Content uploaded by Donald Shoup
Author content
All content in this area was uploaded by Donald Shoup on May 19, 2018
Content may be subject to copyright.
Transport Policy 13 (2006) 479–486
Cruising for parking
Donald C. Shoup
Department of Urban Planning, University of California, Los Angeles, Los Angeles, CA 90095-1656, USA
Available online 24 July 2006
Abstract
Suppose curb parking is free but all the spaces are occupied, and off-street parking is expensive but immediately available. In this case,
you can cruise to find a curb space being vacated by a departing motorist, or pay for off-street parking right away. This paper presents a
model of how drivers choose whether to cruise or to pay, and it predicts several results: you are more likely to cruise if curb parking is
cheap, off-street parking is expensive, fuel is cheap, you want to park for a long time, you are alone in the car, and you place a low value
on saving time. The model also predicts that charging the market price for curb parking—at least equal to the price of adjacent off-street
parking—will eliminate cruising. Because the government sets curb parking prices, planners and elected officials strongly influence
drivers’ decisions to cruise. The failure to charge market rates for curb parking congests traffic, pollutes the air, wastes fuel, and causes
accidents. Between 1927 and 2001, studies of cruising in congested downtowns have found that it took between 3.5 and 14 min to find a
curb space, and that between 8 and 74 percent of the traffic was cruising for parking.
r2006 Elsevier Ltd. All rights reserved.
Keywords: Parking; Pricing; Congestion
1. Introduction
My father didn’t pay for parking, my mother, my
brother, nobody. It’s like going to a prostitute. Why
should I pay when, if I apply myself, maybe I can get it
for free.
George Costanza
When a resource is communally owned, the right of
‘‘first possession’’ means that anyone who captures the
resource has the right to use it. Free curb parking is an
example of communal ownership, because drivers occupy it
on a first-come, first-served basis. If all the curb spaces are
occupied, drivers must cruise to find a space vacated by a
departing car. Cruising for parking probably began soon
after the wheel was invented.
2. Cruising in the 20th century
Cruising creates a mobile queue of cars that are waiting
for curb vacancies, but no one can see how many cars are
in the queue because the cruisers are mixed in with other
cars that are actually going somewhere. Perhaps because
cruising is invisible, most transport economists and
planners have neglected it as a source of congestion.
Nevertheless, a few researchers have attempted to estimate
the volume of cruising and the time it takes to find a curb
space. They have analyzed videotapes of traffic flows,
interviewed drivers who park at the curb, and have
themselves cruised. Table 1 shows the results of every
study of cruising I have been able to find. Between 8 and 74
percent of the traffic was cruising for parking, and the
average time to find a curb space ranged between 3.5 and
14 min. The wide variance in the estimates of cruising
surely reflects reality. On most streets most of the time,
none of the traffic is cruising, but on some streets some of
the time, most of the traffic may be cruising.
Are these studies dating back to 1927 only of historical
interest? The data were probably not very accurate when
they were collected, and the results depend on the time and
place where they were collected. The data are selective
because researchers study cruising only where they expect
to find it. Conditions have also changed since many of the
observations were made. Nevertheless, cruising itself has
not changed, and the studies at least show that searching
for curb parking has wasted time and fuel for many
ARTICLE IN PRESS
www.elsevier.com/locate/tranpol
0967-070X/$ - see front matter r2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tranpol.2006.05.005
Tel.: +1 310 825 5705; fax: +1 310 206 5566.
E-mail address: shoup@ucla.edu.
decades. Because curb parking is underpriced and over-
crowded in the busiest parts of most of the world’s big
cities, the sun never sets on cruising.
Even a small search time per car can create a surprising
amount of traffic. Consider, for example, a congested
downtown where it takes three minutes to find a curb space
and the parking turnover is 10 cars per space per day. Each
curb space generates 30 min of cruising time per day. If the
average cruising speed is 10 miles an hour, each curb space
generates five vehicle miles traveled (VMT) per day. Over a
year, this cruising amounts to 1825 VMT, greater than
halfway across the United States, for each curb space.
Because this cruising adds to traffic that is already
congested, it makes a bad situation even worse.
Where on-street parking is cheaper than off-street
parking, cruising is individually rational. Collectively,
however, it congests traffic, causes accidents, wastes fuel,
pollutes the air, and degrades the pedestrian environment.
Cities create all these problems when they underprice curb
parking. Underpricing of anything creates a shortage, and
curb parking is no exception. Underpriced curb parking is
gross mismanagement of scarce urban land, with wide-
spread ramifications for transportation, cities, the econo-
my, and the environment.
3. Choosing to cruise
When we cruise for parking on crowded streets, we
rarely seem to think about how we end up in this mobile
purgatory. How do you choose whether to cruise or to pay?
A simple model of the benefits and costs of cruising can
help answer this question. The model predicts several
results: you are more likely to cruise if curb parking is
cheap, off-street parking is expensive, fuel is cheap, you
want to park for a long time, you are alone in the car, and
you place a low value on saving time.
To set the scene for the model, suppose curb parking is
free but so crowded that you have to spend time hunting
for a space. You can park off-street without waiting, but
you have to pay for it. Given the tradeoff between spending
time to find curb parking or spending money to pay for off-
street parking, should you cruise or pay?
Drivers do not explicitly calculate whether to cruise or to
pay, but several factors influence the decision. To help
understand the choice, consider the following variables
(and their dimensions).
1
pprice of curb parking ($/h)
mprice of off-street parking ($/h)
tparking duration (h)
ctime spent searching for parking at the curb (h)
ffuel cost of cruising ($/h)
nnumber of people in the car (persons)
vvalue of time spent cruising ($/h/person)
We can use these seven variables to compare the time
and fuel cost of cruising with the money cost of parking
off-street. I will, for the moment, assume that the walking
time from the parked car to the final destination is the same
for both curb and off-street parking.
First, consider how much you save on parking if you can
find a curb space. The price of parking at the curb is p
dollars per hour, and the price of parking off-street is m
dollars per hour, so parking at the curb rather than off-
street saves mpdollars per hour. The amount you save by
parking at the curb is the duration (t) multiplied by the
difference between the prices of off-street and curb
parking, or t(mp). For example, if curb parking is free,
off-street parking costs $1 an hour, and you park for two
hours, you save $2 by parking at the curb.
2
Second, cruising has a fuel cost. If your car consumes
fuel at a rate of fdollars per hour of cruising for parking,
and you cruise for chours, the total cost of fuel spent
ARTICLE IN PRESS
Table 1
Twentieth century cruising
Year City Share of traffic
cruising (percent)
Average search
time (min)
1927 Detroit (1) 19%
1927 Detroit (2) 34%
1933 Washington 8.0
1960 New Haven 17%
1965 London (1) 6.1
1965 London (2) 3.5
1965 London (3) 3.6
1977 Freiburg 74% 6.0
1984 Jerusalem 9.0
1985 Cambridge 30% 11.5
1993 Cape Town 12.2
1993 New York (1) 8% 7.9
1993 New York (2) 10.2
1993 New York (3) 13.9
1997 San Francisco 6.5
2001 Sydney 6.5
Average 30 8.1
Note: The numbers after Detroit, London, and New York refer to
different locations within the same city.
Sources: Simpson (1927),Hogentogler et al. (1934),Huber (1962),Inwood
(1966),Bus+Bahn (1977),Salomon (1984),O’Malley (1985),Clark
(1993a, b),Falcocchio et al. (1995),Saltzman (1994), and Hensher (2001).
1
For the model, I assume that drivers are indifferent between curb and
off-street parking if the price and the time required to find a space are the
same in both cases. There are no time limits, and drivers pay in a linear
proportion to the number of minutes parked in both cases. In reality, curb
parking is often more convenient if a space is available right in front of the
final destination because the walking distance from the car to the
destination is shorter. Curb parking is also often available in smaller
increments (such as 6 or 10 min) while off-street parking is available only
in larger increments (such as 20 or 30 min), and the price per minute
declines for longer stays.
2
I assume that you know in advance how long you want to park, and
that you pay only for the exact time that you park. The parking charge is a
linear function of the number of minutes you park, with no advance
commitment to how long you park. Shoup (2003) describes parking meters
that allow you to pay for the exact time that you park, determined ex post.
D.C. Shoup / Transport Policy 13 (2006) 479–486480
for cruising is fc.
3
For example, if the fuel cost is $1
per hour and you cruise for 6 min (0.1 h), the fuel
cost is 10b. (If drivers ignore the cost of fuel for cruising,
f¼0.)
Third, cruising has a time cost. The value you place on
saving time depends on your income and on many factors
that are unique to each trip: whether you are in a hurry, the
weather, the scenery, safety, your health, and so on. The
value of time will, of course, vary from person to person,
but even the same person will place a higher or lower value
on time depending on the circumstances. Each person’s
time cost of cruising is the value of time (v) multiplied by
the time spent cruising (c), or vc. Because every person in
the car must spend the same time cruising, the total time
cost for everyone in the car is the number of people in the
car (n) multiplied by each person’s time cost (vc), or nvc.
4
So if you are alone in the car, value time savings at $9 an
hour, and cruise for 6 min before parking, your cost of time
spent cruising is 90b. Adding one passenger in the car
doubles the time cost to $1.80. A second passenger makes it
$2.70, and so on.
The money saved by parking at the curb and the cost of
cruising for a curb space are,
tðmpÞ, (1) money saved by parking at
the curb
fc, (2) money cost of cruising for
curb parking
nvc, (3) monetized cost of time spent
cruising for curb parking
fc þnvc ¼cðfþnvÞ, (4) money and (monetized) time
cost of cruising for curb
parking
At what point does cruising for curb parking become
more expensive than paying to park off street right away?
Let c* denote the time that equates the time-and-fuel cost
of cruising with money cost of off-street parking. There is
no cost difference between cruising and paying if you
expect to spend exactly c* minutes to find a curb space, so
you are indifferent between the two choices.
5
This
equilibrium occurs when the money saved from parking
at the curb, t(mp), equals the money and time cost of
cruising, c*(f+nv). So if you expect that it will take longer
than c* to find a curb space, you should pay to park off-
street. But if you expect that it will take less than c*, then
you should cruise.
The break-even point occurs where the cost of cruising
equals the savings from parking at the curb.
cðfþnvÞ¼tðmpÞ(5)
The search time at which you are indifferent between
cruising and paying is,
c¼tðmpÞ
fþnv (6)
At time c*, you realize no net savings by parking at the
curb instead of off-street. The money the city loses from
underpriced curb parking does not accrue to you or to
anyone else, but is instead dissipated in time and fuel spent
cruising. And because each driver in congested traffic
imposes time delays on all other drivers, cruising makes
all drivers worse off, including those who are not trying
to park.
4. Equilibrium search time: an example
We can use an example to illustrate the equilibrium
search time. Suppose you want to park for one hour
(t¼1), off-street parking costs $1 an hour (m¼1), and
curb parking is free (p¼0). You thus save $1 by parking at
the curb rather than off-street. If you drive 10 miles an
hour and your car gets 20 miles per gallon of gasoline, the
cruising consumes half a gallon of gasoline an hour. If
gasoline costs $2 a gallon, the fuel cost is $1 an hour
(f¼1). You are alone in the car (n¼1) and your time is
worth $9 per hour saved (v¼9). The equilibrium search
time, c*, is
c¼tðmpÞ
fþnv ¼1ð10Þ
1þ19¼0:1h ¼6 min
In this case it is worth spending up to 6 min to find a curb
space. If fuel costs $1 an hour, and you cruise for 6 min
(0.1 h), you spend 10bfor fuel ($1 0.1). You save $1 on
parking for an hour, so your net saving from parking at the
curb is 90b($1 saving on parking minus 10bspent for
fuel). In a sense, you ‘‘earn’’ $9 an hour for the time spent
cruising (90bsaved for 0.1 h of cruising). If you value time
savings at $9 an hour, 6 min is the search time that leaves
you indifferent between searching for curb parking and
paying to park off-street immediately. You are no better
off parking free at the curb after searching for 6 min than if
you had paid $1 to park off-street immediately.
6
ARTICLE IN PRESS
3
The fuel cost per hour of cruising is the cost of fuel per gallon divided
by the miles cruised per gallon and then multiplied by the cruising speed.
For example, if gasoline costs $1 per gallon, and your car gets 20 miles per
gallon, cruising costs 5bper mile. If you cruise at 20 miles per hour, the
fuel cost of cruising is $1 per hour.
4
The cost of time spent cruising, v, may differ among persons in the car.
If everyone’s value of time is weighted equally, we can interpret vas the
average value of time.
5
The time spent walking from the car to the final destination is neglected
here.
6
If it takes less than 6 min to find a curb space, you earn more than $9
an hour by cruising, so cruising for free curb parking is cheaper than
paying to park off-street immediately. If it takes more than 6 min to find a
curb space, you earn less than $9 an hour by cruising, so paying to park
off-street immediately is cheaper than cruising for free curb parking. This
appears to suggest that cruising should allocate curb parking to drivers
who place a lower value on saving time spent driving. Nevertheless, a
driver who has a higher value of time and wants to park for a longer
duration may be willing to cruise longer than a driver who has a lower
value of time who wants to park for a shorter duration because
c¼tðmpÞ=ðfþnvÞ, with parking duration in the numerator and value
of time in the denominator.
D.C. Shoup / Transport Policy 13 (2006) 479–486 481
This example suggests two results. First, ‘‘free’’ curb
parking is not really free. Although cruising’s costs are not
directly out-of-pocket (like money put in a parking meter),
they are incurred in the form of time and fuel used to find a
curb space. ‘‘Free’’ curb parking leaves the driver no better
off, but everyone else is worse off because cruising congests
traffic and pollutes the air; the city also loses the money it
would have received if it had charged the market price for
curb parking. Second, since the time spent cruising is the
price of curb parking, this price depends on each person’s
opportunity cost of time.
7
In this example, solo drivers who
value time savings at more than $9 an hour should pay to
park right away, and those with a lower value of time
should cruise. Free curb parking thus attracts solo drivers
who place a low value on saving time. Areas where curb
parking is free, many cars are solo driven, and drivers
place a low value on saving time will therefore have long
search times.
This equilibrium-search-time example also suggests that
raising the price of curb parking reduces the parkers’ time-
and-fuel cost of cruising by as much as it increases their
money payments for parking. The net burden on parkers is
zero, and the price rise converts private waste into public
revenue.
5. The wages of cruising
Cities create the incentive to cruise when they charge less
for curb parking than the price of adjacent off-street
parking. To examine this incentive, I collected data on the
price of curb and off-street parking for an hour at noon at
the same location—City Hall—in 20 cities throughout the
US.
8
Table 2 shows the results. The average price is $1.17
an hour for curb parking, and $5.88 an hour off-street.
Cruising saves drivers the most money in New York, where
the price of off-street parking is $14.38 for the first hour,
but curb parking is only $1.50. Cruising saves money in all
cities except Palo Alto and San Francisco. In the 20 cities,
the average price of curb parking is only 20 percent of the
price of off-street parking, and the highest price of curb
parking is only $2 an hour.
The supply of parking affects its price. Boston’s
high price of off-street parking ($11) stems in part
from a cap the city has placed on the number of off-
street parking spaces available downtown. The parking
inventory is frozen at its 1975 level—35,500 spaces.
Developers who want to build new parking spaces
must buy licenses owned by existing parking facilities that
close.
9
This supply cap drives up the market price of off-
street parking and produces an ironic outcome: combined
with the low price of curb parking, the higher price of off-
street parking increases the incentive to cruise. Boston
limits the private off-street parking supply, but fails to
price its own public curb parking properly. A survey in
2003 found that the average price for off-street parking
in the Boston CBD was $390 a month, and $30 a day.
10
In
contrast, Boston charges the same price ($1 an hour) for all
ARTICLE IN PRESS
Table 2
The wages of cruising for parking at city hall (curb parking one hour at
noon)
City State Price of parking for
one hour
Savings for
finding a curb
space
Curb Off-street
(1) (2) (3) $/h (4) $/h (5) ¼(4)–(3) $
Baltimore MD $2.00 $6.00 $4.00
Berkeley CA $0.75 $1.00 $0.25
Boston MA $1.00 $11.00 $10.00
Buffalo NY $1.00 $3.00 $2.00
Cambridge MA $0.50 $4.00 $3.50
Chicago IL $1.00 $13.25 $12.25
Houston TX $0.25 $1.50 $1.25
Long Beach CA $2.00 $2.50 $0.50
Los Angeles CA $1.50 $3.30 $1.80
New Orleans LA $1.25 $3.00 $1.75
New York City NY $1.50 $14.38 $12.88
Palo Alto CA $0.00 $0.00 $0.00
Pasadena CA $1.00 $6.00 $5.00
Philadelphia PA $1.00 $3.00 $2.00
Portland OR $1.00 $1.50 $0.50
San Diego CA $1.00 $6.00 $5.00
San Francisco CA $2.00 $2.00 $0.00
Santa Barbara CA $0.00 $5.00 $5.00
Santa Monica CA $0.50 $4.20 $3.70
Seattle WA $1.00 $8.00 $7.00
Average $1.17 $5.88 $4.71
Assumptions: A solo driver parks for one hour at noon on a weekday.
7
Smolensky, Tideman, and Nichols (1972, 95) say, ‘‘Queues can be
viewed as prices assessed in time, and time prices, like money prices, ration
according to the tastes, income and opportunity costs of buyers.’’
8
The cities are an opportunistic sample of places where my research
assistants and I visited and were able to gather the data. Nevertheless, the
sample shows that curb parking is probably much cheaper than off-street
parking in many big and small cities. City Hall was chosen because it is a
standard reference point that everyone can recognize. The data were
collected in 2001–2003.
9
Boston Transportation Department (2001). The Boston Air Pollution
Control Commission administers the ‘‘parking freeze’’ in Boston Proper
(the downtown). The number of spaces available to the general public is
frozen at the 1975 level, but the Boston Air Pollution Control Commission
may grant exemptions to private off-street parking that is available
exclusively to employees, guests, or customers in a building. Residential
parking isn’t capped. The total off-street parking supply increased by only
9 percent between 1977 and 1997, and in 1997 the 35,500 public parking
spaces represented 60 percent of the total 59,100 off-street spaces in
Boston Proper. Additional freezes apply in East Boston, South Boston,
and at Logan Airport. Portland, Oregon, had a similar limit on the
number of parking spaces—known as the parking lid—in the CBD. It was
replaced in 1995 by limit on 0.7 spaces per 1000 square feet of net leasable
area, in part because historic buildings without any parking were losing
nearby surface parking lots and were increasingly difficult to lease
(Portland TriMet, 2002, 3–9).
10
Colliers International (2003, pp. 28–29). The highest price for
unreserved parking in Boston’s CBD was $600 a month, and the lowest
was $285 a month.
D.C. Shoup / Transport Policy 13 (2006) 479–486482
meters in the city. Far from using prices to manage the
demand for curb parking, Boston underprices curb parking
and thus encourages drivers to cruise for it.
Boston’s off-street parking cap makes sense as a way to
reduce congestion on routes to the city, but its failure to
follow through with market prices for curb parking
increases congestion in the city. Everyone would criticize
off-street parking operators if long lines of cars regularly
spilled into the streets and congested traffic because the lots
and garages were always full. Nevertheless, cities create the
same result with curb parking by underpricing it, and
nobody notices because the cars hunting for curb parking
are hidden in the general traffic flow.
6. Two pricing strategies
Cities can use two pricing strategies to discourage
cruising. The first is to charge the market price for curb
parking. When the prices of curb and off-street parking are
equal (p¼m), the equilibrium cruising time (c*) is zero.
c¼tðmpÞ
fþnv ¼tð0Þ
fþnv ¼0
If curb parking costs the same as off-street parking, why
drive around hunting for a curb space? Since curb parking
(after you spend time and money to find it) costs the same
as off-street parking, you don’t save any money by
cruising. If all curb spaces are occupied and there is no
economic incentive to cruise, you should park off-street
without wasting time and fuel.
11
If curb parking remains free, however, a second strategy
to discourage cruising is to reduce the price of off-street
parking to zero. The logic is the same: if off-street parking
is free, why drive around looking for a curb space? Since
the prices of curb and off-street parking are again equal
(m¼p¼0), the equilibrium time is again zero.
c¼tðmpÞ
fþnv ¼tð0Þ
fþnv ¼0
Cities can therefore eliminate cruising either by charging
market prices for curb parking or by requiring enough off-
street spaces to reduce the price of off-street parking to
zero. The price of curb parking is one of the few policy
variables that cities control directly, but almost all
American cities have chosen the wrong policy: require
plentiful off-street parking rather than charge fair market
prices for scarce curb parking.
7. Elasticities
Table 3 shows how each of the variables in the model
affect the decision whether to cruise or to pay. The second
column shows the partial derivatives of c* (the maximum
time a driver is willing to cruise) with respect to the
variables in the first column. Six factors affect the decision
to cruise: (1) the price of curb parking, (2) the price of off-
street parking, (3) parking duration, (4) the price of fuel,
(5) the number of persons in the car, and (6) the value
of time.
The third column shows the elasticity of c* with respect
to each variable. The coefficient of elasticity is denoted by Z
(the Greek letter eta). These elasticities show how a small
change in each variable increases or decreases the time a
driver is willing to cruise. Five results stand out.
First, Z
p
(the elasticity of search time with respect to the
price of curb parking) depends only on the prices of curb
and off-street parking. The elasticity is low when curb
parking is relatively cheap, which means that raising the
price—say, doubling it from 10bto 20ban hour—will
have little effect on curb vacancies. This result may lead
some to conclude that the demand for curb parking is
inelastic, and that raising the price of curb parking will not
produce vacancies. But as the price of curb parking
approaches the price of off-street parking, a small increase
can create curb vacancies and reduce congestion. The
demand for curb parking may be completely inelastic until
its price exceeds that of off-street parking, at which point it
can suddenly become very elastic.
12
Because the variables
ARTICLE IN PRESS
Table 3
Equilibrium search time
Variable Partial derivative of c* Elasticity of c*
p(curb parking price) qc
qp¼ t
fþnv o0Zp¼ p
mpo0
m(off-street parking
price)
qc
qm¼þ t
fþnv 40Zm¼þ m
mp40
t(parking duration) qc
qt¼þmp
fþnv 40Zt¼þ1
f(fuel cost of cruising) qc
qf¼tðmpÞ
ðfþnvÞ2o0Zf¼ f
fþnv o0
n(number of persons) qc
qn¼tvðmpÞ
ðfþnvÞ2o0Zn¼ nv
fþnv o0
v(value of time) qc
qv¼ntðmpÞ
ðfþnvÞ2o0Zv¼ nv
fþnv o0
Notes: The length of time (c*) a motorist is willing to search for curb
parking is: c¼tðmpÞ
fþnv . The elasticity (Z
i
)ofc* with respect to variable iis:
Zi¼qc=qi
c=i.
11
If drivers prefer curb to off-street parking, the price of curb parking
must rise above the price of off-street parking to create curb vacancies and
discourage cruising.
12
For example, when the price of curb parking is 25ban hour and the
price of off-street parking is $1 an hour, Zp¼ð:25Þ=ð1:25Þ¼0:33;
therefore, raising the price by 10 percent reduces the time drivers are
willing to cruise by only 3.3 percent. But when the price of curb parking is
75ban hour, Zp¼ð:75Þ=ð12:75Þ¼3; therefore, raising the price by 10
percent reduces the time drivers are willing to cruise by 30 percent. Even
this large reduction in search time may not produce many curb vacancies,
however. If curb parking is cheaper than off-street parking, the main effect
D.C. Shoup / Transport Policy 13 (2006) 479–486 483
that affect the elasticity vary from place to place and
according to the time of day, we cannot estimate a single
price elasticity of demand for curb parking at even one
location, let alone all locations.
Second, when curb parking is free (p¼0), the elasticity
of search time with respect to the price of off-street parking
(m) is always +1. Therefore, reducing the price of off-street
parking by 10 percent reduces the time you are willing to
cruise by 10 percent.
Third, the elasticity of search time with respect to
parking duration (t) is always +1. A longer parking
duration justifies a proportionally longer cruise. For
example, a driver who wants to park twice as long as
another is willing to hunt twice as long to find a curb space,
all else equal. Curb parking is best suited for short stays,
but drivers who want to park for a long time have a
stronger incentive to search for the curb spaces.
Fourth, the elasticity of search time with respect to fuel
cost (Z
f
) depends on the relative values of fuel (f), the
number of people in the car (n), and the value of their time
(v). If fuel cost is much lower than time value, an increase
in fuel cost has little effect on willingness to cruise. If
cruising costs $1 an hour for fuel, and time is worth $9 an
hour, a solo driver’s elasticity of cruising with respect to
price of fuel is only 0.1. Raising the price of gasoline by
10 percent therefore reduces the time drivers are willing to
cruise by only 1 percent.
13
Finally, the elasticity of search time is the same with
respect to the number of people in a car (n) and the value of
their time (v). More people in a car and a higher value of
their time thus have the same effect on willingness to cruise.
The lone motorist who values time at $8 an hour and a
carpool of four persons who each value time at $2 an hour
are both willing to cruise for the same length of time, all
else equal.
8. Right-priced curb parking: an illustration
The top panel of Fig. 1 illustrates the case where curb
parking is underpriced, all spaces are occupied, and cars
are circling the block looking for a space. It is based on
observations in Westwood Village, a commercial district in
Los Angeles next to the UCLA campus.
14
The average
block has eight curb spaces on each side, the average
cruising time before finding a curb space is 3.3 min, and
two cruisers are circling each block.
The small distances cruised by each driver add up
quickly because the turnover rate is 17 cars a day per space.
With 470 meters in the Village, 7990 cars park at the curb
each day (17 470). Because so many cars park at the curb,
a small cruising time for each driver creates an amazing
amount of traffic. Although the average driver cruises
only 3.3 min before parking, cruising in the aggregate
creates 440 vehicle-hours of cruising every day. At an
average speed of about 8 miles per hour, cruising around
the 15 blocks in the Village creates more than 3500 VMT
every day, or more than the distance across the United
States.
In contrast, the bottom panel of Fig. 1 illustrates what
happens if the price of curb parking is set high enough to
produce one vacancy for each eight spaces (a 12.5 percent
vacancy rate). There is no incentive to cruise because
drivers can find a vacant curb space anywhere they want to
park, or they can pay to park off-street, and the search time
is therefore zero. Now imagine that the price of curb
parking varies in response to parking demand to keep
roughly one out of every eight spaces vacant in the Village
throughout the day. Underpriced curb parking creates an
ARTICLE IN PRESS
Fig. 1. Curb parking prices and crusing.
(footnote continued)
of raising the price of curb parking is to reduce cruising, not to produce
curb vacancies, because drivers continue to take any curb spaces that
become available.
13
If raising the price of gasoline by 10 percent reduces the search time by
1 percent, it shaves 3.6 s off the 6 min (360 s) you were previously willing to
cruise.
14
Shoup (2005, Chapter 14).
D.C. Shoup / Transport Policy 13 (2006) 479–486484
astonishing amount of cruising, and right-priced curb
pricing can eliminate cruising.
9. Complications
The decision to cruise is far more complex than a simple
model can portray, of course, and I will suggest six
complications. First, the value of time savings is not
constant. Different people place different values on time
savings, and the same person may place different values on
saving time on different days, at different hours, and for
different trips. Even for a specific trip, the value of saving
time may increase as you cruise, because the likelihood that
you will arrive late at your destination increases the longer
you hunt for curb parking. You may therefore drive
around for a while and then, growing desperate, pay to
park off-street.
Second, you don’t know in advance how long it will take
to find a curb space. In a normal queue you can see how
many people are ahead of you and how fast the queue is
moving, and you can roughly estimate how long it will take
to be served. But to cruise is to wait in a queue of unknown
length, where the next person called to the window is
determined by lottery. You may find a curb space in the
next minute, or it may take half an hour. How long it will
take to find a curb space is therefore a random variable
with a distribution and an expected value that are both
unknown. Previous experience with cruising in the area
may give you an idea of what to expect, but you don’t
know how many other drivers are cruising, or how
frequently curb spaces turn over. Nevertheless, clues
gathered while driving around may alter your guess about
how long it will take to find a curb space. You may see that
other cars ahead of you also appear to be cruising, and this
information lowers the likelihood of finding a curb space
soon. You may therefore drive around for a while to size
up the probability of finding a curb space within a
reasonable time.
15
Third, curb and off-street parking are not perfect
substitutes for one another, and drivers do not choose
between them only on the basis of price. The walking
time from the parked car to the final destination and
back must also be considered. Off-street parking may be
less convenient than curb parking because the added
driving and walking time needed to access the higher
levels of an aboveground garage may be significant, and
underground garages may seem unsafe. W. Vickrey
explained that curb parking is in most cases more
convenient than off-street parking, and said that curb
parking prices would therefore have to be higher than the
prices for off-street parking to create curb vacancies and
discourage cruising.
16
The right price for curb parking is
thus the price that leads to a few vacant spaces, as
illustrated in the bottom panel of Fig. 1, and this price may
be higher than the price of adjacent curb parking.
Fourth, you may not know the cost of parking in all
locations, or how far you will have to walk to the final
destination. You may thus drive to your destination and
then begin looking around, or if the price of off-street
parking is low, you may simply park at your destination
without making any comparisons.
Fifth, where curb parking is limited by the length of stay,
this limitation will reduce cruising by those who want to
park for a long time. Time limits are difficult to enforce,
however, and surveys often show that more than half the
cars parked in zones with time-limited free parking either
violate the time limit or are in an illegal space. In Seattle,
Washington, for example, a study of on-street parking in
35 neighborhoods found that the average parking duration
in spaces with a one-hour time limit was 2.1 h.
17
Finally, there are options beyond the simple choice
between cruising and paying. You can also park in an
illegal curb space and risk getting a ticket. Or you can drive
to a nearby area where curb parking is readily available or
where off-street parking prices are lower, and then walk
farther to your destination. But despite these and other
complications, the basic lesson is the same: if cities charge
too little for curb parking, drivers will cruise.
Drivers do not use a mathematical model when deciding
whether to cruise, but all the model’s assumptions are
reasonable, and all its predictions are testable hypotheses.
A model cannot predict how everyone will behave, but it
does suggest how to behave if you want to be rational in
your own cruising. You can even test the model’s
predictions by referring to your own experience. Suppose
that you want to park at a site where curb parking is free,
but all the curb spaces are occupied. Off-street parking
costs $1 an hour, but is immediately available. Would you
be more likely to cruise if you have several passengers in
your car? If you are in a hurry to reach your destination? If
you intend to park for only a short time? If your answers to
these questions are no, the model correctly predicts your
choices.
10. Conclusion: an invitation to cruise
Where curb parking is underpriced and overcrowded,
some drivers search for a curb space rather than pay to
park off-street. Research throughout the last century
showed that cruising is common in congested traffic, and
a model of how drivers decide whether to cruise or to pay
predicts that charging the fair market price for curb
ARTICLE IN PRESS
15
Thompson and Richardson (1998) develop a model of parking search
behavior that explains how drivers choose among off-street parking
facilities. They consider parking choice as a search process in which drivers
make a number of linked decisions based on updated knowledge gained
from experience. Surprisingly, their model suggests that long-term
experience doesn’t necessarily lead to better parking choices.
16
Vickrey (1954, p. 64). Similarly, Topp (1993, p. 85) says, ‘‘The most
convenient parking spaces are the on-street spaces where parking fees are
also usually less than in parking garages. Those spaces—and even the
illegal ones—generate more search traffic and waiting cars than spaces in
parking garages.’’
17
Shoup (2005, 297).
D.C. Shoup / Transport Policy 13 (2006) 479–486 485
parking can eliminate cruising. City governments therefore
play a large part in choosing whether drivers cruise,
because they set the prices for curb parking. Cruising is
simply a driver’s individual response to the public pricing
policy. Underpriced curb parking is a perverse subsidy
because it encourages drivers to do something that harms
other people and may not even benefit the drivers
themselves. Cities must then throw good money after bad
by spending more to fix the congestion and pollution
problems they have created. Getting the price of curb
parking right will benefit everyone, and every city can do it.
Acknowledgments
I am grateful to the University of California Transporta-
tion Center, the Lincoln Institute of Land Policy, and the
United States Department of Transportation for financial
support. I am also grateful to Jeffrey Brown, Matthew
Dresden, Alexandra Evans, Amy Ford, Mason Gaffney,
Daniel Hess, Kevin Holliday, Hiroyuki Iseki, Stephen Ison,
David King, Douglas Kolozsvari, Christopher Locke,
Michael Manville, Anne McAulay, Eric Morris, Jeremy
Nelson, Todd Nelson, Paul Philley, Lisa Schweitzer, Pat
Shoup, Alexander Smith, Paul Sorenson, and two anon-
ymous referees for excellent editorial advice. This article is
adapted from The High Cost of Free Parking (Shoup,
2005).
References
Boston Transportation Department, 2001. Parking in Boston. City of
Boston, Boston Available at http://www.cityofboston.gov/accessBos-
ton/pdfs/parking.pdf [Accessed July 19, 2005].
Bus+Bahn, 1977. Kommentar u
¨berflu
¨ssig, p. 2.
Clark, P., 1993a. An Assessment of the Likely Impact of Changes in Short
Term Metered Parking Prices on Parkers in the Central City Area.
Town Planning Branch, Cape Town City Council, TP 608, Cape
Town, South Africa.
Clark, P., 1993b. Policies to Manage Parking in the Central City Area.
City Planning Department, TP 608/PC, Cape Town, South Africa.
Colliers International, 2003. North America CBD Parking Rate Survey,
The Parking Professional, September, pp. 27–31.
Falcocchio, J., Jose, D., Elena, P., 1995. An inquiry on the traffic
congestion impacts of parking and pricing policies in the Manhattan
CBD,’’ prepared for the New York City Department of Transporta-
tion Division of Parking and the University Transportation Research
Center, Region II. Polytechnic University of New York Transporta-
tion Training and Research Center, New York, February 1995.
Hensher, D., 2001. Modal diversion. In: Kenneth, B., David, H. (Eds.),
Handbook of Transport Systems and Traffic Control. Pergamon,
Amsterdam, pp. 107–123.
Hogentogler, C.A., Willis, E.A., Kelley, J.A., 1934. Intangible economics
of highway transportation. In: Proceedings of the Thirteenth Annual
Meeting of the Highway Research Board, Washington, DC, December
7–8, 1933, pp. 189–205.
Huber, M., 1962. Street travel as related to local parking. in: Proceedings
of the 41st Annual Meeting of the Highway Research Board,
pp. 333–352.
Inwood, J., 1966. Some effects of increased parking meter charges in
London, RRL Report No. 7, Harmondsworth, England: Road
Research Laboratory.
O’Malley, M., 1985.‘‘Cruising for parking in Harvard Square: a model to
evaluate city parking policies. John F. Kennedy School of Govern-
ment, Harvard University, June 17, 1985. Selected by the American
Planning Association’s Transportation Planning Division for the
national award as Best Student Paper for 1985.
Salomon, I., 1984. Toward a behavioural approach to city centre parking:
the case of Jerusalem’s CBD. Cities 3 (3), 200–208.
Saltzman, R., 1994. Three proposals for improving short-term on-street
parking. Socio-Economic Planning Sciences 28 (2), 85–100.
Shoup, D., 2005. The High Cost of Free Parking. Planners Press, Chicago.
Simpson, H., 1927. Downtown storage garages. The Annals CXXIII,
82–89.
Vickrey, W., 1954. The economizing of curb parking space. Traffic
Engineering, November, pp. 62–67. Later incorporated in testimony to
the Joint Committee on Washington, DC, Metropolitan Problems in
1959, which was republished in the Journal of Urban Economics 36,
1994, 42–65.
ARTICLE IN PRESS
D.C. Shoup / Transport Policy 13 (2006) 479–486486
