Evans School Review
Vol. 2, Num. 1, Spring 2012
A Benefit-Cost Analysis of
Road Pricing in Downtown
Seattle
By Steven Danna, Keibun Mori, Jake Vela & Michelle Ward
ABSTRACT
This benefit-cost analysis evaluates the key benefits and costs
associated with potential cordon-based road pricing in downtown
Seattle to determine if this road pricing option will be beneficial
for the region. The accrued benefits quantified in this analysis are
(1) reduced travel time, (2) increased travel reliability, (3) reduced
emissions, and (4) reduced traffic accidents. On the cost side, the
study measures (1) capital and operation costs of the toll collection
system, (2) additional subsidy for transit agencies to meet higher
demand, and (3) both direct and indirect financing costs. The toll
revenues from road pricing and changes in other government
revenues such as gas tax are neutral for the society as a whole,
because they represent cash transfer between government and
residents and do not cause adverse impacts on social welfare unless
the toll rates are dramatically high. The results over a thirty-year
period show that there will be net benefits of 585 million dollars in
present value for the region in the base scenario. The extensive
sensitivity and uncertainty analysis also demonstrate that the
project will remain favorable in conservative scenarios. The study
qualitatively examines possible impacts on other factors such as
retail sales and land use but concludes that they are unlikely to
undermine the key results in this analysis. These findings suggest
that in sum, road pricing in downtown Seattle would have positive
impacts for the city and region.
Keibun Mori completed his MPA at the Evans School focusing on energy and environmental
policy. He is specialized in forecasting long-term energy demand and analyzing policy solutions to
curb greenhouse gas emissions, with experience in assisting energy policy overhaul at the Washington
State Department of Commerce and for the Ministry of the Environment of Japan. He can be
contacted at [email protected].
26
Evans School Review
Vol. 2, Num. 1, Spring 2012
I.
INTRODUCTION
Seattle has the ninth most congested roadways in America; in 2010, an average driver in the
Seattle metro area spent 44 hours in traffic delays (Texas Transportation Institute 2011). Traffic
congestion causes fuel waste, which amounted to 46 million gallons of gasoline in the region in
2010, and this leads to decline in air quality and additional greenhouse gas (GHG) emissions
(Texas Transportation Institute 2011). Economists argue that traffic congestion is the
consequence of a market failure, as the private costs of driving (mostly fuel costs) do not account
for its social costs, such as the adverse impacts on roadway capacity and air quality (Downs
2007).
Road pricing can simultaneously solve these problems by internalizing these social costs
into the private costs of driving; the added costs induce some drivers to take a trip by other
means such as public transportation (Johansson and Mattsson 1994). Road pricing can be in a
variety of forms such as a gas tax, toll on freeways and bridges, and vehicle miles traveled
(VMT) tax, but road pricing on vehicles crossing a designated cordon has gained traction among
policy makers in recent decades (Johansson and Mattsson 1994). This cordon-based road pricing
is effective in deterring single-occupancy vehicles (SOV) to enter the city center, and Singapore
established the world’s first cordon-based road pricing scheme in 1975 to alleviate its persistent
traffic jam in its city center (SMOT 2012). The invention of electronic tolling systems in the late
20th century reduced the costs of toll collection and prompted several European cities such as
London, Stockholm, and Milan to follow Singapore’s success in the 2000s (Rotaris et al. 2010).
Evans (2007) conducted benefit-cost analysis on the road pricing in London, finding that the
benefits from road pricing, primarily shorter travel time and increased travel reliability, exceed
its costs by a ratio of 1.5. Similarly, Eliasson’s (2008) benefit-cost analysis shows that a trial
program in Stockholm yielded significant net benefits from reduced congestion, and it would
take only four years to recoup the initial capital investment.
The demonstrated success of the London and Stockholm road pricing projects indicates
that road pricing may be a feasible policy option for Seattle to alleviate the traffic delay and its
adverse impacts. To estimate the costs and benefits for Seattle, this study assumes that the City
of Seattle will follow the example of these cities by:
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
27
Evans School Review

Vol. 2, Num. 1, Spring 2012
Constructing and maintaining open-road electronic tolling facilities, similar to the
tolling system on the SR-520 Bridge connecting Seattle and Bellevue

Charging $2.15 per entry on weekdays only, an initial rate used in Stockholm
(Eliasson 2008), to the vehicles entering the designated cordon, defined by the
Puget Sound to the west, by the I-5 to the east, by Denny Way to the north, and S
King St. to the south (see areas defined by the blue lines in Figure 1)

Providing subsidies to King County Metro, a local transit agency, to increase the
level of transit service to accommodate increased demand for transit

Exempting special service vehicles such as transit and emergency vehicles
The following section outlines the methodology used to predict behavioral change and its
impacts. The subsequent sections estimate and aggregate various costs and benefits, then provide
qualitative discussions of other possible effects. After uncertainty and sensitivity analyses, this
study concludes with policy implications for Seattle.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
28
Evans School Review
Vol. 2, Num. 1, Spring 2012
Figure 1. Designated Cordon and Weekday Traffic Volume
Source(s): SDOT 2003.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
29
Evans School Review
Vol. 2, Num. 1, Spring 2012
II.
METHODOLOGY
The central assumption of this analysis is that vehicle trips into downtown Seattle have social
costs that drivers do not consider, in addition to its private costs such as the costs of fuel,
insurance, and maintenance. The social costs consist of the decreased reliability and increased
travel time, traffic accidents, and emissions from fuel consumption. Since drivers tend to neglect
these social costs when choosing to drive into the city center, the total number of vehicle trips is
larger than is otherwise socially optimal. Figure depicts a simplified market for vehicle trips
before (left) and after (right) road pricing. The demand curve represents the marginal value of
each vehicle trip to drivers.
Currently drivers take vehicle trips until their marginal costs of driving into downtown
equalize to the marginal benefits, but at this equilibrium point, the number of trips exceeds the
socially optimal point. This leads to a loss of social welfare, called deadweight loss, as shaded in
blue in Figure 2. By charging vehicles entering downtown, road pricing increases the private
costs of vehicle trips to match its social costs. This decreases the number of vehicle trips and
subsequently lowers or eliminates the welfare loss.
Figure 2. Market for Vehicle Trips before (Left) and after Road Pricing (Right)
To forecast this change in the number of vehicle trips to the city center, this study uses
the price elasticity of demand for vehicle trips. The elasticity represents the percentage change in
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
30
Evans School Review
Vol. 2, Num. 1, Spring 2012
vehicle trips for every percentage change in the price of a trip. For example, an elasticity of -0.5
implies that increasing the cost of driving by 10% will lower vehicle trips of 5%. Using the price
elasticity of demand, the following equation can determine the change in the number of vehicle
trips:
where
represents the price elasticity of vehicle trips into the city center.
Table 1 summarizes the parameters used to estimate the impacts on traffic volume. The
price elasticity of the number of vehicle trips in the table is from the survey on the observed
traffic change with similar road pricing schemes in London, Stockholm, and Milan (Transport
for London 2008; Evans 2008; Rotaris et al. 2010). Seattle’s drivers may be more or less
responsive to road pricing, so the Monte Carlo simulation addresses this uncertainty in the price
elasticity in the sensitivity analysis. When assuming the average toll rate to be in the range of $2
and $3 per entry, this model suggests that road pricing would curb the traffic volume by 12% to
30% as shown in Figure 3.
Table 1. Summary of Parameters for Traffic Volume Estimation
Parameter
Value
Note
Elasticity
-0.4 ~ -0.6*
Derived from observed data in
London, Stockholm, and Milan
Driving cost (fuel, insurance, vehicle
$0.5 per mile
AAA’s survey (US average)
12.8 miles
PSRC’s survey
Number of annual weekday vehicle
125 million
City of Seattle’s survey
trips to the city center
trips
costs)
Average commute distance to the city
center
Source(s): American Automobile Association (AAA) 2008; Evans 2008; Puget Sound Regional Council (PSRC)
2007; Seattle Department of Transportation 2003.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
31
Evans School Review
Vol. 2, Num. 1, Spring 2012
40
35
Millions Trips
30
25
Elasticity=-0.4
20
Elasticity=-0.5
15
Elasticity=-0.6
10
5
0
Toll=$2.00
Toll=$2.50
Toll=$3.00
Figure 3. Number of Annual Averted Vehicle Trips
These forecasted changes in the number of vehicle trips are the basis for the rest of this
analysis. In order to estimate its impacts on other parameters, this study assumes constant
relationships between the changes in vehicle volume and those for other parameters. Table 2
shows these constant relationships in elasticities, which express the degree of response to
changes in vehicle volume for each parameter. These elasticity estimates come from the
observed data in London, Stockholm, and Milan (Roritas et al. 2010), and they can predict the
percentage changes of various parameters for the Seattle metro area with respect to changes in
vehicle trips.
Table 2. Elasticities of Various Parameters to Changes in Vehicle Trips
Parameter
Elasticity
% Change ($2.15 toll per entry)
Greenhouse gas (GHG) emissions
0.5042
-12.0%
Carbon monoxide (CO) emissions
0.5882
-14.0%
Nitric oxide (NO) emissions
0.3571
-8.5%
Particulate matters (PM) emissions
0.5882
-14.0%
Traffic accidents
0.1513
-3.6%
0.1471
-3.5%
-0.6480
+9.2%
Total travel time
(travel time outside the cordon included)
Transit ridership
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
32
Evans School Review
Vol. 2, Num. 1, Spring 2012
III.
DIRECT COSTS AND BENEFITS
Capital and Operation Costs
It is challenging to estimate the capital and operation costs for road pricing in Seattle since the
costs vary from location to location and data from comparable projects are scarce. Stockholm
can provide useful figures for Seattle with some adjustments to account for differences in
roadway configuration and average income, because Seattle’s size and geographical features
roughly correspond to those for Stockholm (Popenoe 2001). Adopting the cost estimates for
Stockholm by Eliasson (2008), the capital costs would be $362 million,1 including the costs of
planning, public outreach, and construction of the toll system. The operation would cost $38.9
million per year, including the costs of toll collection, enforcement, and the toll system
replacement.2
For the financing costs for the capital expenditures, this analysis considers three
alternative financing methods:

Twenty-year bond structured with level debt-service payments

Twenty-year bond structured with level principal payments

One-time tax levy (no financing)
For the two bonding options, the assumed annual effective coupon rate is 5%, based on
fixed-coupon bonds issued in Washington State in the past several years. Using this baseline
coupon rate, the capital costs including financing costs would be $674 million for the level-debt
service option.
It is also possible to levy property or sales tax to fund capital costs of infrastructure
projects in Seattle. In this case, there are no direct financing costs for Seattle, but taxes do have
indirect costs called “shadow price” (Campen 1986). This shadow price, or marginal excess tax
burden (METB), represents welfare loss caused by taxes that create economic inefficiency.
Judging from the findings by Ballard et al. (1985a) and Jorgensen et al. (1990), the analysis sets
METB at 17% of the tax levy. When accounting for this shadow price of tax levy, the capital
1
A simple conversion from Swedish Krona to US dollar is $272 million, but it was inflated by a third because the
cordon for Seattle proposed in this paper requires the number of toll points to be approximately 33% higher than
Stockholm’s.
2
A simple conversion from Swedish Krona to US dollar is $32 million, but it was inflated to $38.9 million per year
to account for wage difference for administrative jobs between the two cities, based on the assumption that the half
of the operation costs are labor costs.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
33
Evans School Review
Vol. 2, Num. 1, Spring 2012
costs will amount to $424 million. The toll revenues themselves do not incur METB because toll
is an excise tax on goods with negative externalities (Zerbe and Dively 1994).
Transit Subsidy
The large number of averted vehicle trips is likely to increase the demand for mass transit.
Transport for London (2008) found an increase in transit ridership entering and exiting the
downtown cordon in London, with an increase in the speed of city buses and a decrease in wait
time for passengers due to greater reliability in the bus schedule. Similarly, in Stockholm, transit
ridership rose by 6% with road pricing, with most of the increase directly attributed to road
pricing (Eliasson 2008). The quality of transit system and land use patterns influence the degree
to which road pricing impacts transit ridership, but this study assumes that the commuters in
Seattle will show similar responses to road pricing, and that the impact on destination choice
would be negligible. Table 3 summarizes the parameters used to estimate the additional subsidy
for transit to provide an affordable alternative and avoid overcrowding of the transit system.
Table 3. Summary of Parameters for Transit Subsidy Estimation
Parameter
Value
Average vehicle occupancy rate
1.59 per vehicle
Costs per boarding
$3.67 per boarding
Farebox recovery rate
50%3
Average daily downtown inbound transit ridership
87,7204
Source(s): King County Dept. of Transportation 2007; King County Dept. of Transportation 2008; Puget Sound
Regional Council 2009; Puget Sound Regional Council 2010; Seattle Department of Transportation 2003.
Using these parameters and the number of averted vehicle trips, the estimated additional
transit subsidy to support road pricing would be $61 million in 20125. In reality, the needed
subsidies may be considerably less, because increased level of ridership could result in higher
farebox recovery rate and lower marginal costs of providing transit service.
3
The fare box recovery rate, a ratio of fare revenues over operating costs, was 24.6% for the entire system in 2008; however, the
bus routes serving downtown Seattle have much higher recovery rates because of the shorter route lengths and higher ridership
levels particular during daytime when additional capacity is necessary for the road pricing program. For this reason, this study
assumes that the fare box recovery rate will be roughly 50%, which is a rate of several highest ridership routes.
4
73100/weekday * 20% growth (2005-2008)
5
Additional subsidy =Δ vehicle volume * average vehicle occupancy * cost per boarding * (1- farebox recovery rate)
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
34
Evans School Review
Vol. 2, Num. 1, Spring 2012
IV.
INDIRECT COSTS AND BENEFITS
Travel Time Savings
Travel time savings occur when consumers benefit from reduced travel times. The average travel
time is currently 0.55 hour per trip (PSRC 2009), and the estimated reduction from the lower
traffic volume would be 0.0137 hour, using the following algorithm:
PSRC (2009) indicates that the value of travel time is correlated to the wage rate of
passengers for private vehicles, plus the value of cargo for commercial vehicles. 6 Using this
approach, PSRC (2009) estimates the value of time is $18.75 per hour for private vehicles, and
$50 per hour for commercial vehicles in Seattle.7 When multiplying the travel time savings by
the change in vehicle volume and value of time, the benefit of travel time savings will amount to
$77 million.
Travel Time Reliability
There are also benefits from improved travel time reliability; the reduced traffic volume lowers
the variations in travel time, which in turn causes further time savings as commuters have less
needs to budget additional time to avoid late arrival. Studies from London and Stockholm
suggest the reliability benefits be approximately one-third of the benefits of travel time savings
(Eliasson 2008; Santos and Fraser 2006). Assuming this relationship between travel time benefits
and reliability holds for Seattle, the value of travel time reliability will be $26 million.
Air Pollutants and Greenhouse Gasses
Each vehicle trip into the city center produces emissions from fuel combustion. These emissions
include greenhouse gasses (GHG), carbon monoxide (CO), nitrogen monoxide (NO), particulate
matter (PM), and volatile organic compounds (VOC). The emissions of these compounds
contribute to poor health conditions and climate change (IPCC 2007).
6
The approximate split between private and commercial vehicles is 75% private and 25% commercial
Texas Transportation Institute (2011) says the value of time for private vehicles is $16.30 per hour, and that for
commerce vehicles is $88.12.
7
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
35
Evans School Review
Vol. 2, Num. 1, Spring 2012
Table 4 summarizes the current emissions levels of these pollutants (Puget Sound Clean
Air Agency 2008), percentage changes induced by road pricing, which are estimated from the
elasticities of the emission level of air pollutants to the changes in vehicle volume, and values for
monetization (Muller and Mendelsohn 2007; Muller et al. 2009; McCubbin and Delucchi 1999).
Table 4. Summary of Current Emission Levels, Estimated Changes, and Monetization Values
Current Emissions
Parameters
(ton/year)
Estimated Change
Value
(per ton)
GHG Emissions
2,682,559.47
-8.5%
$27.0
CO Emissions
93,791.13
-9.8%
$45.4
11,579.71
-6.0%
$81.0
VOC Emissions
7,589.60
-8.6%
$838.0
PM Emissions
205.67
-9.8%
$7408.0
NO Emissions
8
Source(s): Puget Sound Clean Air Agency 2008; Muller and Mendelsohn 2007; Muller et al. 2009; McCubbin and
Delucchi 1999.
Multiplying these parameters for each pollutant suggests the value of the lower emissions
amount to $7.17 million per year.
Traffic Accidents
The decrease in traffic volume is likely to lower the number of traffic accidents, which
deteriorate social welfare through property damages, injuries, or loss of life. Rotaris et al. (2010)
report that road pricing in London lowered traffic accidents by 3.6%. Using the elasticity of
traffic accidents to traffic volume change, the estimated drop in the number of accidents would
be 2.4%. With the baseline number of accidents in Seattle (King County Dept. of Transportation
2009)9 and monetization values (PSRC 2009) listed in Table 5, the reduced accidents would
improve the social welfare by $1.4 million per year.
8
Since data on the change in VOC is unavailable, the analysis uses average of the other computed percent changes
(approx. 8.5%).
9
The report shows the total number of accidents in King County, and the analysis assumes that one third of them
involve vehicles destined to/from downtown Seattle.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
36
Evans School Review
Vol. 2, Num. 1, Spring 2012
Table 5. Summary of Traffic Accidents Estimation and Monetization Values
Parameters
Number of Incidents
Estimated Change
Value
(per accident)
Fatal
4
-2.4%
$2,500,000
Non-fatal
590
-2.4%
$75,500
Property damage only
595
-2.4%
$2,600
Source(s): King County Dept. of Transportation 2009; PSRC 2009.
V.
OTHER CONSIDERATIONS
Distributional Effects
Public policy often creates winners and losers, and these distributional effects tend to draw
criticism from the opponents of the policy. This study does not quantify them due to the
difficulties in quantification and redistribution, as recommended by Zerbe and Dively for most
policy options (1994), but qualitatively assesses the following three effects to see if they can
significantly affect the analysis results.
One effect is the potential decline in retail sales within the cordon, which could occur if
shoppers avoid downtown businesses because of the added travel costs. There have been several
studies on this effect, most notably the analysis by Quddus et al. (2005) on retail sales in London.
Quddus et al. (2005) compare the sales inside and outside the cordon before and after the
introduction of the road pricing with econometric analysis. They conclude that there were some
effects on particular franchises but no significant effect on the overall retail sales. Daunfeldt et al.
(2009) also find similar effects in Stockholm. Furthermore, since most weekday shoppers are
downtown workers, the toll should have minimal effect on their spending habits, and downtown
businesses may even benefit from road pricing due to the increased transit service and reduced
traffic congestion. These findings suggest that the potential adverse effects on retail sales within
the cordon are negligible.
The second potential effect is the reduced competitiveness of the freight industry. The
Port of Seattle is the eighth busiest port in the nation in cargo volume, and many of the container
terminals are located just south of the city center (Enghirst et al. 2011). The port faces
competition with other ports in the West Coast, including the Port of Tacoma located about 30
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
37
Evans School Review
Vol. 2, Num. 1, Spring 2012
miles south of Seattle. It is possible that the road pricing in the city center will increase the costs
for freight trucks and shift the demand away from the Port of Seattle and local freight carriers.
Enghirst et al. (2011) however find that through GPS-survey, none of the port-related trucks
passes through the city center. Furthermore, the averted vehicle trips from road pricing would
alleviate the congestion on local roadways and thus may improve the freight mobility.
The last distributional effect in this discussion is the potential change in land use patterns.
As discussed earlier, road pricing is likely to reduce significant amounts of vehicles entering the
cordon. This translates into the reduced demand for parking spaces, which could lead to
redevelopment of commercial parking lots, particularly concentrated in the northern and southern
edges of downtown Seattle. In addition, changes in commute patterns may have distributional
effects on housing prices. These effects on land use, however, are unclear and controversial. On
one hand, European Federation for Transport and Environment (2003) and Banister (2002) argue
that road pricing in theory has minimal effects in citywide land use patterns. On the other hand,
land use modeling by Hargreaves and Echenique (2008) implicates potential displacement of
low-income population in or near the city center, and econometric analysis by Zhang and Shing
(2006) finds that the housing price just outside the cordon increased at higher rates than other
areas. However, Hargreaves and Echenique (2008) also note that increased level of transit
service can easily offset the adverse effects, and Zhang and Shing (2006) caution that the
causation of the housing price change is not clear and needs further assessment through longterm observation. In sum, land use effects are too complex to predict and quantify, but even if the
effects are negative, it is possible to offset them through countermeasures such as increased level
of transit service.
Long-Term Effects
The long-term effects of road pricing are also controversial. Eliasson (2008) argues that the
traffic volume may return to the previous level after a certain period due to acclimatization effect,
a phenomenon that the effect of road pricing on driving behavior fades away as commuters
become accustomed to toll. Triple convergence, another phenomenon that freed-up road capacity
encourages people to drive more (Downs 2007), could also offset a portion of the traffic volume
decline if the lower level of traffic jam prompts some new transit riders to resume driving.
Eliasson (2008) however argues that these long-term effects are likely to be insignificant, since
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
38
Evans School Review
Vol. 2, Num. 1, Spring 2012
his observation period for road pricing in Stockholm was long enough to capture these effects. In
fact, in London the traffic volume reduction remained stable for years until the toll rate changed
(Transport for London 2008). Furthermore, in contrast to the effect of acclimation and triple
convergence, the decline in traffic volume could become larger in the long run because road
pricing can alter the consumer behavior in housing choice and car ownership. For these reasons,
these long-term effects are unlikely to undermine the effects of road pricing.
Government Revenues
Various government entities, ranging from the federal government to the local transit agency,
will receive or lose some revenues from road pricing. One obvious example is the toll revenues
from road pricing itself, which would amount to $109 million per year in this analysis. Other
changes include additional transit fare revenues for transit agencies, fewer revenues from
commercial parking tax and metered parking fee for the city, and fewer gas tax revenues for the
state and federal government. These revenue changes are simply cash transfer between the
governments and residents; whereas it is possible that these excise taxes beyond socially optimal
point can compromise social welfare by overcompensating the negative externalities, the study
assumes that there will be no change in social welfare as in most benefit-cost analyses (Zerbe
and Dively 1994). Therefore, while it is important to recognize and quantify the changes in
government revenues for financial analysis, the magnitude of the changes does not affect the net
benefits for the region.
VI.
RESULTS
The study sets the analysis period to thirty-years, a standard analysis period for infrastructure
projects. The key element in aggregating the benefits and costs is a discount rate, an instrument
to express future costs or benefits at today’s equivalent value. All benefit-cost analyses are
highly sensitive to a discount rate, and there have been various studies and recommendation on
the selection of a discount rate. For instance, the Congressional Budget Office (CBO 1998)
recommends a rate of 2% while the Office and Management and Budget (OMB 1992) suggest its
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
39
Evans School Review
Vol. 2, Num. 1, Spring 2012
upper bound be 7%. Zerbe and Dively (1994) recommend a more moderate rate ranging from
2.5% to 5.5%. For previous road pricing studies, Eliasson (2008) uses a rate of 4% for his
Stockholm analysis, and Santos and Fraser (2006) use a rate of 3.5% for their London analysis.
This analysis first uses a rate of 5% as discussed earlier, but later tests other rates in sensitivity
analysis.
In addition to this discounting, the projected population growth requires the adjustment of
the benefits and costs accordingly.10 It also assumes that the benefits and costs related to fuel use
would decrease with improving fuel economy, and that the value of time and labor costs would
increase with personal income growth. The costs of toll, driving, fuel, capital replacement, and
monetization values will grow at the rate of inflation, but the discount rate of 5% internalizes
these changes.
Table 6. Discount Rate and Growth Factors
Parameters
Rate
Discount rate
5.00%
Population growth
1.30%
Fuel efficiency growth
1.25%
Personal income growth
1.45%
Source(s): Bureau of Transportation Statistics 2011; Washington State Office of Financial Management 2011.
Table 7 shows the discounted costs and benefits over the thirty-year analysis period with
the level debt-service financing option. When aggregated, the net benefits are positive at $585
million in present value, showing that road pricing in downtown Seattle would have positive
impacts on the area.
Table 7. Discounted Benefits and Costs over Thirty-year Period
Discounted Costs
Capital costs
$357,459,493
Discounted Benefits
Travel time savings
$1,987,258,411
Operation costs
$802,829,103
Travel time reliability
$662,419,470
Transit subsidy
$1,038,042,253
Lower GHG emissions
$89,862,081
Air quality improvement
$19,802,735
Fewer traffic accidents
$24,098,531
NPV Benefits
$5,837,101,056
NPV Net Benefits
$585,110,379
NPV Costs
$5,251,990,677
10
The capital replacement portion of the operating cost, assumed to be about half, is not inflated while the labor
portion is inflated at income growth rate.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
40
Evans School Review
Vol. 2, Num. 1, Spring 2012
Finally, while the analysis period is thirty-year, the results find that the benefits will begin to
exceed the costs after the ninth year of implementation. This implies that it is likely that the
project would result in net benefits even if it lasts significantly less than thirty years.
VII.
SENSITIVITY ANALYSIS
The uncertainty of various parameters, particularly the price elasticity, toll rate, and discount rate,
requires extensive sensitivity analysis. Monte Carlo simulation can test the sensitivity of the
results over various parameters at one time (Mooney 1997). This simulation incorporates
information about the range and certainty of the estimates by expressing parameters with
probability distributions rather than with single point estimates. It then tests a large number of
possible values for each parameter and draws a distribution of possible outcomes. This
distribution gives a better picture of the possible benefits of the future than a single point
estimate can.
Table 8 summarizes the varied parameters with their plausible range and certainty. The
simulation uses normal distribution for parameters found through real-world observation, and
uniform distribution on other parameters.
Table 8. Summary of Parameters for Monte Carlo Simulations
Parameters
Distribution
Parameters
Discount Rate
Uniform
Min: 0.035, Max:0.080
Trip Costs
Normal
Mean:$6.40, Std. Dev: $0.64
Toll Rate
Uniform
Min: 2, Max: 3
Elasticity of Trips
Uniform
Min: -0.6, Max: -0.4
Capital Costs
Normal
Mean: $362,000,000, Std. Dev: $72,400,000
Operation Costs
Normal
Mean: $38,900,000, Std. Dev: $7,780,000
Fuel Efficiency Growth Rate
Normal
Mean: 1.25%, Std. Dev: 0.125%
Annual Income Growth Rate
Normal
Mean: 3.10%, Std. Dev: 0.310%
Annual Population Growth Rate
Normal
Mean: 1.30%, Std. Dev: 0.130%
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
41
Evans School Review
Vol. 2, Num. 1, Spring 2012
Figure 4 shows the results of the simulation with these parameters for 100,000 trials; in
most cases, the net benefits are positive, and the expected net present value is $824 million.
Figure 4. Results of Monte Carlo Simulation
The study also uses traditional sensitivity analysis to test the importance of three crucial
parameters in the model: the toll rate, price elasticity of trips, and discount rate. Figure 5 shows
how the net benefits in present value change for different values of these parameters. While the
results are highly sensitive to all three variables, even in the conservative scenarios the predicted
benefits will be positive.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
42
Million dollar
Evans School Review
Vol. 2, Num. 1, Spring 2012
1,400
1,200
1,000
800
600
400
200
0
$2.0 $2.5 $3.0
-0.4 -0.5 -0.6
Toll Rate
Elasticity
3.5% 5.3% 7.0%
Discount Rate
Figure 5. Summary of Sensitivity Analysis
VIII. CONCLUSION
The expected growth of the region is likely to increase traffic congestion in Seattle in the coming
decades (PSRC 2010). A policy solution is necessary to address traffic congestion and its
negative externalities for healthy growth of the metro area. After considering numerous benefits
and costs, this analysis indicates that cordon-based road pricing in Seattle can significantly lower
its traffic volume and result in better social welfare. Shorter travel time and better reliability are
the major contributors to this positive outcome, but the secondary benefits such as fewer harmful
emissions and traffic accidents also play significant roles in improving the welfare.
Despite using the best available information in determining the net benefits in present
value, the study uses a number of assumptions to predict the impacts. The elasticity-based
modeling approach itself is one of the major limitations of the study, and more sophisticated
origin-destination traffic modeling (OD model) will provide better information on the impacts of
road pricing, particularly on traffic volume. However, an important policy implication from the
extensive sensitivity and uncertainty analysis is that road pricing is very likely to have positive
net benefits even in conservative scenarios. This benefit-cost analysis therefore successfully
demonstrates that road pricing can play a crucial role in improving social welfare in the region
and enabling the city to achieve healthy growth.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
43
Evans School Review
Vol. 2, Num. 1, Spring 2012
Bibliography
Ballard CL, JB Shoven, and J Whalley. “General equilibrium computations of the marginal
welfare costs of taxes in the United States.” America Economic Review, 75 no. 1 (1985a):
128–138.
Ballard CL, JB Shoven, and J Whalley. “The total welfare cost of the United States tax system: a
general equilibrium approach.” National Tax Journal, 38 no. 2 (1985b): 125–140.
Banister, David. The Integration of Road Pricing with Land Use Planning. May 2002.
Bureau of Labor Statistics. Seattle–Tacoma–Olympia, WA: National Compensation Survey,
December 2009. Dec 2010: 15.
Campen, James. Benefit, Cost, and Beyond: The Political Economy of Benefit-Cost Analysis.
Ballinger Pub. Co., 1986: 35-40
Congressional Budget Office (CBO). “The Economic Effects of Federal Spending on
Infrastructure and Other Investments,” CBO Paper, 1998: 13.
Daunfeldt, Sven-Olov, Rudholm Niklas, and Ulf Rämme. “Congestion charges and retail
revenues: Results from the Stockholm road pricing trial.” Transportation Research Part A:
Policy and Practice, 43 no. 3 (2009): 306-309.
Downs, Anthony. Still Stuck in Traffic: Coping with Peak-Hour Traffic. Washington, D.C.:
Brookings Institute Press, 2007.
Eliasson, Jonas. Cost-Benefit Analysis of the Stockholm Congestion Charging System. Centre for
Transport Studies, Royal Institute of Technology, Sep 2008.
Enghirst Abigail L., Jacob B. Gonzalez, Keibun Mori, Maura Rowell, Anfisa Rovinsky, Amber
L. Trout, Sungchan Lim, and Justin R. Vickerman. Port of Seattle Truck Traffic Study.
University of Washington, Mar 2011.
European Federation for Transport and Environment. Congestion pricing in London: An
European Perspective. 2003.
Evans, Reg. Central London Congestion Charging Scheme: Ex-post Evaluation of the Quantified
Impacts of the Original Scheme. Transport for London, Jun 2007.
Evans, Reg. Demand Elasticities for Car Trips to Central London as revealed by the Central
London Congestion Charge. Transport for London, Sep 2008.
Hargreaves, Anthony J, and MH Echenique. “Cambridge Futures: Forecasting the Effect of
Congestion Charging on Land Use Transport.” Road Congestion Pricing in Europe.
Implications for the United States. Edward Elgar Publishers, 2008: 98-117.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
44
Evans School Review
Vol. 2, Num. 1, Spring 2012
Intergovernmental Panel for Climate Change (IPCC), Climate change 2007: the physical science
basis: summary for policymakers. IPCC Secretariat, 2007
Johansson, Börje, and Lars-Göran Mattsson. Road Pricing: Theory, Empirical Assessment and
Policy. Springer, 1994.
Jorgenson DW, and KY Yun. “Tax Reform and US Economic Growth.” Journal of Political
Economy, 98 no. 5 (1990): 151-193.
King County Dept. of Transportation. 2008 Annual Management Report. 2009.
King County Dept. of Transportation. 2008 Route Performance Report. 2009.
King County Dept. of Transportation. 2008 Traffic Safety Report. 2009.
McCubbin, Donald and Mark Delucchi. “The Health Costs of Motor-Vehicle-Related Air
Pollution.” Journal of Transport Economics and Policy, 33 no. 3 (1999): 253-286.
Mooney, Christopher. Monte Carlo Simulation: Quantitative Applications in the Social Sciences.
Sage Publications, 1997.
Muller, N and R Mendelsohn. “Measuring the Damages of Air Pollution in the United States.”
Journal of Environmental Economics and Management, 54 (2007): 1-14.
Muller, Nicholas, Robert Mendelsohn, and William Nordhaus. Environmental Accounting for
Pollution: Methods with an Application to the United States Economy. May 2009.
Office of Management and Budget (OMB). “Guidelines and Discount Rates for Benefit-Cost
Analysis of Federal Programs.” Circular No. A-94 Revised, 1992.
Popenoe, David. Private Pleasure, Public Plight: Urban Development, Suburban Sprawl, and
the Decline of Community. Transaction Publishers, 2001: 43-51.
Puget Sound Clean Air Agency. 2005 Air Emission Inventory for King, Kitsap, Pierce, and
Snohomish Counties. Mar 2008.
Puget Sound Regional Council (PSRC). Puget Sound Trends. Rep. Vol. 10, 2007.
Puget Sound Regional Council (PSRC). Traffic Choices Study. Apr 2008.
Puget Sound Regional Council (PSRC). Transportation 2040. 2010.
Puget Sound Regional Council (PSRC). Vision 2040. Dec 2009.
Quddus, Mohammed, Alon Carmel, and Michael GH Bell. The impact of the congestion charge
on retail: the London experience. Greater London Authority, May 2005.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
45
Evans School Review
Vol. 2, Num. 1, Spring 2012
Rotaris, Lucia, Romeo Danielis, Edoardo Marcucci, and Jerome Massiani. The urban road
pricing scheme to curb pollution in Milan, Italy: Description, impacts and preliminary cost–
benefit analysis assessment. Mar 2010.
Santos, Georgina and Fraser, Gordon. “Road Pricing: Lessons from London.” Economic Policy,
21 no. 46 (2006): 263-310.
Seattle Dept. of Transportation (SDOT). “Transportation in the City Center Today.” Seattle
Urban Mobility Plan. 2003.
Singapore Ministry of Transport (SMOT). “Electronic Road Pricing.” May 2011.
Sweden Statistics. Wage and salary structures and employment in the primary municipalities.
2007.
Texas Transportation Institute. “Performance Measure Summary - Seattle, WA.” 2011 Annual
Mobility Report, Sep 2011.
The American Automobile Association (AAA). “Your Driving Costs Study.” Apr 2011.
Transport for London. Central London Congestion Charging: Impacts Monitoring, Sixth Annual
Report. Jul 2008.
US Dept. of Transportation (USDOT). Average Fuel Efficiency of U.S. Passenger Cars and
Light Trucks. Dec 2011.
US Dept. of Transportation (USDOT). Motor Vehicle Fuel Consumption and Travel. Dec 2010.
US Dept. of Transportation (USDOT). “Summary of Travel Trends.” 2009 National Household
Travel Survey. Jun 2011: 33.
Washington State Office of Financial Management (OFM). 2011 Long-Term Economic and
Labor Force Forecast for Washington, 2011.
Zerbe, Richard O Jr. and Dwight Dively. Benefit-Cost Analysis: In Theory and Practice,
HaperCollin Paper Publisher, 1994: 139-142, 290.
Zhang, Yi and Hui-Fai Shing. The London Congestion Charge and Property Prices: An
Evaluation of the Impact on Property Prices Inside and Outside the Zone. London School of
Economics, Aug 2006.
A Benefit-Cost Analysis of Road Pricing in Downtown Seattle
46