The Price Doesn’t Matter if You Don’t Have to Pay: Legal Exemption as an Obstacle to
Congestion Pricing
Michael Manville*
Lewis Center for Regional Policy Studies
Institute of Transportation Studies
Department of Urban Planning
UCLA
3250 School of Public Affairs
Box 951656
Los Angeles, California 90095-1656
[email protected]
[email protected]
(323) 333-0149
Jonathan Williams
Fehr and Peers Associates
*Corresponding Author
Abstract: Transportation analysts frequently recommend pricing as way to combat road
congestion. Market prices for curb parking are a particularly attractive way to implement this
approach, both because paying to park is a less alien idea than paying to drive and because the
travel involved in cruising for parking is almost entirely socially wasteful. However, the
literature on performance-priced parking has thus far ignored the possibility that pervasive
nonpayment—through rent-seeking, opportunism, or fraud—will dilute the efficacy of market
prices. In this paper we use an original survey of thousands of parking meters in Los Angeles,
California to show that at any given time almost 40 percent of vehicles parked at meters are both
not paying and not breaking any laws. We document, in other words, a high incidence of legal
nonpayment. We also show that across an entire day, vehicles that are legally nonpaying
consume more meter hours than vehicles that pay or that occupy spaces illegally. The
implications of this research are twofold. First, legal nonpayment costs local governments
significant amounts of revenue each year. Second, and more importantly, the high rate of legal
nonpayment threatens to undermine the effectiveness of policies built around market-prices for
curb parking. We suggest policy reforms for cities confronted with high levels of legal
nonpayment. Chief among these reforms is an end to the practice of granting free, time-unlimited
parking to vehicles displaying disabled placards. These laws are present in at least 23 states, and
pose a serious threat to the effectiveness of market-priced parking.
Keywords: Parking
Transportation Rent-seeking Pricing
1
It is now commonplace for transportation analysts to recommend pricing as way to
combat road congestion and its attendant externalities, such as lost time, air pollution and carbon
emissions (Arnott, 2005; Shoup 2005). Pricing motor vehicle travel, however, is logistically
complicated and politically contentious, which explains why pricing, though almost universally
endorsed by academics, remains rare in practice (Guiliano 1992; Levinson 2007). Obstacles to
congestion pricing include technological difficulties, resistance from drivers, concerns about
equity, mistrust between governments and drivers, and mistrust among governments themselves
(Levinson 2007; Schaller 2010; Manville and King 2010). While these obstacles are substantial,
most scholars assume that once pricing is implemented, it will be effective, and that its
effectiveness will bolster its popularity. There is some evidence to support this idea: road
congestion pricing in both London and Stockholm significantly reduced traffic congestion, and
tolling was more popular after implementation than prior to it (Santos 2008; Harsman and
Quigley 2010).
Researchers have for the most part neglected the possibility that pricing, once
implemented, will not be effective. Yet there is considerable reason to believe pricing could fail.
The effectiveness of congestion pricing rests heavily on the assumption that most or all users will
pay the charge. If that assumption doesn't hold—if nonpayment is pervasive—the benefits of
pricing will erode.
Why might some people not pay? Roads and parking spaces have long been unpriced (or
underpriced) goods. Many people benefit from this status quo, will resist efforts to change it, and
will look for ways to escape payment if pricing begins. These attempts at nonpayment could be
legal or illegal. Examples of illegal nonpayment abound in London, where the congestion toll
relies on cameras that photograph a vehicle's license plate as it crosses the tolling cordon. The
2
city then uses the license plate to identify and bill the driver. Santos (2008) discusses an increase
in license plate theft by drivers wishing to avoid the charge, and also mentions the black-market
availability of liquid crystal display license plates that can be activated from inside a vehicle.
When activated, the LCD plates obscure a vehicle's true license number and attach a fake number
in its place.
More mundane examples of illegal evasion include drivers who park at metered parking
spaces without paying. These drivers might arrive at spaces fully intending not to pay, perhaps
because they know the area is poorly policed, or that the fine for nonpayment is small, or
because they know their errand is quick and their probability of being caught is low (Becker
1993). Or they may be willing to pay but have no coins, and no place to find them.
Illegal evasion is a problem for congestion pricing, but probably a self-limiting one.
Stealing a license plate or buying an evasion device is a crime in itself (in many US states,
license plate theft is a felony), and the number of people willing to commit a serious crime, and
risk severe punishment, in order to avoid prices is probably small. Conversely, the number of
people willing to park without paying might be large, but the damage this behavior inflicts on
the pricing system might be small, because each illegal parking session is likely to be short. The
probability of punishment rises the longer a vehicle is illegally parked, so while parking without
paying deprives the local government of revenue, it may not impede turnover and vacancy.
Legal nonpayment, by contrast, is potentially more problematic. Legal nonpayment
occurs when drivers acquire privileges or credentials that exempt them from payment and also—
in the case of parking—from time limits. The London congestion charge exempts taxis, who
successfully lobbied to avoid the toll. (The taxi share of traffic in the charging zone rose from 16
3
percent prior to congestion tolling to 21 percent in 2008).1 London's charge system also gives a
90 percent discount to residents in the charge zone, and fully exempts anyone holding a
European Union disabled placard. Similarly, American diplomats in London consider
themselves exempt from the congestion charge, and drive into the cordon zone facing neither
payment nor punishment (Santos 2008). United Nations diplomats, by dint of their status,
needn't pay fines for parking violations they commit in New York City. In 1996 alone, diplomats
were responsible for over 160,000 parking violations in New York City. In 2002, when New York
won the right to revoke the diplomatic plates of almost 200 vehicles that had more than three
violations apiece, diplomatic parking violations fell by over 95 percent (Fisman and Miquel
2007). In 2010, New York City cracked down on what it called excessive use of parking permits
by government officials, and sought to revoke over 25,000 government parking credentials.
Unlike a driver who steals a license plate or buys a black-market device to evade
detection, people using a credential that exempts them from payment do not necessarily break
the law (they break the law if the credential is acquired fraudulently or misused). And unlike
drivers who park without paying, an individual with a credential needn't worry about using it for
long periods. Because the risk of acquiring and using the credential can be low, legal
nonpayment is likely to be more pervasive and more damaging to the pricing system than illegal
nonpayment.
In this article we examine the problem of legal nonpayment at parking meters in the City
of Los Angeles, California. Los Angeles is an appropriate place to examine this problem, for two
reasons. First, the LA urbanized area is the most traffic-congested region in the United States
(Texas Transportation Institute 2010), so any policy reform designed to reduce congestion—and
1
See www.tfl.gov.uk/corporate/about-tfl/publications/1482.aspx) and
(www.tfl.gov.uk/roadusers/congestioncharging/6722.aspx).
4
any obstacle to such a reform—is particularly relevant in Los Angeles. Second, the City of Los
Angeles, in partnership with the federal government, is currently embarking on an $18 million
pilot project in downtown LA to test the efficacy of congestion-priced street parking; the federal
government is in a similar partnership with the City of San Francisco. Legal nonpayment, if
sufficiently pervasive, could undermine the validity of this experiment.
We focus on three types of legal nonpayment: failed meters, the use of government
credentials, and—in particular—the use of disabled placards, which in California and many other
states allow drivers to park free for unlimited time. We find that legal nonpayment is widespread,
and that most drivers who occupy spaces without paying do so by displaying disabled placards.
The use of the placards results in low turnover at parking spaces and large amounts of lost
revenue for the City of Los Angeles.
In the next section of this article we briefly outline the advantages of market-priced
parking, then discuss the problem of legal nonpayment in more detail. Section III outlines our
research strategy, and Section IV presents our results. In Section V we offer policy
recommendations, the most important of which is a recommendation to end the payment
exemption for holders of disabled placards. As part of this discussion, we take up the inevitable
question of whether many placard users are committing fraud. Answering that conclusivelt is
beyond the scope of our article, but there is reason to believe fraud is widespread. However,
neither our analysis nor our argument for eliminating the exemption depend on fraud being
pervasive.. We present information from 2000 Census microdata suggesting that disabled
payment exemptions are poorly-targeted interventions, and that better ways exist to help disabled
motorists. Section VI concludes.
II. THE LOGIC OF PRICED PARKING
5
The classic form of congestion pricing is dynamic road tolling—prices placed on the
road that vary with demand, usually by time of day and day of the week. Road tolling is, in
economic terms, "first best," or ideal (Arnott 2005), because it directly targets the activity that
leads to congestion: driving in times and places where demand for scarce road space is high.
"First best" is different than "feasible", however, and road congestion tolls are still rare, largely
because of political resistance from elected leaders and the driver-voters t they answer to
(Guiliano, 1992).
Market prices for curb parking are an appealing alternative to road congestion pricing for
four reasons. First, although charging people to park is not perfectly equivalent to charging them
to drive, almost all vehicle trips end in a parking space, so the demand for parking will vary
closely with demand for driving. Thus market-priced parking can be a powerful and effective
way to reduce congestion. Second, the type of driving that priced parking eliminates is almost
entirely socially wasteful. One concern that surrounds road congestion pricing is that priced
roads will make it more difficult to engage in productive activities, such as going to work or
school. Whatever the merits of this argument, it is less valid when parking, rather than driving, is
priced. Market-priced parking eliminates cruising, and drivers cruising for parking are not on
their way anywhere; they have already arrived, and are simply driving around in an effort to
lower their parking costs. Third, paying for parking, while not common, is nevertheless more
common than paying to drive, and voters are accustomed to it even if they don't like it. Fourth,
cities can introduce paid parking incrementally and autonomously. Effectively implementing
road congestion pricing can require tolling an entire network, which can in turn require
cooperation across local governments, and permission from state and federal agencies. Cities,
however, already have the authority to charge for parking, and they can introduce parking
6
charges neighborhood by neighborhood.
These advantages of priced parking have been well elucidated by Shoup (2005) and
Verhoef (2005), among others. Shoup (2005) suggests that cities set a target occupancy rate of 85
percent for each block-side, and adjust the price to maintain that occupancy rate. The 85 percent
rate ensures that arriving vehicles always have a place to park, which reduces cruising. To gain
political approval, Shoup suggests using the resulting meter revenue to provide local services in
the areas where the meters are located. Note that both the economic and political aspects of this
proposal hinge on payment; the economic gains from reduced cruising are a result of drivers
having to pay, and so too are the political benefits that come from revenue-funded local public
services.
III. THE DILEMMA OF NONPAYMENT
Figure 1 shows the problem that motivates this article. Taken from sensors placed below
parking spaces on a block of Hollywood Boulevard, the figure shows consistently high
occupancy throughout the day—vacancy peaks at about 20 percent—but a consistently low level
of payment. At just after 1 p.m., for example, about 85 percent of parking spaces are occupied.
Eighty-five percent is Shoup's (2005) target occupancy rate, but it is doubtful that this rate has
much to do with the parking's price, since fewer than half of the occupied spaces are collecting
any revenue. At both 10 a.m. and noon there is no vacancy at all, which suggests prices should be
higher. But in both cases fewer than 50 percent of drivers have paid, suggesting that a higher
price would do little to create turnover.
7
$1 per hour
Vacant
Unpaid
Paid
Figure 1: Parking Occupancy and Payment on Hollywood Boulevard
Why might people be able to park without paying? Consider four reasons.
1. Inadequate Enforcement: A driver might park at a metered spot and gamble that she
will be done with her business before an enforcement official arrives and cites her. A
driver might also set out to deceive enforcement personnel, either by placing an
empty citation envelope on her windshield to make it appear as though she has
already been cited, or placing cones around her vehicle to make it seem like she is on
some official business.
2.
Meter failure: Many cities have old parking meters designed for an era when parking
was less expensive, often 10-25 cents an hour. If local governments increase the
hourly rate to encourage turnover, which may mean rates of $1 to $4 an hour, the
meters can quickly fill up with coins and fail. Older meters are also prone to
8
vandalism. Scofflaws can tamper with old meters by jamming them with paper clips
or spraying hardening foam into their coin slots. In many cities, a failed meter is a
free meter. In Los Angeles, for example, drivers arriving at a failed coin-only meter
park free. These drivers must still abide by any time limits, but in practice enforcing
time limits is difficult, particularly when the meter is failed.
3. Government Identification: State and local government officials, including police
officers and other civic officials, often have credentials that enable them to park free
and without time limits at curbside spaces. In some cities this exemption is a formal
rule; in others it is simply a common practice not to cite government vehicles.
4. Disabled Placards: In at least 23 states, and in a number of municipalities in other
states, vehicles displaying a disabled placard are allowed to park free, park without
time limits, or both. In California, a vehicle with a disabled placard gets free, timeunlimited parking at curbside parking spaces. Table I shows a list of states that grant
free parking to holders of disabled placards.
9
Table 1: Parking Privileges for Holders of Disabled Placards, by State
State
Benefit
Code
Alaska
Arkansas
California
Connecticut
Florida
Hawaii
Idaho
Illinois
Indiana
Kansas
Maryland
Massachusetts
Michigan
Minnesota
New Hampshire
North Dakota
Oregon
Rhode Island
South Carolina
Tennessee
Texas
Utah
Vermont
Virginia
Washington
Washington, DC
Wisconsin
Free parking, no time limit
Discretion of counties and cities
Free parking, no time limit
Free parking, no time limit
Free parking, no time limit
Free parking, 2.5 hours or meter maximum
Free parking, no time limit
Free parking, no time limit
Free parking, no time limit***
Free parking, 24 hour time limit
Free parking, up to four hours**
Free parking, no time limit
Free parking, no time limit
Free parking, no time limit
Free parking, no time limit
Free parking, no time limit
Free parking, no time limit
Free parking
Free parking, no time limit
Free parking
Free parking, no time limit
Free parking, for "reasonable periods of time"
Free parking, no time limit
Free parking, four hour time limit
Free parking
Free parking, twice the posted time limit
Free parking, no time limit*
28-10-181; 28-10-495
27-15-312
V C Section 22511.5
14-253 e
316.1964
§291-55
49-410
Sec. 11?1301.1.
IC 9-18-18-2
8-1, 126
§ 13-616
Ch40, Sec22; Sec22A
257.675
169.345
265:73
39-01-15
811.635
§ 31-28-4
56-3-1965
55-21-105.
681.006.
41-1a-414
§ 304a
46.2-1245
46.61.582
346.5
*Vehicles in 30-minute meters excluded.
**Applies only if meter space is not ADA compliant
***Disabled veterans only
The table shows the 23 states where we were able to identify parking privileges for drivers holding disabled
placards. It is possible that some of the remaining 27 states, or some municipalities within those states, also offer
privileges to placard holders. For instance, to the best of our knowledge New York State does not offer a payment
exemption to holders of disabled placards. New York City does, however; individuals certified for a placard in
New York City are allowed to park free without time limits (see
http://www.nyc.gov/html/dot/html/permits/pppdinfo.shtml). We list, for each state, the nature of the entitlement
and the relevant section of the state code.
Reasons (1) and (2) are serious but solvable. If the bulk of the nonpayment is the result of
10
inadequate enforcement, local governments can combat the problem through more and better
policing. Indeed, sensors embedded beneath parking spaces can alert authorities immediately if a
space is occupied but not paid for. If meter failure is the source of most nonpayment, the city
can upgrade its meters. Newer parking meters rarely break down, and when they do, they send a
signal to centralized controllers, who can dispatch a repair team quickly. Moreover, because new
meters are linked to a network, motorists who arrive at failed new meter are usually not allowed
to park free, but rather required to pay at another meter, or via their mobile phones. Upgraded
meters also usually accept debit and credit cards, which may also serve to reduce illegal parking,
if many people fail to pay because they lack coins.
The third and fourth reasons for legal nonpayment, however, are more difficult to address,
as they can be solved neither by technological improvement nor increased policing. (If a
credential is being used fraudulently, policing can help, but detecting fraudulent use of
credentials is time-consuming and difficult, as we discuss later).2 Disabled credentials are likely
a larger problem than government credentials, because disabled placards are easier to acquire.
The power to distribute government credentials is relatively centralized,3 and few people are
qualified to hold such credentials. Further, local governments have both the incentive and ability
to minimize the use of government credentials at local parking meters, because local revenue
will fall as the use of credentials rises.
By contrast, a broad set of conditions qualifies Californians for permanent or temporary
disabled parking credentials, and a wide variety of professionals, most of them not under the
2
The prevalence of legally exempt vehicles could also undermine other enforcement efforts. If enforcement officers
routinely respond to spaces where sensors indicate nonpayment and find that the vehicles are not in violation, they
may become progressively less likely to respond, and more scofflaws will go unpunished.
3
This was not the case in New York City, where multiple agencies could distribute credentials. The city's major
reform has been to consolidate credential distribution. See "City's Parking Permit Problem Worse than Thought,"
New York Sun, March 6, 2008. http://www.nysun.com/new-york/citys-parking-permits-problem-worse-thanthought/72414/
11
authority of a local government, are empowered to certify people for these credentials. In
California, doctors, midwives, nurse practitioners, optometrists and chiropractors are all able to
certify that individuals have impaired mobility or vision, and this certification, when presented to
the Department of Motor Vehicles (DMV), qualifies a person for a disabled placard. In practice,
people receive disabled credentials for both long- and short-term conditions, ranging from
serious permanent disabilities to sprained ankles and pregnancy.
For municipal officials, the placards pose a dilemma. The placard grants free and timeunlimited parking to its holder, so its value rises with the price of parking. In a world where all
parking is free and time-unlimited, the value of the disabled placard is low. As the price of
parking rises, however, so too does the value of the permit and the incentive to acquire one. The
cost of acquiring a permit, however, remains the same; a trip to see a doctor or other professional
who can certify an individual as impaired. We should therefore see a process where more people
use permits in the highest-priced areas. This process might involve premeditation, where people
set out to acquire permits as a reaction to high prices, or might be more subtle, as people react to
higher prices by using the placard of a friend or relative. It may also take place because disabled
people are less able to alter their behavior. Where able-bodied people might be able to avoid
prices by walking, taking transit or parking in less-accessible off-street spaces, disabled people
might find such adjustments more costly, and continue to use curb spaces as other drives
substitute away from them. In any event, as the share of exempt vehicles rises, turnover will fall,
and the price of parking—if it is responsive—will rise again. This in turn will increase the value
of the permit, and the cycle will start over. Taken to its logical conclusion, this cycle will
eventually result in placard holders completely driving out paying customers. The increased price
of parking—designed to create turnover in the most congested areas—will have accomplished
12
precisely the opposite.
The logic outlined above yields four hypotheses:
1. Both illegal and legal nonpayment will be problems in areas with priced parking, but legal
nonpayment will be more pervasive. A person who can avoid prices with impunity will consume
more of the priced good. Individuals who park illegally must avoid detection; individuals who
park legally without paying face no such constraint.
2. Drivers with disabled credentials will be more likely to use curb parking. If priced properly,
street parking should be a premium amenity. The convenience it affords drivers with quick
errands should give it a high per-hour rate relative to off-street parking. However, drivers
shielded from the price will be over-represented at curb spaces, the proportion of curb spaces
occupied by legally exempt vehicles will rise with the price of parking.
3. Vehicles with credentials exempting them from payment will park longer than vehicles whose
drivers have to pay, and longer than vehicles that park at failed meters. People without
credentials include not just paying motorists, but also motorists who are illegally avoiding
payment and motorists who park at failed meters. For people parked illegally, the risk of being
cited rises with the duration of the parking sessions. And a person who parks at a failed meter is
likely to stay for a shorter period of time than an individual with a credential for two reasons.
First, to the extent it is enforced, the time limit still applies to failed meters. Second, in most
cases finding a failed meter is a matter of chance. Because motorists don't know beforehand that
they will be able to park free—because they likely set out with an expectation of paying—they
may also have other destinations to reach.
4. The incidence of meter failure will also correlate with the price of parking. In areas with high
meter rates, meters will fail more often, both because they fill up with coins and because people
13
have a greater incentive to vandalize them. However, this hypothesis should not hold in places
with newer meters.
IV. DATA AND METHODS
We are not the first to note the prevalence of legal nonpayment in street parking.
Anecdotal evidence, as well as data gathered by consulting firms, suggests that disabled placards
are pervasive in downtown areas where parking is costly. A reporter for the Oakland North
newspaper found that on a random weekday 44 percent of the vehicles parked in downtown
Oakland were displaying disabled placards.4 In 2009 a study by Desman Associates examined
380 meters in Los Angeles and found that only 5 percent of the vehicles parked at meters
displayed disabled placards. However, those 5 percent of vehicles took up over 17 percent of the
available meter time.5 Similarly, a study of meter parking duration near hospitals in downtown
Seattle found that over 40 percent of drivers with placards parked for more than four hours in
what was otherwise a two hour zone6.
Parking is a function of both space and time. At any given time, parked cars occupy a
certain proportion of space, and in any given space, a parked car will consume a certain portion
of time. To measure both, we conducted two different kinds of surveys. The first type of survey,
designed to ascertain how many parking spaces are occupied but not paid for at any given time,
was cross-sectional. The City of Los Angeles has approximately 40,000 curb parking meters,
which are divided into 80 Parking Meter Zones (PMZs). The zones vary widely in size; we
4
5
6
See ―Disabled Placards in Downtown Oakland: Are they Legit?‖ By Anrica Deb. Oakland North, August 20,
2010. http://oaklandnorth.net/2010/08/20/disabled-parking-placards-in-downtown-oakland-are-the-legit/
See Desman Associates (2009). Financial Analysis and Condition Appraisal: Los Angeles Public Parking
System. Chicago, IL. Curiously, this study only examined "high-turnover" neighborhoods, and as a result it may
suffer from selection bias; credentials, after all, are likely to inhibit turnover.
See Heffron Transportation (2009). "First Hill Neighborhood On-Street Parking Study." Seattle, WA.
14
selected 13 of the largest zones and sent researchers, usually in pairs but occasionally alone, to
walk routes through these zones. The researchers recorded whether a parking space was
occupied, whether an occupied meter was paid or unpaid, and, if unpaid, the type of nonpayment
observed. The areas we chose had street parking rates ranging from $1 an hour (usually on the
west side of Los Angeles) to $4 an hour (in parts of downtown). Almost every space was
evaluated both morning and evening, and no surveys were conducted within 30 minutes of a towaway or no-stopping restriction. Drivers sitting in their vehicles (and not actively engaged in
loading or unloading) at unpaid meters were marked as unpaid, in accordance the California
vehicle code. The observations were grouped into block sides, and researchers also collected data
on various block amenities (recording, for instance, whether there was public off-street parking
on that block side).
Most of LA's parking meters are older, coin-operated single-space meters. In recent years,
however, the city has begun transitioning to new meters, including single-space meters that
accept credit cards and multi-space meters that allow drivers to pay at any station. Multi-space
meters also allow drivers to pay via mobile phones. Because the pay station meters permit
remote payment, they introduce some potential bias into the survey. Remote payments register
directly at a central computer at the city's Department of Transportation—if a person parks and
pays with his cell phone, the meter station nearest his space classifies his vehicle as unpaid.
Similarly, if a person parks and pays at a meter two blocks away, the pay station nearest his space
would incorrectly show his vehicle in violation. (Parking enforcement officers avoid this
dilemma by carrying handheld devices that communicate directly with the central parking
server.) Because our surveyors examined only the meter closest to any given parking space, it is
possible that we undercounted payment and overcounted illegal nonpayment. However, if it is
15
reasonable to assume that most drivers use the meter closest to their vehicle, then the bias
introduced by multi-space meters is probably low. As it was, we found relatively few spaces that
were unpaid and lacked an exempting credential.
In total, we assembled 11,987 observations of approximately 5,150 unique meters on
over 500 block sides. Since Los Angeles has roughly 40,000 parking meters, the survey covered
approximately 12 percent of all the city's meters. Almost every meter was observed a minimum
of two times. In light of LA's pilot project, we paid particular attention to the downtown (Little
Tokyo, Civic Center, and Chinatown neighborhoods). Table 2 shows the locations and counts of
our surveys; Appendix II shows these locations on a map.
Table 2: Parking Meter Observations by Areas Surveyed
Single-Space
Multi-Space
Total Meter Unique Hourly
Neighborhood
Meter Obseverations
Meter Observations Observations Spaces Meter Rate
Westwood Village
588
267
855
423
$1.00
Brentwood
328
0
328
167
$1.00
Westwood Boulevard
384
586
970
494
$1.00
Little Tokyo
1,794
0
1,794
407
$1-$1.50
Sawtelle
1,075
0
1,075
531
$1.00
Hollywood
1,202
227
1,429
753
$2.00
Chinatown
712
0
712
359
$2.00
Civic Center
627
508
1,135
304
$1-$4.00
Santa Monica Western
169
236
405
201
$1.00
Silver Lake
494
108
602
304
$1.00
Venice
477
440
917
225
$2.00
Studio City
371
286
657
424
$1-$1.50
Beverly-Fairfax
1,108
0
1,108
557
$1.00
Total
9,329
2,658
11,987
5,149
Block Side
Observations
106
29
124
216
188
215
96
145
18
64
104
30
216
1,551
The second type of survey, designed to ascertain how much time vehicles consume in a
given place, was a continuous observation survey. In these surveys, the research team observed
one block side of meters for the duration of the metering period, usually between eight and ten
hours. The surveyors observed all parking activity for that duration, recording the start and end
16
time of every parking session; the time and length of payment; and any visible reason for
nonpayment. We conducted five such surveys: Table 3 shows the times, locations, and other
details.
Table 3: Location and Time of Continous Observation Surveys
Date
Meters
Flower Street, Downtown Los Angeles
8-Mar-10
14
Hope Street, Downtown Los Angeles
25-Mar-10
11
Weyburn Avenue, Westwood Village
7-Apr-10
10
Fashion District, Downtown Los Angeles
29-Jun-10
9
Little Tokyo, Downtown Los Angeles
16-Jun-10
10
Meter Hours
Meter Type
8 a.m. - 6:30 p.m. Pay Station
8 a.m. - 6 p.m. Conventional
8 a.m. -8 p.m.
Conventional
8 a.m. - 6 p.m. Pay Station
8 a.m. - 8 p.m. IPS
Meter Rate ($/Hr)
4
4
1
3
2
Time Limit
2 hours
2 hours
2 hrs/15 min.*
1 hour
2 hours
*Eight meters had a two-hour time limit; two meters had a 15-minute time limit.
V. RESULTS
We used the walking surveys to test the idea that legal nonpayment would be more
pervasive than illegal nonpayment, that the use of disabled placards would increase with parking
meter rates, and that the incidence of failed meters would also rise with meter rates, but fall with
the presence of newer, multi-space pay stations.
Table 4 shows our results. Across all surveyed neighborhoods, we observed 61 percent
meter occupancy, which in the absence of any distortions would suggest that prices in many
places are too high (assuming target occupancy is 85 percent). Of the occupied spaces, however,
less than half (48 percent) were paid, and this ranged from a high of 70 percent along Santa
Monica Boulevard to a low of 30 percent in the Civic Center area of Downtown Los Angeles.
This high level of nonpayment constitutes a significant distortion, and at least suggests that if
more vehicles had to pay the price of parking would be lower. Payment correlates inversely with
the share of meters where disabled placards were observed, and also correlates inversely with the
presence of vehicles that simply failed to pay (―Percent Expired‖). Disabled placards were
observed at 27 percent of all meters, while illegal nonpayment was observed at 13 percent.
17
Table 4: Summary Results of Cross-Sectional Meter Surveys, by Neighborhood and in Total
All Spaces
Occupied Spaces
Total
Percent
Percent Percent Percent Percent Percent
Meters Occupied
Paid
Govt
Expired Disabled Failed
Westwood Village
Brentwood
Westwood Boulevard
Little Tokyo
Sawtelle
Hollywood
Chinatown
Civic Center
Santa Monica Western
Silver Lake
Venice
Studio City
Beverly-Fairfax
Total
Unpaid Spaces
Percent Percent Percent Percent
Govt Expired Disabled Failed
Percent
New Meters
855
328
970
1,794
1,075
1,429
712
1,135
405
602
917
657
1,108
86
67
48
61
48
53
59
75
89
43
67
62
56
51
58
37
57
38
47
37
30
70
39
62
66
37
3
1
1
4
5
2
9
7
1
1
1
1
1
9
12
14
13
11
17
11
15
12
19
17
16
8
29
31
28
22
32
27
35
44
18
12
19
12
27
11
0
21
5
14
8
8
5
1
30
2
6
27
6
3
1
8
8
4
14
10
1
2
1
3
2
19
29
22
30
18
31
18
21
28
31
46
45
13
53
68
43
50
51
49
56
62
39
18
48
34
42
22
0
33
12
23
15
12
7
1
49
4
18
43
31
100
60
57
0
18
0
73
58
18
48
55
0
11,987
61
48
3
13
27
10
6
25
50
19
37
"New Meters" refers to either multi-space pay stations or intelligent single-space parking meters
The final four columns of Table 4 examine the reasons for nonpayment. Government credentials
are an issue but a minor one, accounting for 6 percent of nonpayment overall and never climbing
above 14 percent in any neighborhood. Meter failure is a substantial problem, accounting for 19
percent of overall nonpayment overall and nearly half of nonpayment in some neighborhoods. As
the final column of the table shows, however, meter failure is less of an issue in zones where
newer, computerized meters are prevalent.
The most important contrast is between illegal nonpayment and the use of disabled
placards. Overall, disabled placards account for 50 percent of all nonpayment, twice the share
accounted for by meter scofflaws who simply park and don’t pay. There is ample room to
improve the illegal nonpayment problem: overall only 6 percent of the vehicles illegally parked
at expired meters had citations. Note too that our research method probably understates the
placard problem, because approximately 5 percent of the disabled placards we observed were at
spaces with time on the meter. We marked these as ―paid‖, but in reality a good portion of these
vehicles had pulled into spaces where the previous occupant had time left over.
The incidence of disabled placards appears to be weakly correlated with price, as the
18
placards are most common in Civic Center, parts of which have $4 an hour parking, and high
shares also occur in Hollywood, Venice, and Chinatown ($2 an hour). On the other hand, placard
use is also quite high in Westwood Village, where parking rates are only $1 an hour. Other
factors, such as the presence of elderly residents in the area, or the presence of medical facilities
(such as UCLA’s large hospital complex) no doubt contribute to the prevalence of placards.
Because our surveys were carried out iteratively over the course of months, it is possible
that any relationships we observe between placard use and price (or meter failure and price) are
spurious, and actually the result of events happening in those times and places that we did not
observe. To partially account for this possibility, we estimated three regressions that hold these
effects constant; the results suggest that that the relationships between meter rates and legal
nonpayment are not artifacts of unobserved place and time effects; a one dollar increase in the
price of parking (starting from a $1 base—there is no unpriced parking in the sample) is
associated with a roughly 5 percentage-point increase in the share of spaces showing disabled
placards. We present these regressions in Appendix I.
Continuous Observation Surveys
We hypothesized that individuals with credentials exempting them from meter rates
would park on the street for longer durations than individuals who have to pay for their time, and
that the share of time consumed by vehicles with credentials (as opposed to the share of space
consumed, which we just measured) would rise with meter rates. Our results lend credence to
these hypotheses. Table 5 shows the share of time consumed by occupancy for each
neighborhood, and overall. In every survey disabled placards consumed the most unpaid time,
and in all observations but one (Little Tokyo) they consumed the most occupied time. Even on
19
Hope Street—which was an outlier in that 8 of the 11 meters were broken on the day of the
survey—disabled placards accounted for the largest share of meter time.
Table 5: Percent of Time Consumed by Type of Occupancy and Neighborhood
Flower Street
Hope Street
Fashion District
Little Tokyo
Weyburn
Percent
Percent
Occupied
95
66
91
94
90
Paid
8
3
21
43
62
Percent
Percent
Percent Percent
Disabled Government Expired
81
4
1
25
9
6
38
10
22
28
21
2
21
0
7
Failed
0
23
0
0
0
Rate
($)
4
4
3
2
1
Setting Hope Street aside, the disabled share does vary positively with meter rate;
disabled placards consumed an astonishing 81 percent of the meter time on Flower Street, where
the rate was $4 an hour. Note too that, as we hypothesized, the amount of time consumed by
illegal parkers was in most cases quite low. In the cross-sectional analysis above we showed that
disabled placards occupied twice as many spaces as illegally parked cars. The continuous
observation surveys show that placard-carrying vehicles (outside of the Fashion District)
consume well more than twice as much time; ranging from a factor of three to a factor of 80.
The source of these massive differences is not that a large number of drivers hold
disabled placards. Rather those drivers with placards stay parked much longer. Table 6 shows the
average parking session length for regular (paid or unpaid), disabled, and government vehicles.
Because so many of Hope Street’s meters were failed, we examine its working and broken meter
spaces separately. Overall, disabled placards account for only 12 percent of all parking sessions,
and never more than 25 percent in any survey. But whereas the average ―regular‖ parker stays for
32 minutes, the average vehicle with a disabled placard stays for 229 minutes. (The average
20
vehicle with a government credential stays for 121 minutes.) A number of these averages are
biased upward by single extraordinarily long parking sessions—suggesting the potential damage
that can be done to pricing systems when even one vehicle is exempt from payment.
Table 6: Average Length of Parking Session by Type of Occupancy
Regular Average
Longest
Area
Sessions Length (Min.) Session (Min.)
Flower Street
56
21
159
Hope Street (Operating Meters)
25
26
184
Hope Street (Failed Meters)
70
20
93
Hope Street (All Meters)
95
22
184
Little Tokyo
70
47
175
Weyburn
128
38
181
Fashion District
80
28
600
Disabled Average Longest
Sessions Length Session
19
383
621
3
173
426
6
195
505
9
188
505
8
247
614
13
117
549
14
138
600
All Observations
429
32
600
63
229
621
Notes: "Regular" sessions include both paid and unpaid meter sessions that involved no payment-exempting credentials.
Govt.
Average Longest
Sessions Length Session
1
2
2
1
94
94
6
79
163
7
81
163
11
140
175
0
0
0
3
181
517
22
121
Disabled Pct
of Sessions
25
10
7
8
9
9
14
517
12
Table 6 also shows that, as we hypothesized, individuals parked at failed meters do not
stay longer than other non-credentialed vehicles. The average length of a parking session at a
failed meter on Hope Street was 20 minutes—shorter than the average session at a functioning
Avg Parking Duration
400
350
Non-Credentialed
Disabled Placard
Holder
Government
Placard-Holder
300
Minutes
250
200
150
100
50
0
Flow er St.
Hope St. (op)
Hope St. (all)
Westw ood
Hope St. (failed)
Little Tokyo
Fashion District
meter (26 minutes), and much shorter than the average session for a vehicle on Hope Street with
a disabled placard (188 minutes).
Lost Revenue: Nominal and Effective Meter Rates
21
The primary value of market-priced parking is to efficiently allocate parking spaces; the
social benefits arise from the revenue being collected, not from its being spent. Nevertheless,
cities rely on meter revenue to finance public services, and when motorists don’t pay posted
prices, the city collects less revenue. Hence one way to illustrate legal nonpayment’s impact on
pricing is to examine revenue collections. Table 7 shows the potential and actual daily revenue
from each of the areas we continuously observed. ―Potential revenue‖ assumes 85 percent
occupancy at current meter rates for all hours of meter operation (the issue of whether this rate is
correct rate will be discussed later); ―actual revenue‖ is the amount of revenue actually collected.
Because we monitored the meters minute-by-minute, we are able to account for driver
overpayment, and the fact that some vacant spaces generated revenue (i.e., when a driver paid for
fifteen minutes but left after ten). The nominal meter rate is the posted rate, ranging from $1 to
$4 an hour.
Recall from Table 6 that occupancy rates in these surveys ranged from 66 to 95 percent.
Table 7 shows that despite such high occupancy, revenue collection was small. Flower Street was
95 percent occupied, yet collected less than one-tenth the revenue it should have collected if it
was 85 percent occupied. All told, meters in our observation collected between 4 and 77 percent
of their potential revenue. In all cases but Hope Street, where vacancy and meter failure were
pervasive, the bulk of the revenue losses were a result of legal nonpayment (disabled parking and
to a lesser extent government parking). Revenue losses from scofflaws (―unpaid expired‖) were
much smaller than losses from legal nonpayment. The meters on Flower and Hope Streets, which
have nominal meter rates of $4 an hour, actually collected an average of 28 and 14 cents an hour,
respectively.
22
Table 7 : Potential and Actual Revenue from Observed Parking Meters
Number of Number Nominal Meter Potential Unpaid Unpaid
Unpaid
Unpaid
Location
Meters
of Hours
Rate ($/hr) Revenue Vacant Disabled Government Expired
Flower Street
Hope Street
Little Tokyo
Westwood
Fashion District
14
11
10
10
9
12
12
12
10
10
$4.00
$4.00
$2.00
$1.00
$3.00
$499.80 $31.60 $478.60
$374.00 $153.00 $112.00
$204.00
$9.85 $65.64
$102.00
$8.34 $24.73
$229.50 $21.80 $101.75
$0.00
$38.00
$45.44
$0.00
$27.25
$8.53
$29.00
$5.26
$6.94
$59.00
Unpaid Actual Percent of Effective Meter
Failed Revenue Potential Rate ($/hr)
$0.00 $47.44
$93.00 $15.33
$0.00 $113.80
$0.00 $78.68
$0.00 $60.65
9.5
4.1
55.8
77.1
26.4
Notes: Calculations are taken from minute-by-minute observations of parking meters. "Actual revenue"
includes revenue from all individual parking sessions, as well as overpayment by individual drivers (vacant spaces collecting revenue).
"Potential Revenue" is based on a target of 85 percent paid occupancy throughout the day.As such, columns do not
sum to "potential revenue" but rather sum to the total if all parking was always occupied.
Flower Street’s meters operate six days a week; in a well-functioning pricing system this
single block-side would collect $156,000. At the observed rate of payment, however, it would
collect less than $15,000—a loss of over $140,000 a week for the city. The problem with this
calculation, of course, is that it assumes that $4 an hour is the correct price for parking on this
block side. It may well be that a rate of $3, or $5, is the price that yields a consistent 85 percent
occupancy rate. Yet we have no way of knowing what the correct price is, because so many
drivers don’t have to pay. This, again, is the real problem posed by legal nonpayment. The price
signal, which is the backbone of all the efficiencies inherent in market-based pricing, is
completely obscured.
VI. POLICY RECOMMENDATIONS AND THE PROBLEM OF FRAUD
Anecdotal and journalistic accounts of placard use almost invariably raise the possibility
that many holders of placards are using them fraudulently. Figure 2 shows the growth in disabled
placards in California since 1990. In 1990 just over 2 percent of California residents were issued
placards, by 2009 the share had more than tripled, to almost 7 percent. In 2000, the last year for
which reliable data are available, 20 percent of the population of Los Angeles County reported
some type of disability.7 Our evidence suggests that on average disabled placards occupied 27
7
This information comes from the 2000 US Census, Summary Tape File 3. Evidence on disability is also
available from the Survey on Income and Program Participation (SIPP), but not at the county level.
23
$0.28
$0.14
$0.95
$0.79
$0.67
percent of the street parking spaces, and in some high-value areas consumed over 40 percent of
the spaces. We also showed a large and statistically significant relationship between placard use
and the price of parking. Neither the growth in placards nor the over-representation of placards
in high-priced areas, however, are necessarily indicate of fraud. As the population ages, we
should expect a higher prevalence of disability and therefore more disabled placards. Similarly,
more disabled people may be legitimately taking advantage of the entitlement, either because
they are more aware of it, because the price of parking has risen, or because newer technologies
make it easier for disabled people to drive. And the high proportion of disabled placards at curb
spaces could simply indicate, as mentioned before, a more inelastic demand for curb parking
among people with disabilities.
Figure 2: Placard Growth in California, 1990-2009
. All that said, anecdotal and journalistic evidence suggest that placard fraud is rampant.
24
Stakeouts conducted by a local news affiliate in Los Angeles indicate that many placard users
have acquired their credentials illegally, either through a black market purchase or (more
commonly) by using the placard of an older infirm relative.8 Our own surveyors repeatedly
witnessed what appeared to be fraud.9 Placard fraud also appears to be common in the United
Kingdom.10
However, fraud is difficult to identify, because not all disabilities are readily evident.
People who appear able-bodied may nevertheless have a serious impairment.11 The issue is
further complicated because some behaviors that a majority of people might consider
inappropriate may not legally rise to the level of fraud. A person who uses a disabled placard
assigned to someone else, or who use a temporary placard after it expires, is committing fraud.
But an individual assigned a placard for a temporary condition—say, a sprained ankle—and who
continues to use the credential after the condition has healed but before the placard has expired,
is in a more legally ambiguous position For that matter, fraud, if present, may occur in a medical
office rather than on the street: patients might persuade medical professionals to certify them for
placards even if their condition does not warrant it. An infamous case at UCLA illustrates this
problem: in 1999 members of the UCLA football team provided false information to doctors and
8
See Grover and Goldberg 2010, and Goldberg 2010. In a personal communication, Grover told us of
interviewing an able-bodied downtown merchant who used a placard to reduce the cost of doing business. His
immediate family, who all worked with him, used placards as well. They secured the placards through an elderly
relative, who reported her placard stolen four times. The Department of Motor Vehicles sent her a replacement
each time, and she gave it to a different family member. Also see Shin (2010), who reports that thieves in
Washington DC break into cars and steal placards for resale.
9
For example, in two instances seemingly able-bodied office workers parked and paid for 15 minutes, entered an
office building and then returned with a placard, which they hung in the vehicle for the remainder of the work
day. Surveyors also watched a man hang a placard in a van, load a dolly with heavy boxes from the van, and then
bounce the dolly down a flight of stairs into a subterranean food court. He remained parked for over 10 hours.
10
. A series of investigations in the United Kingdom led British officials to claim in September 2010 that over half
of the nation’s 2.5 million disabled placards were being used fraudulently. See
http://www.dailymail.co.uk/news/article-1313428/Fraudsters-cost-taxpayers-14-7-million-using-unauthoriseddisabled-parking-badges.html
11
Shin (2010) reports that the possibility of a difficult-to-observe disability makes police officers in Washington,
DC reluctant to confront people they suspect of being placard frauds.
25
received legal disabled placards. The players were eventually caught, but detecting fraud of this
sort is extremely time-intensive, as it might involve wrongdoing by both drivers and medical
professionals.12
Combating fraud is made still more difficult because placards are assigned to people
rather than vehicles. Enforcement officers must therefore confront placard users at the moment
they arrive or depart. Thus where conventional meter violators become easier to identify and
punish the longer they stay parked, placard frauds become more difficult to catch, because the
effort consumes so much time. As our continuous observation surveys showed, many placard
users stay parked for extremely long periods.13 Further, on-street enforcement would do little to
stem inappropriate behavior by medical professionals, or people who lie to medical
professionals. And efforts to regulate placards at the level of the medical professional—to
essentially have the state appropriate some of the autonomy of doctors and other qualified
providers—would almost certainly fail politically and would in any case most likely be
counterproductive. State transportation officials have neither the time nor the expertise to
determine who does or does not have a disability.14
12
13
14
Some journalistic accounts suggest that this practice is common. Assemblywoman Joan Quigley, a state
representative who wrote New Jersey’s law granting parking privileges to the disabled, said that in Jersey City,
"It became a bit of a scam. It looked like every doctor would write a letter saying this person or that person
shouldn’t have to walk." See
http://blog.nj.com/njv_paul_mulshine/2010/01/handicapped_parking_permits_ne.html. Similarly, in 2004 Barry
Siegel, a Los Angeles County Commissioner on Disability, said that ―all people have to do is lie a little bit to
their doctor and they can get a placard.'' See
http://www.thefreelibrary.com/DISABLED+LOSING+OUT+ABUSE+OF+PARKING+PLACARDS+HURTIN
G...-a0114762056
Two examples: In August of 2010 agents from the California DMV caught 18 fraudulent placard users in Los
Angeles. The stakeout lasted six hours and involved multiple investigators. See
http://www.contracostatimes.com/california/ci_15883262?nclick_check=1. Similarly, Seattle has cracked down
on placard fraud, but a Seattle Times article
(http://seatlletimes.newsource.com.html/localnews/2012735051_parkingside27m.html) suggests the tremendous
labor costs involved. The enforcement officer profiled in the article drives an unmarked car, and—because she
needs to apprehend violators as they slide behind the wheel—often hides for extended periods of time in hedges
near vehicles displaying placards. She has tracked some violators for months.
Although New York City attempts such a program, by choosing the physicians allowed to certify people as
disabled. ―People who live in New York City as well as non-City residents are eligible for a New York City permit if
26
We make two points. First, the incidence of fraud is immaterial to the problem posed by
the placards. In 2009 Los Angeles County had about 650,000 valid temporary and permanent
placards. The City of Los Angeles constitutes about 40 percent of the population of Los Angeles
County. If placards were distributed evenly countywide, 40 percent of these placards would be
for residents in LA, meaning that the City of Los Angeles would have more than 6 valid placards
for each of its 40,000 parking meters. Given the long duration placard-users stay parked, this
quantity is more than enough to undermine any attempt at market-priced parking, regardless of
whether the placards are legitimate.
Second, assuming some fraud does exist, the difficulty of enforcement suggests that the
best option is to change the law itself. Eliminating the payment exemption for placard holders
would eliminate the demand for placards by all but people with legitimately disabilities.. The
time-limit exemption should be left in place, for two reasons. First, disabled people may need
more time to carry out their tasks. Second, all parking should migrate away from time limits, as
time limits are prima facie evidence that the prices are wrong. Eliminating time-limits for
disabled people can be a stepping-stone to eliminating them overall.
The natural objection to removing the payment exemption is that doing so might harm
disabled people. If the law delivers important assistance to disabled people, and if legitimately
disabled people account for a large share of placard users, then removing the exemption might
impose large costs on a disadvantaged group. Fully exploring this question is beyond the scope
of this article, and as we have noted we cannot determine the precise incidence of fraud.
However, assume there is no fraud at all. Removing the payment exemption could harm
certified by a New York City Health and Hospitals Corporation (HHC) physician or other Department of
Health designated physician as having a disability which: is permanent; severely affects their ability to walk; and
requires the use of a private automobile for transportation.‖ See
http://www.nyc.gov/html/dot/html/permits/pppdinfo.shtml.
27
disabled people who are physically unable to pay, and those who are financially unable to pay. A
person physically unable to operate a parking meter is probably severely disabled. Nationally,
about two percent of disabled people never leave the home, and almost 60 percent of the
homebound are severely disabled (Bureau of Transportation Statistics, 2002). Of the severely
disabled who do leave the home regularly, many cannot drive themselves. About 4 percent of the
disabled report needing assistance from another person when they are outside the home and rely
on relatives or paratransit services to help them get around (Bureau of Transportation Statistics
2002). For these people, a driver or attendant would presumably be able to feed the meter, just as
he or she would also pay for gas, for parking in a private lot, and so on. Thus the law would harm
those individuals whose disability is severe enough to preclude operating a meter, but not severe
enough to preclude solo driving. While almost certainly some individuals fall into this category,
they may not be numerous enough to warrant a blanket payment exemption for all the disabled.15
Similarly, a law requiring holders of disabled placards to pay for parking would harm
those disabled individuals too poor to pay for parking, but not so poor that they have no access to
a vehicle. Justifying a payment exemption to disabled people in general because some disabled
people are poor requires that three conditions be satisfied. First, justifying a program for the poor
disabled, as opposed to the poor in general, requires some evidence that the disability is a cause
rather than a mere correlate of the poverty, and second requires evidence and that being poor as a
result of being disabled entitles a person to income supports that other poor people don't deserve.
Third, delivering a benefit to all people with disabled placards is an inefficient way to help
disabled people with low incomes, so the payment exemption makes sense only if there is no
15
Williams (2010) points out that California’s disabled parking law was originally written shortly after the Korean
War, and only exempted from payment veterans who had lost, or lost the use of, their hands, and who physically
could not operate a meter. Over time the scope of the law expanded to include people with a much broader set of
impairments. A natural question is how veterans who have lost use of their hands could drive; the legislative
record from this law offers no answers.
28
better way to target benefits.
The first two conditions are easier to satisfy than the third. The causality between
poverty and disability runs in two directions, but there appears to be little question that disability
increases the odds of being poor (Brault 2005; Lustig and Strauser 2007; Fremstad 2009; She
and Livermore 2009). Disability can reduce earnings because it physically precludes regular
employment, because disabled people might suffer labor market discrimination, or because
medical expenses can lead to bankruptcy. Conversely, being poor can increase the odds of
disability, if stress and poor nutrition lead to deteriorating health and insufficient medical care
causes minor conditions to become debilitating. And there is some reason to believe that poor
disabled people warrant more income support than poor people who are not disabled: as Sen
(2009:254) notes, disabilities ―make it harder to convert income into capability.‖
The third part of the case, however, is harder to make. A parking payment exemption is at
once over- and under-inclusive. Table 8 presents tabulations from the PUMS that make some of
these relationships more explicit. About 80 percent of the disabled live in a household with at
least one vehicle available, compared to 90 percent of the non-disabled. Just over 20 percent of
the disabled are poor, compared to 13 percent of the non-disabled (19 percent in LA County).
However, only half of those who are disabled and poor have access to a vehicle, and only about
ten percent rely on a private vehicle to get to work. Perhaps more importantly, only about 15
percent of disabled car owners are poor, and only 12 percent of the poor are both disabled and
have access to a vehicle.
29
Table 8: Transportation Characteristics of the Disabled in the USA and Los Angeles County, 2000
USA
Los Angeles
Percent of Disabled in a Household with at least One Vehicle Available
80
79
Percent of Non-Disabled with at least one Vehicle Available
92
90
Percent of Disabled in Poverty
21
23
Percent of Non-Disabled in Poverty
13
19
Percent of Poor Disabled with at least one Vehicle Available
49
54
Percent of Poor Disabled Commuting by Private Vehicle
11
12
Percent of Poor who are Disabled and have Vehicle Available
14
12
Percent of Disabled who are Poor and have Vehicle Available
13
16
N
14,076,739
471,162
Source: 2000 U.S. Census PUMS 5-Percent Sample. Calculations for USA are probability-weighted.
Calculations for Los Angeles County are unweighted.
The payment exemption therefore provides no benefit to the most disabled (those who are
shut-ins or cannot use automobiles) and delivers an unnecessary benefit to the majority of
disabled people who are not poor. The law also provides no benefits to those disabled people
who are most poor (and likely don’t own vehicles)16 and no benefit to the majority of the poor
who aren’t disabled. In and of itself this does not mean the exemption should end; perhaps
ideally laws intended to help the poor should help all the poor, and laws intended to help the
disabled should help all the disabled. In practice, however, policy interventions rarely meet such
abstract criteria.
Nevertheless, if we want to deliver benefits to poor disabled people, there appears to
ample room to improve our targeting. Even assuming a complete absence of fraud, a law
exempting placard holders from payment appears to offer legitimate help to only a small group
of people, and that help is not delivered to the worst-off. These benefits need to be weighed
against the law’s costs. The law substantially impedes efforts to fight congestion and localized air
16
There is relatively little published evidence about the interaction between poverty, vehicle availability and
disability. In Appendix II we present a regression using IPUMS data suggesting that the odds of a disabled
person having access to a vehicle are about 60 percent smaller than the odds a non-disabled person will.
30
pollution (whose burden often falls disproportionately on the poor),17 and to raise revenue. There
is no clear rationale for allowing disabled people to impede vehicular turnover and impose
pollution and congestion costs on others, or to impose costs on those who would benefit from
public services financed by meter revenue.
One approach might be to remove the payment exemption and dedicate some of the
increased revenue to programs that improve access for disabled people. These programs might
include better paratransit service, more curb cuts, or repairs that bring sidewalks into compliance
with the Americans with Disabilities Act—allowing people with impaired mobility to move
around more easily (Shoup 2010). Such improvements would benefit all residents, but disabled
people in particular. Similarly, if market-priced parking ensures the continuous availability of
one or two spaces on every block, it could benefit disabled drivers who need parking spaces
close to their destination.
VII. CONCLUSIONS
We have shown that legal nonpayment can be a substantial obstacle to congestion-priced
parking. Our data is confined to Los Angeles but the dynamic we observe in LA may well apply
to many large cities. Exemptions for drivers who hold disabled placards can prevent pricing
from sufficiently increasing turnover and reducing congestion.
Our surveys revealed a number of obstacles to market-priced parking, but most of these
problems can be easily solved. Broken meters can be replaced. New meters do not fail nearly as
much as old meters, and pricing programs should be accompanied by meter upgrades. Similarly,
the problem of illegal nonpayment, while not inconsequential, can be combated with better
17
See for example Houston et al (2004).
31
enforcement.18 And illegal nonpayment will likely regulate itself to some extent, because people
who break the law to park without paying will do so for short durations to avoid being caught.
Legal nonpayment, by contrast, is harder to manage, and the fraud that often surrounds it
is difficult to detect. For this reason states should consider ending the practice of granting free
parking to holders of disabled placards. Politically this reform will be hard. Entitlements are
difficult to remove, and undoing an entitlement designed to benefit the disabled, in order to more
effectively charge drivers to park, is doubtless a tall political order. Federalism compounds this
challenge; most of these statutes are state laws that impose local costs. The best approach to
reforming these laws may be to focus on the evidence of massive fraud that the laws have
engendered. Fraud is not necessary to make the economic case against these laws, but it will be
useful in making the political case for changing them.
One cost of these laws, which we have thus far downplayed, is the cloud of suspicion
they throw over legitimate users of disabled placards. When a credential enables the evasion of a
price, widespread fraudulent use of the credential dilutes both price mechanism and the original
meaning of the credential itself. In any event, laws that grant free parking to disabled people
help neither most of the disabled nor the most disabled; they also help neither most of the poor
nor the poorest. More importantly, the externalities of this poorly targeted subsidy threaten to
undermine a transportation reform that will deliver large benefits to all citizens. Cities moving
toward market-priced parking should also move to eliminate laws the enable legal nonpayment.
ACKOWLEDGMENTS
We owe thanks to the University of California Multi-Campus Research Initiative on
Sustainability for funding, Ariel Strauss and Donald Shoup for helpful comments, and most of all
18
Enforcement is time-consuming in a city as large as Los Angeles, but see Kleiman (2009) for a discussion of
―dynamic concentration‖ of enforcement personnel. This approach could help crack down on illegal
nonpayment.
32
to the many graduate students who helped conduct the surveys.
REFERENCES
Arnott, R. 2005. City Tolls – One Element of a Successful Policy Cocktail. CESifo DICE Report,
Ifo Institute for Economic Research at the University of Munich, 3(3): 5-11.
Becker, Gary. 1993. Nobel Lecture: The Economic Way of Looking at Behavior. Journal of
Political Economy. 101(3): 385-409.
Brault, Matthew. 2008. Americans With Disabilities: 2005. Current Population Reports, P70-117,
Washington, DC.: U.S. Census Bureau.
Bureau of Transportation Statistics. 2002. Freedom to Travel. BTS Report S03-08.
Fisman, R. and Miquel E. 2007. Corruption, Norms and Legal Enforcement: Evidence from
Diplomatic Parking Tickets. Journal of Political Economy. 115(6): 1020-1048.
Fremstad, Shawn. 2009. Half in Ten: Why Taking Disability into Account is Essential to
Reducing Income Poverty and Expanfing Economic Inclusion. Washington, DC: Center for
Economic and Policy Research.
Giuliano, G. 1992. An Assessment of the Political Acceptability of Congestion Pricing.
Transportation. 19(4):335-358.
Grover, Joel and Matt Goldberg. 2010a. No Place to Park. March 3, KNBC LA.
http://www.nbclosangeles.com/news/local-beat/No-Place-to-Park-85105762.html
Grover, Joel. 2010b. Parking Peril. http://www.nbclosangeles.com/news/localbeat/Parking_Peril_Los_Angeles.html.
Harsman, B., and Quigley, J. 2010 (forthcoming). Political and Public Acceptability of
Congestion Pricing. Journal of Policy Analysis and Management.
Houston, Douglas, Jun Wu, Paul Ong, and Arthur Winer. 2004. Structural Disparities of
Urban Traffic in Southern California: Implications for Vehicle-Related Air Pollution
Exposure in Minority and High-Poverty Neighborhoods. Journal of Urban Affairs.
26(5):565-592.
Kleiman, Mark. 2009. When Brute Force Fails.Princeton, NJ: Princeton University Press.
Levinson, D. and Odlyzcko A. 2008. Too Expensive to Meter: The Influence of Transactions
33
Costs in Transportation and Communication. Philosophical Transactions.366(1872): 2033-2046.
Lustig, Daniel and David R. Strauser. 2007. Causal Relationships between Poverty and
Disability. Rehabilitation Counseling Bulletin. 50(4):194-202.
Santos, G. 2008. London Congestion Charging. Brookings-Wharton Papers on Urban Affairs:
177-234.
Schaller, B. 2010. New York City’s Congestion Pricing Experience and Implications for Road
Pricing Acceptance in the United States. Transport Policy.17(4): 266-273.
Sen, Amartya. 2009. The Idea of Justice. Cambridge, MA: Harvard Belknap.
She, Peiyun and Gina A. Livermore. 2009. Long Term Poverty and Disability Among WorkingAge Adults. Journal of Disability Policy Studies. 19 (4): 244-256
Shin, Anys. 2010. Misuse of Handicapped Parking Placards is Widespread. Washington Post.
September 26, 2010. http://www.washingtonpost.com/wpdyn/content/article/2010/09/26/AR2010092603823.html?sid=ST2010091905153
Shoup, D. 2005. The High Cost of Free Parking. Chicago. Planner’s Press.
_______. 2010. Putting Cities Back on their Feet. Journal of Urban Planning and Development.
September: 225-233.
Texas Transportation Institute. 2010. Urban Mobility Report. College Station, TX.
Verhoef, E., P. Niljkamp, et al. 1995. The Economics of Regulatory Parking Policies.
Transportation Research A. 29(2): 141-156.
34
APPENDIX I: LOCATION OF SURVEYED NEIGHBORHOODS
35
36
APPENDIX II: OLS Estimates of Influence of Parking Price on Disabled Placard Prevalence
Table A1: OLS Estimates: Determinants of Legal Nonpayment
Percent
Percent
Variable
Disabled Placards Disabled Placards
Parking Rate Per Hour
0.045***
(0.01)
-0.002*
0
0
-0.003
(0.02)
Total Spaces
Public Off-Street Lots
Price of Parking Over $2 (Binary)
-0.002*
0
0
-0.006
(0.02)
0.151***
(0.03)
Percent New Meters
Constant
Time Fixed Effects
Day Fixed Effects
Month Fixed Effects
Neighborhood Fixed Effects
R2
N
* p<0.05, ** p<0.01,
***
0.255***
(0.03)
0.322***
(0.03)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
0.03
1,294
0.04
1,294
p<0.001
Dependent variable is the share of vehicles on a block side displaying a disable placard.
Robust standard errors in parentheses. All regressions have fixed effects to control for
day of week, month of year, time of day, and neighborhood. The regression suggests that
a one-dollar increase in the parking rate is associated with a five percentage-point
increase in the share of vehicles on a block side that display a disabled placard.
37
APPENDIX II: Poverty, Disability and Vehicle Availability
Table A2: Logit Regressions Examining Poverty, Disability and Vehicle
Availability in the US and Los Angeles County, 2000
USA
LA County
Disabled
-0.642***
-0.631***
(0.00)
(0.00)
Poverty
-2.183***
-1.765***
(0.00)
(0.00)
Male
-0.008***
0.005***
(0.00)
(0.00)
Black
-1.001***
-0.489***
(0.00)
(0.00)
Less than HS Education
-1.34***
-0.391***
(0.00)
(0.00)
Age
0.014***
0.020***
(0.00)
(0.00)
Age Squared
-0.000***
-0.000***
(0.00)
(0.00)
Constant
3.230***
2.777***
(0.00)
(0.02)
Pseudo R2
0.19
0.13
N
14,076,739
471,162
Notes: Standard errors in parentheses. Data are from the 2000 IPUMS 5 percent
sample. Dependent variable in each regression is binary, and coded 1 if an
individual has a vehicle available. The regression results suggest that a
the odds a disabled person will have access to a vehicle are about 63 percent less
than the odds a non-disabled person will have access to a vehicles, while
the odds a poor person will have access to a vehicle are between 170 and 218
percent smaller than the odds a non-poor person will. Hence a person who is both
disabled and poor is much less likely to have access to a private
vehicle than the average American.
38