast food w/ drive-thru

316.1

10.0

31.6

Fast food w/o drive-thru

388.6

11.7

33.3

Bank w/ drive-thru

145.6

4.2

34.4

Bank w/o drive-thru

95.0

0.6

150.8

Convenience market

443.5

1.4

314.6

Per 1000 Square Feet Sources: Institute of Transportation Engineers (1987a, b)

 

 

The parking turnover rate at furniture stores is only 1.8 cars per parking space per day, implying slow business. At churches it is a busy nine cars per space per day, heralding a religious awakening. At government office buildings it is also nine cars per space per day, suggesting that the state has not withered away. At tennis courts it is 23.2 cars per space per day, implying very short games but many of them.

 

These turnover rates are unreliable because the underlying parking and trip generation rates are often based on scant evidence (the parking or trip generation rate is based on only one survey for 4 of the 19 land uses). The surveys of parking generation for each land use were probably conducted at different sites and at different times from the surveys of trip generation. These bizarre turnover rates also suggest a more serious problem: the parking and trip generation rates are misleading guides to transportation and land use planning.

 

 

The Tail That Wags Two Dogs

Free parking is an unstated assumption behind both parking generation rates and minimum parking requirements. Transportation engineers do not consider the price of parking as a variable in estimating parking generation rates. When urban planners set parking requirements they make the same mistake. Urban planners interpret the ITE parking generation rates as the demand for parking, neglecting the fact that demand has been observed only where parking is free. The following five steps describe the dysfunctional interaction between transportation engineers and urban planners.

  1. Transportation engineers survey parking occupancy at sites that offer ample free parking and lack public transit. The ITE summarizes the peak parking occupancies observed at each land use and reports the parking generation rate.
  2. Urban planners use the parking generation rates to set minimum parking requirements for all land uses. Because the required parking supply is so large, the market price of parking is zero, and most new developments offer free parking.
  3. Transportation engineers survey vehicle trips to and from sites that offer free parking. The ITE summarizes the data on vehicle trips observed at each land use and reports the trip generation rate.
  4. Transportation planners design the roads and highways to satisfy the trip generation rates. Therefore, the transportation system provides enough capacity to satisfy the expected demand for vehicle trips to and from land uses that provide free parking.
  5. Urban planners limit land use density so that new development will not generate more vehicle trips than nearby roads and highways can carry.

 

 

In this five-step process, the unstated assumption of free parking underpins planning for both transportation and land use. Peak parking occupancy observed at sites that offer free parking becomes the minimum number of parking spaces that all development must provide. Ubiquitous free parking then stimulates the demand for vehicle travel. The observed travel demand becomes the guide for designing the transportation system that brings cars to the free parking. Planners limit development density to prevent traffic congestion around the sites that offer free parking. Because of this circular reasoning, free parking is the tail that wags two dogs–transportation and land use.

 

 

The Cost of Complying With Minimum Parking Requirements

Theory and data play small roles in setting parking requirements, and so we should not be surprised that the requirements often look foolish. This foolishness is a serious problem because minimum parking requirements increase development cost and they powerfully shape land use, transportation, and urban form. While urban planners rarely consider the cost of parking requirements, developers rarely have the luxury of not considering this cost.

 

The Cost of Parking Spaces

What does it cost a developer to comply with minimum parking requirements? We can estimate this cost by taking into account the number of required parking spaces and the cost per space. The Appendix presents evidence that aboveground structured parking often costs about $10,000 per space and that underground parking often costs about $25,000 per space. The most common parking requirement for an office building is four spaces per 1,000 square feet of floor area. If aboveground parking costs $10,000 per space, the cost of providing the required parking is $40 per square foot of floor area (4 x $10,000 ¸ 1,000). If underground parking costs $25,000 per space, the cost of the required parking is $100 per square foot of floor area (4 x $25,000 ¸ 1,000).

 

In Los Angeles the average construction cost of an office building, excluding the cost of parking, is about $150 per square foot. Therefore, in this example, the cost of four aboveground parking spaces per 1,000 square feet of office space increases the cost of the office space by 27 percent ($40 ¸ $150). The cost of four underground parking spaces per 1,000 square feet of office space increases the cost of the office space by 67 percent ($100 ¸ $150).

 

Because motorists park free for most vehicle trips, they clearly do not pay the cost of providing parking spaces. If motorists do not pay for parking spaces, who does? Minimum parking requirements bundle the cost of parking spaces into the cost of development, and thereby increase the cost of all the goods and services sold at the sites that offer free parking. These requirements "externalize" the cost of parking, so that you cannot reduce what you pay for parking by consuming less of it. Minimum parking requirements bypass the price system in the markets for both transportation and land.

 

The Cost of Parking Spaces Compared with the Cost of Cars

Minimum parking requirements increase the supply and reduce the price–but not the cost– of parking. To reveal the size of the resulting subsidy for parking, we can compare the value of parking and cars with what motorists pay for parking and cars.

 

Table 2 shows the number of registered vehicles and the capital value (in current dollars) of these vehicles for the years 1985 to 1995. For example, 202 million vehicles were registered in 1995, and this stock of vehicles was valued at $1,079 billion, or $5,352 per vehicle. How does this value of vehicles compare with the value of parking spaces?

 

 

Table 2 The Value of Motor Vehicles in the United States

Year

Registered Vehicles

Capital Value of Vehicles

 

(million)

Total (US$ billion)

Per Vehicle (US$/vehicle)

1985

172

614

3575

1986

176

688

3918

1987

179

731

4085

1988

184

790

4286

1989

187

833

4446

1990

189

868

4595

1991

188

879

4673

1992

190

910

4778

1993

194

961

4952

1994

198

1032

5211

1995

202

1079

5352

Sources: Katz and Herman (1997) for capital values and Federal Highway Administration (1995) for number of vehicles. Values are expressed in current dollars of each year.

 

 

Minimum parking requirements are intended to satisfy the expected peak demand for parking at every land use–at home, work, school, banks, restaurants, shopping centers, movie theaters, and hundreds of other land uses from airports to zoos. Because the peak parking demands at different land uses occur at different times of the day or week, and may last for only a short time, several off-street parking spaces must be available for every motor vehicle. Although no one knows the number of parking spaces per car, Victor Gruen (1973) estimated that for every car there must be at least one parking space at the place of residence and three to four spaces elsewhere.

 

Suppose there are four parking spaces per vehicle. If the average vehicle is worth $5,352 and if there are four parking spaces per vehicle, the average vehicle value per parking space is $1,338 ($5,352 ¸ 4). Therefore, if the average land-and-improvement value of a parking space exceeds $1,338, the average value of four parking spaces exceeds the average $5,352 value per vehicle they serve. Because $1,338 is a very modest sum for both the land and construction cost of a parking space, the total value of all parking spaces probably exceeds the total value of all vehicles.

 

Motorists pay for their vehicles (worth $1.1 trillion in 1995) but they park free for 99 percent of automobile trips. Motorists pay so little for parking because parking requirements bundle the cost of parking into the cost of development. Parking is free for most automobile trips only because its cost has been shifted in to higher prices for everything else. Everyone pays for parking whether they use it or not. Cars have many external costs, but the cost of parking in cities may be far greater than all these other external costs combined. By hiding a huge share of the cost of owning and using cars in cities, minimum parking requirements intensify all the other problems of external cost (such as air pollution and traffic congestion), making an already bad situation far worse.

 

Minimum parking requirements distort transportation and land use. They are not, however, the first example of an unwise professional practice that has produced unintended consequences. A medical analogy illustrates the problem.

 

 

An Analogy: Lead Poisoning

Parking requirements in urban planning resemble lead therapy in medicine. Lead has antiseptic properties because it is toxic to microorganisms, and until the twentieth century physicians prescribed lead to treat many ailments. One popular medical treatise recommended the using lead as a therapy for abscesses, burns, cancer, contusions, gout, gunshot wounds, inflammation, itch, piles, rheumatism, ruptures, sprains, stiffness of the joints, and ulcers.

 

Early physicians did not realize that lead is toxic to humans, and lead poisoning went largely unnoticed as a medical problem until the end of the nineteenth century. Nevertheless, a few early critics had recognized lead’s harmful effects. As a printer, Benjamin Franklin had much contact with lead, and he wrote to a friend in 1786,

The Opinion of this mischievous effect from lead is at least above sixty years old; and you will observe with Concern how long a useful Truth may be known and exist, before it is generally receiv’d and practis’d on.

 

Lead continued to be used as medicine for more than a century after Franklin’s warning, and folk remedies continue to use it as an ingredient today. Lead has local antiseptic properties, but any local benefit comes at a high price to the whole person. Lead exposure occurred in many ways unrelated to medicine when no one suspected that lead was harmful.

 

Minimum Parking Requirements: Urban Lead Therapy

Like lead therapy, minimum parking requirements produce a local benefit–they ensure that every land use can accommodate all the cars "drawn to the site." But this local benefit comes at a high price to the whole city. Minimum parking requirements increase the density of both parking spaces and cars. More cars create more traffic congestion, which in turn provokes calls for more local remedies, such as street widening, intersection flaring, intelligent highways, and higher parking requirements. More cars also produce more exhaust emissions. Like lead therapy, minimum parking requirements produce a local benefit but damage the whole system.

 

Minimum parking requirements resemble other primitive medical practices that were adopted without good theory and careful empirical research. Describing a leading medical text written in 1896, Lewis Thomas (1981, 40) says,

The public expectation then, as now, was that the doctor would do something. There was no disease for which a treatment was not recommended…. Every other page contains a new, complex treatment always recommended with the admonition that the procedure be learned by rote (since it rarely made any intrinsic sense) and be performed precisely as described. Acute poliomyelitis had to be treated by subcutaneous injections of strychnine; the application of leeches; the administration of belladonna, extract of ergot, potassium of iodide, and purgative doses of mercury; the layering of thick ointments containing mercury and iodine over the affected limbs; faradic stimulation of the muscles; ice-cold shower baths over the spine; and cupping …each of these with a dosage schedule to be followed precisely, some of them singly, others in various combinations… All of this has the appearance of institutionalized folly, the piecing together of a huge structure of nonsensical and dangerous therapy, and indeed it was. The pieces were thought up and put together almost like thin air, but perhaps not quite. Empiricism made a small contribution, just enough in the case of each to launch it into fashion.

 

 

I suspect that, looking backward a century from now, urban planners will see minimum parking requirements to have been no better than physicians now see lead therapy: a poison prescribed as a cure. Like many discredited and abandoned medical practices, minimum parking requirements are "institutionalized folly."

 

Many parking spaces are provided voluntarily rather than in response to requirements. And far from being a poison, parking is an indispensable part of the transportation system. What is poisonous, however, is for planners to require massive overdoses of parking.

 

Sometimes a disaster must occur to stimulate the reexamination of customary practices. Minimum parking requirements have produced no single disaster, but evidence of their harm confronts us everywhere–traffic congestion, air pollution, energy imports, the orientation of the built environment around the automobile, perhaps even global climate change. Although not their sole cause, minimum parking requirements magnify all these problems.

 

Likening parking requirements to lead poisoning is a criticism of current planning practice, not of individual planners. Physicians who prescribed lead were making an honest mistake. Urban planners who prescribe parking requirements are, I believe, also making an honest mistake. Although many planners may agree with this criticism, they may also feel that it is unhelpful unless the critic can propose a better way to deal with the parking problem. I will propose an alternative: cities should price on-street parking rather than require off-street parking.

 

 

 

An Alternative: Let Prices Do The Planning

Minimum parking requirements are a mistake but they respond to a real problem–spillover parking. If a land use does not provide enough off-street parking, some motorists drawn to the site will park on nearby streets, competing for the scarce curb parking supply. Urban planners know that this spillover parking creates enormous political problems. If spillover parking from a new development congests the adjacent curb parking, everyone nearby will angrily ask planners and politicians, "How could you let this happen?"

 

To prevent parking spillover where adjacent curb parking is free, new land uses must provide enough off-street spaces to satisfy the demand for free parking. Free curb parking explains why planners consciously or unconsciously base off-street parking requirements on the demand for free parking. In his survey of planning directors in 144 cities, Richard Willson (1996) asked "Why does your city have minimum parking requirements?" The most frequent response was the circular explanation "to have an adequate number of spaces." In effect, planners treat free parking as an entitlement, and they consider the resulting demand for free parking to be a "need" they can measure.

 

Because parking requirements are so ingrained in planning practice, complaining about them may seem futile, like complaining about photosynthesis or gravity. If free parking were an entitlement and the goal is to prevent parking spillover, requiring enough off-street parking to meet the demand at zero price would make sense. But free parking is not an entitlement. As the alternative to requiring off-street parking, consider pricing curb parking.

 

The Market Price for Curb Parking

The market price for curb parking is the price that matches demand with supply and keeps a few spaces vacant. Traffic engineers usually recommend a vacancy rate for curb parking of at least 15 percent to ensure easy parking access and egress. If cities priced curb parking to balance supply and demand with a few vacant spaces on every block, motorists could always find a convenient parking space close to their final destination.

 

 

Figure 4 The Market Price of Curb Parking

 

 

Figure 4 illustrates the policy of market prices for curb parking. Because the supply of curb spaces is fixed, the supply of curb spaces available with a 15 percent vacancy rate is a vertical line positioned above the horizontal axis at an 85 percent occupancy rate. The demand curve slopes downward, and the market-clearing price of parking occurs where the demand curve intersects the vertical supply curve. For example, when parking demand is high (demand curve D1), the price that will yield a 15-percent vacancy rate is high (P1 is 60¢ an hour). When demand is lower (demand curve D2), a price of only 20¢ an hour will yield a 15-percent vacancy rate. When parking demand is lowest (demand curve D3), the vacancy rate will be 50 percent even when parking is free.

 

If the price of parking is set too high, many parking spaces remain vacant, and a valuable resource is unused. If the price of parking is set too low, the occupancy rate reaches 100 percent, and motorists hunting for a vacant space waste time, congest traffic, and pollute the air. Because the demand for parking rises and falls during the day but the supply of parking is fixed, demand-responsive parking prices would necessarily rise and fall to maintain an "inventory" of vacant parking spaces on each block. The lowest price that will yield a vacancy rate of about 15 percent is the market price of curb parking.

 

Obviously, prices cannot constantly fluctuate to maintain a vacancy rate of exactly 15 percent, but they can vary sufficiently to avoid chronic over- or under-occupancy. Commercial parking operators always set prices high enough to avoid regularly putting out the "full" sign, and cities could contract with commercial operators to price curb parking properly, if necessary.

 

Parking Benefit Districts

Elsewhere I have argued that market prices can effectively regulate the off-street parking supply, and that the government’s chief contribution should be to set market prices for curb parking. I have also argued that, to make this pricing solution politically popular, cities could establish Parking Benefit Districts that dedicate curb parking revenue to pay for public services in the neighborhood where the revenue is collected. If the benefits financed by parking charges were visible and local, residents would want to charge market prices for curb parking for the revenue, not because they thought it good public policy. Residents who benefit from parking charges paid by strangers would begin to think like parking lot owners.

 

The economic argument to charge for curb parking is efficiency–the benefits would outweigh the costs. The political argument to create Parking Benefit Districts is distribution–the benefits for neighborhoods would lead residents to vote for the proposal. Parking meters have few friends if their revenue disappears into the city’s general fund. Curb parking revenue needs the appropriate recipient–its neighborhood–before residents will recommend market prices for parking. For example, parking revenue could pay to plant street trees, repair sidewalks, or underground utility wires. Curb parking charges would yield more revenue than the property taxes in many neighborhoods, so many residents could reap enormous benefits. Charging strangers to park in front of your house is like Monty Python’s scheme for Britain to tax foreigners living abroad.

 

Charging for parking does not require a meter at every space. Several payment systems–from high-tech electronic in-vehicle meters and multispace meters to low-tech paper stickers–have eliminated the practical and aesthetic objections to charging for parking. Where the potential revenues are high and the collection costs are low, the transaction costs of charging for parking are not a serious objection. The problem is political, not technical, and dedicating curb parking revenue to its neighborhood can solve the political problem.

 

 

A Model of Parking Choice

If market prices allocated parking spaces, how would motorists decide where to park? A simple model of parking choice will help to answer this question. To anticipate the results, market prices will allocate the most convenient parking spaces to motorists who: (a) carpool, (b) park for a short time, (c) walk slowly, and (d) place a high value on reducing walking time. Conversely, market prices will allocate the peripheral parking spaces to motorists who: (a) drive alone, (b) park for a long time, (c) walk fast, and (d) place a low value on reducing walking time.

 

Variables in the Model of Parking Choice

Suppose the price of parking is highest at the destinations where parking demand is highest, and that the price declines with distance from these destinations. Since the price of parking increases as you drive toward your destination, you will pay more money to park closer to your destination but you will also spend less time walking from your car to your destination. Given the trade-off between money spent on parking and time spent on walking, where should you park your car and walk the rest of the way?

 

To find the optimal parking space, consider the following variables (and their dimensions):

d the distance from parking space to final destination (miles)

p(d) the price of parking at distance d from the final destination ($/hour)

t parking duration (hours)

w walking speed from parking space to final destination (miles/hour)

n number of persons in the car (persons)

v average value of time spent walking ($/hour/person).

 

The total cost associated with parking at any location is the money cost of parking plus the time cost of walking from the parking space to the final destination and back. The money cost of parking equals the parking duration multiplied by the price per hour, or tp(d). The time to walk from the parking space to the final destination and back is 2d/w, the distance walked divided by the walking speed. To convert this time cost of walking into its money equivalent we can multiply the walking time by the dollar value of time, v. Because everyone in the car, n, experiences this time cost, the (monetized) cost of time spent walking equals 2nvd/w. At distance d from the final destination the total cost of parking and walking is therefore

tp(d) + 2nvd/w. (1)

 

The first term of the expression is the money cost of parking, and the second term is the (monetized) time cost of walking from the parking space to the final destination and back.

 

The Optimal Parking Space

What parking location minimizes the total cost of parking and walking? As you drive toward your destination the cost of parking increases and the cost of walking decreases. The minimum total cost of parking and walking occurs where the increase in the money cost of parking balances the decrease in the time cost of walking. If the money cost of parking increases less than the time cost of walking decreases as you approach your destination, you should keep driving. If the money cost of parking increases more than the time cost of walking decreases, you have driven too far.

 

Differentiating equation (1) with respect to d and setting the result equal to zero gives the distance from a final destination that minimizes the total cost of parking and walking.

tp/d + 2nv/w = 0, and -tp/d = 2nv/w. (2)

 

The changes in the money cost of parking (tp/d) and the time cost of walking (2nv/w) are equal in value and opposite in sign for any small movement from the location that minimizes the total cost of parking and walking. A parking space substantially closer to your final destination will increase the money cost of parking by more than it reduces the time cost of walking. A parking space substantially farther from your destination will increase the time cost of walking by more than it reduces the money cost of parking. The optimal parking space perfectly balances greed and sloth.

 

 

An Example

Suppose the price of parking is $1 an hour at your destination, and that the price declines with distance from your destination according to the negative exponential formula

p(d) = $1e-2d. (3)

 

Equation (3) implies that the price of parking, p, declines with distance, d, from the center, and that the slope of the curve relating price to distance also declines with increasing distance from the center (see Figure 5). A negative exponential curve is typical of the relationship between commercial parking prices and the distance from activity centers.

 

Figure 5 The Cost of Parking and Walking

 

 

Suppose that you want to park for 4 hours (t = 4), you are alone (n = 1), your time is worth $8 an hour (v = $8), and you walk 4 miles an hour (w = 4). Figure 5 shows the cost of parking and of walking as a function of parking d miles from your destination. The money cost of parking 4 hours is $4e-2d, which declines with distance from your destination. The time cost of walking is (2x1x$8/4)d, which increases with distance from your destination. The total cost of parking and walking (the upper curve in Figure 5) reaches its minimum value of $3.35 at a distance somewhere between 0.3 and 0.4 miles from your destination. To minimize the total cost of parking and walking you should park about a third of a mile from your destination and walk the rest of the way.

 

Solving equation (2) gives the exact distance that minimizes the total cost of parking and walking. Substituting equation (3) into equation (2) and solving for the optimal distance from a final destination, denoted as d*, gives

d* = [-loge(nv/tw)]/2. (4)

 

Given the values of n = 1 person, v = $8 an hour, t = 4 hours, and w = 4 miles an hour, the value for d* in equation (4) is 0.34 miles. At this distance the price of parking is 50¢ an hour, so the cost of parking four hours is $2. Walking the round trip of 0.68 miles from parking space to final destination and back at four miles an hour will take about 10 minutes. If time costs $8 an hour, 10 minutes will cost $1.35. The minimum total cost of parking and walking to your destination is thus $3.35 for the trip (see Figure 5).

 

The total money-and-time cost curve is flat between 0.25 and 0.5 miles from the destination because the slopes of the money-parking-cost and monetized-time-cost curves are about equal in absolute value but opposite in sign within this range. The total cost of parking and walking is about $3.35 anywhere between 0.25 and 0.5 miles from your destination. Parking less than 0.25 miles or more than 0.5 miles from your destination increases the total cost of parking and walking. For example, the total cost of parking and walking is $4 both at your destination and also at 0.8 miles from your destination.

 

Implications of the Model

Motorists do not make mathematical calculations when choosing where to park. The proposed parking location model merely expresses in mathematical form some of the various factors that motorists surely consider when they pay to park. The model confirms common sense, but several of its predictions are not immediately obvious.

 

First, the number of persons in a car is as important as the value of their time in determining parking location. For example, a carpool of four people who each value time at $5 an hour (nv = 4 x 5) will choose the same location as a solo driver who values time at $20 an hour (nv = 1 x 20), all else equal. A higher vehicle occupancy and a higher value of time justify parking closer to the final destination.

 

Second, parking duration is as important as the value of time in determining parking location. For example, a solo driver who values time at $10 an hour and parks for one hour (v/t = 10/1) will choose the same location as another solo driver who values time at $20 an hour and parks for two hours (v/t = 20/2), all else equal. A shorter parking duration justifies parking closer to the final destination.

 

Third, the number of persons in a car is as important as parking duration in determining parking location. For example, a solo driver who parks for one hour (n/t = 1/1) will choose the same location as a three-person carpool who park for three hours (n/t = 3/3), all else equal.

 

Table 3 shows the derivatives and elasticities of the optimal distance, d*, with respect to the variables that determine it. The derivative of d* is positive with respect to t and w, which implies that the longer you park and the faster you walk, the farther away you should park. The derivative of d* is negative with respect to n and v, which implies that the more people in your car and the higher value of their time, the closer in you should park.

 

TABLE 3 Elasticity of D* With Respect to Parking Choice Variables

 

Variable

Partial Derivative of d*

with Respect to Variable i

Elasticity of d*

with Respect to Variable i

t (parking duration)

d*/t = +1/(2t) >0

Ît = +1/(2d*) >0

w (walking speed)

d*/w = +1/(2w) >0

Îw = +1/(2d*) >0

n (number of persons)

d*/n = -1/(2n) <0

În = -1/(2d*) <0

v (value of time)

d*/v = -1/(2v) <0

Îv = -1/(2d*) <0

Note: d* = [-loge(nv/tw)]/2 and Îi = (d*/i)/(d*/i).

 

 

The elasticities of d* with respect to the variables that determine it decrease with increasing distance from the center (see Figure 6). For example, the elasticity of d* with respect to the parking duration, t, is +1/(2d*). At d* = 0.25 miles from the center, the elasticity of d* with respect to t is +2, so a 10-percent decrease in the length of time you want to park will shift your optimal parking location 20 percent closer to your final destination.

 

Figure 6 Elasticity of Parking Location Choice

 

These predictions are consistent with previous research on parking choices. David Gillen (1978) developed a model of parking location choice similar to the one expressed in equation (4), although he did not consider the number of persons in a car. Using data from Toronto, Gillen found that motorists who pay for parking by the hour are willing to trade a shorter parking duration for a closer parking location.

 

Using trip data from Vancouver, Brown and Lambe (1972) showed that allocating parking spaces by market prices will minimize the total walking time from parking spaces to final destinations. A linear programming model that minimizes total walking time predicted commercial off-street parking prices with an average error of only 20 percent. The price of curb parking was well below the level that would minimize total walking time.

 

Naturally, a simple model of parking prices like the one presented here does not describe most current parking decisions because parking is free for 99 percent of all automobile trips. The model is a simplified description of parking choice, but its assumptions are far more sensible than the assumptions behind minimum parking requirements.

 

 

Efficiency and Equity of Charging For Curb Parking

If curb parking were priced to yield a minimum vacancy rate of about 15 percent in every location, the resulting price gradients would shift predictably throughout the day as demand shifts. The peak parking prices might occur at employment centers during the day, at entertainment centers during the evening, and in high-density residential areas during the night. Many overlapping price gradients would form a three-dimensional parking price surface whose height at any point is the vertical summation of all the individual gradients. The individual gradients would form around many dispersed centers, like anthills covering a terrain that itself has peaks (the central business districts) and valleys (low density neighborhoods). The price of parking at any location would rise and fall during the day, and the local peaks would shift around like kittens fighting under a blanket.

 

Efficiency

Market prices would allocate parking spaces among motorists in a logical way. The more convenient parking spaces would go to carpoolers, those in a hurry, those who want to park for only a short time, those who have difficulty walking, and those more willing to spend money. The best parking spaces could always be reserved for those with physical disabilities. The more distant parking spaces would go to solo drivers, those with time to spare, those who want to park a long time, those who enjoy walking, and those more eager to save money.

 

Even if market prices can efficiently allocate a fixed stock of parking spaces, can market forces alone supply enough spaces to meet the demand for parking? If minimum parking requirements are eliminated, the ratio of parking spaces to cars will decline, and the price of parking will rise. This price rise will have two effects on demand and supply.

 

First, motorists will economize on parking by changing their travel behavior. Shifting to higher occupancy vehicles to spread the cost of parking among more people will reduce the demand for parking. Shifting to walking, cycling, or public transit will also reduce the demand for parking. Shifting vehicle trips to off-peak will reduce the demand for parking at peak hours. Finally, citizens can choose to own fewer cars, and this will reduce the demand for parking.

 

Second, freed from minimum parking requirements, developers will supply parking spaces in response to parking prices. The higher price of parking will encourage developers to voluntarily supply more parking in places where the resulting revenue will cover the cost of providing the parking. Parking will tend to become unbundled from other transactions, and firms that specialize in providing parking will manage more of the parking supply. Off-street parking prices will tend to cover the cost of providing parking spaces, including the cost of land, and these off-street prices will put a ceiling on the price of adjacent curb parking.

 

Flexible market prices can equate demand with the fixed supply of parking in the short run, and these prices will signal where the supply can profitably be increased in the long run. The proper role for the government is to price curb parking to maintain a minimum vacancy rate so that "enough" parking will always be available if motorists are willing to pay for it.

 

Market prices for parking resemble a spot market for land. Demand-responsive parking prices would reveal what parking spaces are really worth, and how motorists are willing to change their travel choices to save money on parking. Motorists could choose parking spaces according to how long they want to stay, how many people are in the car, how they value walking time (are they in a hurry? are they carrying heavy packages? are they tired? are they short of money?) and many other circumstances of time and place that only the individual motorists can know.

 

In contrast to the "spontaneous" order created by market prices and individual choices, urban planners require almost every land use to provide at least enough parking spaces to satisfy the peak demand for free parking. As a result, parking is free for almost every automobile trip because the cost of parking is shifted into higher prices for almost everything else. Minimum parking requirements in zoning ordinances are a disastrous substitute for millions of individual evaluations of what a parking space is worth.

 

Equity

The proposal to price curb parking rather than require off-street parking raises a serious political question. Is it fair to charge motorists for parking? To judge whether charging for parking is fair, it must be compared with the alternative–minimum parking requirements. Minimum parking requirements force everyone to pay for parking through higher prices for all other goods and services, but everyone does not benefit equally from free parking. On average, households with incomes below $10,000 a year own only one car, while households with incomes above $40,000 a year own 2.3 cars. Eight percent of non-Hispanic White households, 19 percent of Hispanic households, and 30 percent of African-American households do not own a car. In total, 10.6 million American households do not own a car, yet even these households indirectly pay the costs imposed by minimum parking requirements. Because cars are not distributed equally in the population, charging motorists only for the parking they use is fairer than requiring everyone to pay for parking whether they use it or not.

 

Market prices would not allocate the best parking spaces only to the rich. With market prices, motorists can pay less for parking if they carpool, stay for a shorter time, or park farther away, and they will pay nothing for parking if they walk, bicycle or ride public transit. Even those who cannot regularly afford to park in the best spaces can park in them on occasions when time is very important. Because income is only one factor that determines the value of time on a particular trip, and because the value of time is only one factor that determines parking location, income is only one of many factors that determine parking location.

 

Given the eternal debate on the merits of markets versus planning, many skeptics will distrust using prices to allocate parking spaces. But even those who doubt the ability of markets to allocate resources fairly may agree that relying on prices to allocate curb parking spaces and using the revenue to fund public services will contribute to a host of social, economic, and environmental goals they support.

 

 

Some Advantages of Market Prices for Parking

Market prices for parking will allow motorists to make many small adjustments to optimize their parking choices according to countless individual circumstances. Compared with minimum parking requirements, demand-responsive parking prices have major advantages.

  1. Charging for parking will give everyone an incentive to consider the alternatives to solo driving for every trip. Motorists will save money if they carpool, park for a shorter time, or walk farther.
  2. Motorists who stay for a short time will tend to choose the higher-priced central spaces. The resulting faster turnover of the central parking spaces will tend to minimize motorists’ total walking time to their final destinations.
  3. Motorists who stay for a long time–commuters, for example–will tend to choose the lower-priced peripheral spaces. They will also save money by riding transit, bicycling, or walking.
  4. Motorists will have flexibility. They can pay extra to park in the central spaces when they are in a hurry, and can save money by parking in the peripheral spaces when they are not in a hurry. Everyone can park in the more convenient spaces at off-peak times.
  5. If market prices reveal the economic value of on-street parking, cities can more rationally allocate scarce roadway space between parking and traffic. The cost of another lane for vehicle movement is the opportunity cost of a lane of parking spaces, and market prices for parking will take some of the guesswork out of regulating the on-street parking supply.

 

 

Conclusions: Time for a Paradigm Shift

Although it would be presumptuous to call urban planning a science, minimum parking requirements in planning resemble a paradigm in science. According to Thomas Kuhn (1996), a paradigm is a conceptual scheme that has gained universal acceptance throughout a profession, and each profession’s practices embody its ruling paradigms.

 

Kuhn argued that scientific education inculcates in students an intense commitment to the existing scientific paradigms. But planning education ignores parking requirements, and therefore does not inculcate in students any commitment to them. Instead, motorists have come to expect the free parking that the requirements produce. The planning profession’s commitment to parking requirements is based not on education and science but on motorists’ yearning to park free.

 

Discussing the difficulty of paradigm shifts in science, Kuhn asks, "How can a conceptual scheme that one generation admiringly describes as subtle, flexible, and complex become for a later generation merely obscure, ambiguous, and cumbersome?" Without doubt, minimum parking requirements are obscure, ambiguous, and cumbersome. In addition, minimum parking requirements impose enormous hidden costs, and they impede our progress toward important social, economic, and environmental goals. Planning for parking deserves a new paradigm.

 

Minimum parking requirements are based on two highly unreasonable assumptions: (1) the demand for parking does not depend on its price, and (2) the supply of parking should not depend on its cost. This neglect of price and cost stems from a belief that planners can assess community needs and can regulate the land market to meet these needs. Regulation is justified in many cases where market prices fail to communicate social costs. But market failure does not justify minimum parking requirements.

 

Without considering the price of parking--as if it were irrelevant--urban planners foretell how many parking spaces every land use needs. In practicing the art of predicting demand without considering price, urban planners resemble the Wizard of Oz, deceived by his own tricks. After he is exposed, the Wizard laments, "I have fooled everyone so long that I thought I should never be found out… [but] how can I help being a humbug when all these people make me do things that everybody knows can't be done?"

 

Letting prices determine the number of parking spaces will transfer to the market an important function that urban planners now perform. But this does not mean an end to planning for parking because planners should regulate many other features of parking that affect the community, such as aesthetics, landscaping, layout, location, pedestrian access, provisions for the handicapped, setback, signage, and stormwater runoff.

 

Pricing curb parking rather than requiring off-street parking will improve urban design, reduce traffic congestion, restrain urban sprawl, conserve natural resources, and produce neighborhood public revenue. Eliminating parking requirements will also reduce the cost of housing and of many other goods and services. In conclusion, deregulating the quantity and increasing the quality of parking will improve transportation, land use, and the environment.

 

APPENDIX: THE COST OF PARKING SPACES

 

How much does a parking space cost? This question has no easy answer because the cost of parking depends on the value of land, which varies greatly among sites. But in the case of structured parking we can account for the value of land as its opportunity cost for surface parking. The number of spaces a parking structure adds to the parking supply is the number of parking spaces in the structure minus the number of surface parking spaces lost as a result of building the structure. The structure’s construction cost (excluding land value) divided by the number of parking spaces added to the parking supply gives the structure’s cost per parking space added, which accounts for land value as the opportunity cost of the surface parking spaces lost (Shoup 1997).

 

This methodology was used to calculate the construction cost per parking space added by twelve parking structures built on the UCLA campus between 1961 and 1991. Each structure's original cost was converted into dollars of 1998 purchasing power by adjusting for construction cost inflation since the structure was built.

 

The average cost of the six structures built in the 1960s was $13,400 per space added, while the average cost of the six structures built since 1977 was $25,600 per space added. The newer parking structures are more expensive because they are smaller and partly or entirely underground, compared with the larger, aboveground structures built earlier. That is, the type of parking structure–not an increase in the real cost of parking spaces (above the rate of inflation of general construction costs)–can explain the higher real cost of new parking spaces.

 

Table 4 Cost of Aboveground and Belowground Parking Spaces

(Costs Per Space Added by Parking Structures in Los Angeles)

Underground (UCLA

Underground (Pershing Sq.)

   

 

1964 Structure

1995 Addition

1983 Structure

1998 Addition

1950

Structure

Current US$

$1,946

$13,712

$19,752

$26,300

$2,500

1998 US$

$12,214

$14,725

$28,540

$26,300

$28,000

The original portion of the Structure 3, built in 1964, contains 1,168 spaces in five aboveground levels; the addition built in 1995 contains 840 spaces in seven aboveground levels. The original portion of Structure 4, built in 1983, contains 448 spaces in two underground levels; the addition built in 1998 contains 1,263 spaces in two aboveground levels. The Pershing Square Garage in downtown Los Angeles contains 2,150 spaces in three underground levels. The ENR Construction Cost Index is used to convert original construction costs to 1998 values.

 

We can test this hypothesis that the type of parking structure explains the increase in cost after 1977. Since the initial study of the twelve structures built between 1961 and 1991, UCLA has built two new campus parking structures as additions to existing parking structures. The first is a 1995 aboveground addition to the aboveground structure built in 1964. The second is a 1998 underground addition to the underground structure built in 1983. Table 4 compares the cost per parking space added by the two original structures and their subsequent additions. The ENR Construction Cost Index is used to convert the original construction costs to 1998 dollars.

 

The cost was $12,214 per space for the original aboveground structure built in 1964, and $14,725 per space for the addition built 31 years later. The cost was $28,540 per space for the original underground structure built in 1983, and $26,300 per space for the addition built fifteen years later. The close match between the cost of each original parking structure and the cost of its later addition suggests that, after correcting for inflation, the cost of building parking structures has changed little in recent decades.

 

To test this finding of cost stability, Table 4 also shows the cost of an underground garage constructed beneath Pershing Square in downtown Los Angeles in 1952. When the original cost of $2,500 per space is converted to its equivalent in 1998 purchasing power, the cost of the Pershing Square garage is $25,700 per parking space, very close to the cost of the two underground garages built at UCLA in 1983 and 1998. In real terms, the cost of building underground parking has not changed in half a century.

 

If these high costs are surprising, it is only because the cost of parking is rarely calculated. Nevertheless, there is other evidence about cost because some cities allow developers to pay a fee in lieu of providing required parking spaces. To justify the in-lieu fees, some of these cities carefully document their cost of providing public parking spaces. In Palo Alto, California, the cost is $17,848 per space added by a municipal parking structure. In Lake Forest, Illinois, the cost is $18,000 per space for the land and construction cost of surface parking lots. In Walnut Creek, California, the cost is $32,400 per space added by a municipal parking structure. In Beverly Hills, California, the average cost was $37,000 per space for the estimated land and construction cost of municipal parking structures. The cost of parking spaces at UCLA is thus in line with the cost of parking spaces in cities that allow developers to pay in-lieu fees.

 

The cost of many surface parking spaces is less than the cost of structured parking spaces, but land values understate the cost of surface parking because developers who are required to provide parking spaces will bid less for land. Therefore, the market value of land subject to a minimum parking requirement will understate the cost of surface parking spaces. For example, when Oakland, California, introduced its parking requirement of one space per 1,000 square feet for apartment buildings, land values fell by 33 percent (Shoup 1997). Willson (1995) estimated that increasing the parking requirement for office buildings in Southern California by 1.3 spaces per 1,000 square feet would reduce land values by 32 percent. Because minimum parking requirements depress land values, low land values do not necessarily imply that minimum parking requirements have a low cost.

 

REFERENCES

American Public Transit Association (1997) 1997 Transit Fact Book. American Public Transit Association. Washington, DC; www.apta.com.

Brierly, J. (1972) Parking of Motor Vehicles. Applied Science Publishers, Second Edition. London.

Brown, S. A. and Lambe, T. A. (1972) Parking prices in the central business district. Socio-Economic Planning Sciences, 6, 133-144.

Chapin, F. S. (1957) Urban Land Use Planning. Harper & Brothers. New York.

Chapin, F. S. (1965) Urban Land Use Planning. University of Illinois Press, Second Edition. Urbana, IL.

Chapin, F. S. and Kaiser, E. (1979) Urban Land Use Planning. University of Illinois Press, Third Edition. Urbana, IL.

Federal Highway Administration (1995) Highway statistics summary to 1995, taken from www.bts.gov/site/news/fhwa/HighwayStats-Summary95/section2.html. Office of Highway Information Management.

Bureau of Transportation Statistics (1997) Federal, State and Local Transportation Financial Statistics, Fiscal Years 1982-94, BTS97-E-02, United States Department of Transportation. Washington, DC.

Gillen, D. (1978) Parking policy, parking location decisions and the distribution of congestion. Transportation, 7, 69-85.

Goulard, T. (1784) A Treatise on the Effects and Various Preparations of Lead: Particularly of the Extract of Saturn, for Different Chirurgical Disorders. Elmsley in the Strand. London.

Greene, D., Jones, D. and Delucchi M., editors (1997) The Full Social Costs and Benefits of Transportation. Springer-Verlag. Heidelberg.

Gruen, V. (1973) Centers for the Urban Environment: Survival of the Cities. Van Nostrand Reinhold Co. New York.

Hanson, S., editor (1995) The Geography of Urban Transportation. Guilford Press. New York.

Institute of Transportation Engineers (1987a) Parking Generation. Institute of Transportation Engineers, Second Edition. Washington, DC; www.ite.org.

Institute of Transportation Engineers (1987b) Trip Generation. Institute of Transportation Engineers, Fourth Edition. Washington, DC; www.ite.org.

Kaiser, E., Godschalk, D., and Chapin, F. S. (1995) Urban Land Use Planning. University of Illinois Press, Fourth Edition. Urbana, IL.

Katz, A. and Herman, S. (1997) Improved estimates of fixed reproducible tangible wealth, 1929-1995. Survey of Current Business, May 1997.

Klose, D. (1965) Multi-story and underground garages [Parkhäuser und Tiefgaragen]. Verlag Gerd Hartje. Stuttgart.

KMPG Peat Marwick (1990) Dimensions of parking, prepared for the United States Department of Transportation, Urban Mass Transportation Administration, Office of Budget and Policy, September 10, 1990.

Kuhn, T. (1957) The Copernican Revolution. Harvard University Press. Cambridge, MA.

Kuhn, T. (1996) The Structure of Scientific Revolutions. University of Chicago Press, Third Edition. Chicago.

Lin-Fu, J. (1992) Modern history of lead poisoning: a century of discovery and rediscovery. In Human Lead Exposure. ed. H. Needleman. CRC Press. Boca Raton, FL.

May, A. D. (1975) Parking control: experience and problems in London. Traffic Engineering and Control. May.

McCord, C. (1953) Lead and lead poisoning in early America. Industrial Medicine and Surgery, 22(9), 393-399.

Meyer, M. and Miller, E. (1984) Urban Transportation Planning. McGraw Hill Book Company. New York.

Muth, R. (1969) Cities and Housing. University of Chicago Press. Chicago.

Jerome Nriagu. 1983. Lead and Lead Poisoning in Antiquity, New York: John Wiley & Sons.

Pisarski, A. (1995) The demography of the U.S. vehicle fleet: observations from the NPTS. 1990 Nationwide Personal Transportation Survey, Special Reports on Trip and Vehicle Attributes. U.S. Department of Transportation. Washington, DC; www-cta.ornl.gov/npts/1995/doc/index.html-ssi.

Pisarski, A. (1996) Commuting in America II: The Second National Report on Commuting Patterns and Trends. Eno Transportation Foundation. Landsdowne, VA; www.enotrans.com.

Planning Advisory Service (1971) An Approach to Determining Parking Demand, Planning Service Report Number 270. American Planning Association. Chicago; www.planning.org.

Planning Advisory Service (1991) Off-Street Parking Requirements: A National Review of Standards, Planning Advisory Service Report Number 432. American Planning Association. Chicago; www.planning.org.

Shoup, D. C. (1992) Cashing Out Employer-Paid Parking. U. S. Department of Transportation. 156 pp. Washington, DC.

Shoup, D. C. (1994) Cashing in on curb parking. Access (http://socrates.berkeley.edu/~uctc), 4, 20-26.

Shoup, D. C. (1995) An opportunity to reduce minimum parking requirements. Journal of the American Planning Association, 61(1), 14-28.

Shoup, D. C. (1997) The high cost of free parking. Journal of Planning Education and Research. 17(1), 1-18.

Shoup, D. C. (forthcoming) In-lieu parking fees. Journal of Planning Education and Research.

Shoup, D. and Breinholt, M. J. (1997) Employer-paid parking: a nationwide survey of employers’ parking subsidy policies. In The Full Social Costs and Benefits of Transportation. ed. D. Greene, D. Jones and M. Delucchi. Springer-Verlag. Heidelberg.

Thomas, L. (1981) "Medicine without science." The Atlantic Monthly. April 1981, 40-42.

Willson, R. (1995) "Suburban parking requirements: a tacit policy for automobile use and sprawl." Journal of the American Planning Association. 61(1), 29-42.

Willson, R. (1996) Local jurisdiction parking requirements: a survey of policies and attitudes. Working Paper, Department of Urban and Regional Planning, California State Polytechnic University, Pomona, California.

Witheford, D. K. and Kanaan, G. E. (1972) Zoning, Parking, and Traffic. Eno Foundation for Transportation. Westport, CT; www.enotrans.com.

 

 

 

 

ACKNOWLEDGMENTS

I am grateful to the Federal Transit Administration and the University of California Transportation Center for financial support. For their many suggestions for improving this paper I am also grateful to Ellison Alegre, Lee Burns, Jeffrey Brown, Eric Carlson, Joy Chen, Elke Daugherty, D. Gregg Doyle, David Gillen, Daniel Hess, Eugene Kim, Kristen Massey, Andrew Mondschein, Virginia Parks, John Pucher, Thomas Rice, Gian-Claudia Sciara, Patricia Shoup, Seth Stark, Richard Willson, and Matthew Zisman.