How much will roads cost to reduce congestion?

August 2006

Building Roads
to Reduce tRaffic congestion
in
ameRica's cities: How mucH
and
at wHat cost?

By David T. Hartgen, Ph.D., P.E. and M. Gregory Fields

Project Director: Robert W. Poole, Jr.

POLICY

STUDY

346

The Galvin Mobility Project

America's insufficient and deteriorating transportation network is choking our cities, hurting our economy, and reducing our quality of life. But through innovative engineering, value pricing, public-private partnerships, and innovations in performance and management we can stop this dangerous downward spiral. The Galvin Mobility Project is a major new policy initiative that will significantly increase our urban mobility and help local
officials move beyond business-as-usual transportation planning.

Reason Foundation

Reason Foundation's mission is to advance a free society by developing, applying, and

promoting libertarian principles, including individual liberty, free markets, and the rule of law. We use journalism and public policy research to influence the frameworks and actions of policymakers, journalists, and opinion leaders.

Reason Foundation's nonpartisan public policy research promotes choice, competition,

and a dynamic market economy as the foundation for human dignity and progress. Reason produces rigorous, peer-reviewed research and directly engages the policy process, seeking strategies that emphasize cooperation, flexibility, local knowledge, and results. Through
practical and innovative approaches to complex problems, Reason seeks to change the way people think about issues, and promote policies that allow and encourage individuals and voluntary institutions to flourish.

Reason Foundation is a tax-exempt research and education organization as defined under IRS code 501(c)(3). Reason Foundation is supported by voluntary contributions from individuals, foundations, and corporations. The views are those of the author, not necessarily those of Reason Foundation or its trustees.

Copyright © 2006 Reason Foundation. Photos used in this publication are copyright © 1996, 2006 Photodisc, Inc. All rights reserved.

The Galvin Mobility Project is made possible by the generous support of Robert Galvin.

The Galvin Mobility Project

raffic congestion is choking our cities, hurting our economy, and reducing our quality
of life. Rush-hour delays rob us of time with our families, and commute times often

dictate where we live and work. The impact our inadequate transportation network has on our economy is alarming. We waste an estimated $63 billion annually in time and fuel while sitting in traffic. Moreover, businesses and their customers bear enormous costs associated with traffic-related logistics problems, delivery delays, poor transportation reliability, and fewer potential employees within commuting distance.

Reason Foundation is developing practical, cost-effective solutions to traffic congestion

with the Galvin Mobility Project, a policy initiative that will significantly increase our urban mobility through innovative engineering, value pricing, public-private partnerships, and
innovations in performance and management. Under the leadership of Reason's Director of Transportation Studies Robert Poole, Reason's original research is building comprehensive policy recommendations that enhance mobility and help local officials move beyond
business-as-usual transportation planning.

The old canard "we can't build our way out of congestion" is not true. Adding capacity

and improving management of roads can eliminate chronic congestion. Public-private

partnerships to build and operate toll facilities have sparked innovations in engineering

and design, overcoming obstacles such as limited right-of-way and noise pollution. Capital markets also provide access to much-needed investment capital and ensure that new
highway capacity is built where it is most needed.

In addition to adding road capacity, changing the way highways are managed can help to maximize the use of the capacity we have. The introduction of Intelligent Transportation System technologies can speed resolution to traffic delays, and electronic toll collection technologies can make extensive tolling practical. More importantly, variable pricing of lanes can keep traffic flowing all day by responding to changing demand.

We can solve our congestion woes. We can upgrade to an innovative, market-driven,

world-class transportation infrastructure. We can change the institutions that guide our
transportation decisions to create greater responsiveness, robustness, and efficiency. The
Galvin Mobility Project provides the ideas and tools needed to make change happen.

R e a s o n F o u n d a t i o n

Building Roads to Reduce Traffic

Congestion in America's Cities: How
Much and at What Cost?

By David T. Hartgen, Ph.D., P.E. and M. Gregory Fields

Executive Summary

his report quantifies the magnitude of traffic congestion and the cost of its removal through the
provision of additional capacity. Other studies in the Galvin Mobility Project examine the role

of other means of addressing congestion as well as financing, project management, and other issues that go along with adding road capacity in urban areas.

We define and quantify severe congestion, in which peak-hour traffic volumes exceed road

capacity, and estimate future congestion if trends continue. With the help of 32 participating

urbanized areas, the report uses sophisticated traffic modeling techniques to determine how much
additional capacity will be needed to relieve severe congestion. These findings are then extended to
all 403 urbanized areas. The report then estimates the cost of providing that additional capacity.
These costs include construction in each state, major bridge widenings, adjustments for induced
travel, and requirements for some elevated or tunnel sections. Detailed results are provided for each
city and state. We also provide a simplified state-level assessment for rural areas and for moderate
urban congestion.

This report finds that severe traffic congestion is pervasive in large regions and is worsening

throughout the United States. In the future even small, urbanized areas are likely to experience

congestion common in mid-sized areas today. The cause of this increase is not wastefulness but

increasing population and preferences for private mobility, combined with limited additions to road
capacity. Nationwide, the number of lane-miles of severely congested roads is expected to increase
from about 39,500 in 2003 to 59,700 in 2030. To relieve severe congestion by providing additional

capacity, an additional 104,000 lane-miles of capacity (about 6.2 percent of current lane-miles) will be needed, costing about $533 billion over 25 years, in 2005 dollars. The amount needed—about $21 billion per year—is about 10-15 percent of the federal highway program over 25 years, about 28 percent of the cost of present urban transportation plans, and about 39 cents per day per
commuter trip. However, the travel time savings are estimated at about 7.7 billion hours annually, so the cost per hour of delay saved is about $2.76. If moderate congestion and rural congestion are also to be addressed, an additional $304 billion will be needed.

We also find that congestion relief through provision of additional capacity is quite feasible, given current budgets. The benefits of an investment in additional capacity would be substantial. In
addition to reduced travel time, other benefits include smoother traffic flow, reduced accidents, improved air quality through lower emissions, lower fuel use and operating costs, more reliable travel, lower logistical costs for manufacturing and delivery, more choices of jobs for workers and businesses and wider choices for consumers.

R e a s o n F o u n d a t i o n

Table of Contents

Introduction ... . 1

Methodology ... .. 4

Eliminating Severe Congestion by Increasing Capacity ... . 6

A. Overall Findings ... . 6

B. Costs by Road Type (Functional Class) ... .. 11

C. Lane-Miles Needed and Costs by Urbanized Area and Region ... . 12

D. Risk Analysis of the Cost to Relieve Severe Congestion ... . 23

E. Additional Costs of Removing Moderate Urban Congestion ... ... 23

F. Rural Congestion ... . 24

E. Comparison with Long-Range Plans ... ... 26

User Benefits of Congestion Reduction ... . 32

Conclusions ... .. 35

A. The Magnitude of Present and Future Congestion ... .. 35

B. Capacity Needs to Eliminate Severe Congestion ... .. 35

C. The Cost of Dealing with Congestion ... .. 36

D. These Costs Are Reasonable Compared to Planned Transportation Spending ... ... 37

E. The Likely Benefits ... . 38

F. Summary ... ... 39

About the Authors... .. 40

Endnotes... . 41

R e a s o n F o u n d a t i o n

Tables and Figures

Tables

Table 1: Trends and Forecasts of Travel Time Indices ... .. 7

Table 2: Cities with 2030 Travel Time Delays Worse Than Today's Los Angeles... .. 7

Table 3: Additional Cities with Travel Time Delays Worse Than Today's Chicago ... .. 8

Table 4: Changes in Traffic Density Forecast, 403 Urbanized Areas ... .. 8

Table 5: Severely Congested Roadways, 403 Urbanized Areas ... ... 9

Table 6: Summary of Findings: Cost to Relieve Severe Congestion... .. 10

Table 7: Individual Costs of Relieving Severe Congestion ... . 11

Table 8: Costs by Roadway Type ... . 12

Table 9: Costs of Relieving Severe Congestion by City Size ($B) ... ... 12

Table 10: Costs of Relieving Severe Congestion for the Top Ten Cities in Population ($B) ... ... 13

Table 11: Costs of Relieving Severe Urban Area (UA) Congestion for the 25 States in the

Northeast/North- Central United States ($Billion)... . 16

Table 12: Costs of Relieving Severe Urban Area (UA)Congestion for the Southeast United States

($B) ... ... 17

Table 13: Costs of Relieving Severe Urban Area (UA) Congestion for the 8 States in the

Southwest United States ($Billions) ... . 19

Table 14: Costs of Relieving Severe Congestion in the Northwest United States ($B)... .. 19

Table 15: Costs of Relieving Severe Congestion for Alaska and Hawaii ($B) ... ... 20

Table 16: States Ranked by Congested Lane Miles in 2030... . 22

Table 17: States Ranked by 2030 Urban Area Lane Miles Needed... .. 22

Table 18: States Ranked by Total Costs of Lane Miles Needed... .. 22

Table 19: Mileage and Costs of Removing Moderate Urban Congestion ... . 24

Table 20: Mileage and Costs of Relieving Severe Rural Congestion ... . 25

Table 21: Mileage and Costs of Relieving Moderate Rural Congestion ... .. 25

Table 22: Costs to Relieve Congestion versus Present Plan Costs... .. 27

Table 23: Average Costs per Resident to Relieve Congestion versus Present Plan Costs,

Weighted by Population ... .. 28

Table 24: User Benefits Analysis for Detroit ... ... 33

Table 25: User Benefits Analysis for Atlanta ... ... 33

Table 26: Cost of Capacity Expansion Per Hour of Delay Saved... ... 34

Table 27: Severely Congested Facilities, 403 Urbanized Areas ... .. 35

Table 28: Summary of Needs and Costs ... .. 36

Figures

Figure 1: An Overview of the Congestion Problem ... .. 2

Figure 2: Urban Areas in the United States Requiring Congestion Relief with Costs to Relieve

Congestion ... . 14

Figure 3: Urban Areas in the Northeast/North Central United States Requiring Congestion

Relief with Costs to Relieve Congestion ($Million)... ... 15

Figure 4: Urban Areas in the Southeast United States Requiring Congestion Relief with Costs

to Relieve Severe Congestion ($Million)... ... 17

Figure 5: Urban Areas in the Southwest Requiring Congestion Relief with Costs to Relieve

Congestion ($Million) ... ... 18

Figure 6: Urban Areas in the Northwest United States Requiring Congestion Relief ($M) ... . 20

Figure 7: Urban Areas in Alaska and Hawaii Requiring Congestion Relief and Costs to Relieve

Congestion ($M). .. .. 21

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 1

P a r t 1

Introduction

rban congestion is a growing problem, as indicated in the annual congestion statistics

reported for large urban regions. The Texas Transportation Institute annually calculates costs

of traffic delay in the nation's largest cities and its 2005 report pegged the costs at $65 billion a
year.¹ Numerous polls show growing frustration on the part of citizens and businesses with
congestion and its deleterious impacts on personal lives and commerce. In a 2001 national travel
survey fully 24 percent of respondents cited congestion as a severe or serious problem; in large
cities 39 percent thought so, but even in smaller cities 21 percent thought so. ² Federal policy
documents cite congestion as "…one of the biggest transportation challenges facing
us…pervasive…affect[ing] more trips, more hours of the day and more of the transportation
system."³

"Congestion is one of the single largest threats to our economic prosperity and way of life."
—U.S. Department of Transportation report released May 16, 2006

Figure 1 summarizes the status of planning for urbanized area congestion. The trend is steeply

rising congestion compared to 2003 levels. Most current long-range transportation plans forecast
worse congestion conditions even after the expenditure of billions of dollars. In most regions the
competition is fierce for scarce dollars for transportation improvements. Many regions have very
large maintenance and repair needs for the existing system. Planning requirements encourage
expenditures for a variety of projects serving numerous objectives other than congestion. Pulled by
competing priorities, many cities and states appear to be focusing largely on other objectives and
are de-emphasizing the congestion problem. Some regions assert that congestion cannot be
eliminated or reduced, or that the extra capacity will just be 'filled up' anyway, and addition of
capacity is very expensive.

Nevertheless, the bottom line is clear: additional actions beyond current plans will be needed if congestion is to be reduced or maintained at current levels.

2 Reason Foundation

Figure 1: An Overview of the Congestion Problem

Congestion

Trend

Present Plans

2003 Congestion
Level

Potential Additional Actions

1995 2003 2030

Several recent studies have reviewed the magnitude and cost of congestion. In addition to the

Texas Transportation Institute's annual mobility study, the American Highway Users Alliance

recently reviewed the delays associated with the worst highway bottlenecks.⁴ They found that

between 1999 and 2004 the number of major bottlenecks in the U.S. had increased from 167 to 233 locations, but that 7 of the worst 18 identified in 1999 had been alleviated. The federal government has estimated the cost of improving highway and bridge infrastructure in both condition and
capacity at about $49.9 billion annually.⁵ These important studies quantify portions of the problem but do not cover all urbanized areas. Nor do they include the 'spreading' effect of congestion as
traffic spreads out in time and space, or possible 'induced' travel as roads are widened. More
recently, several states and urbanized areas, notably Texas,⁶ Washington,⁷ and Atlanta,⁸ have
recognized the cost of congestion to their economic competitiveness and are taking steps to reduce
or eliminate congestion. Most importantly, these initiatives have come primarily from pressure by
the business community, which sees congestion as a serious threat to regional competitiveness. The Transportation Research Board, a national research organization, has recently listed congestion as
its top concern.⁹ However, little comprehensive assessment of the extent of the full congestion
problem or the cost to deal with it has been made.

This report uses estimates from national congestion statistics and sophisticated traffic assignment
methods to review recent trends in congestion and forecast how much there will be after present
transportation plans are completed. It calculates the number and cost of additional lane-miles
needed to eliminate severe congestion. It then compares these costs with the costs of planned
transportation improvements. Detailed results are provided for each state and for each of 403
urbanized areas.

Studies are necessarily limited in scope, based on assumptions, and it is important to know what
this report does not cover. It does not look in detail at ALL congestion, focusing instead on the
most severe congestion in urbanized areas. (In a short section we look briefly at moderate

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 3

congestion, but only at the state level.) It is not intended as a full assessment of various approaches
to dealing with congestion problems, many of which are appropriate along with capacity
expansions. These include a wide range of pricing options (HOT lanes, tolls, and mileage-based
fees), transit services, demand management, activity scheduling, ITS (Intelligent Transportation
System) technologies, operational improvements, and many others. Many of these options are
reviewed in other reports of the Galvin Mobility Project. We do not minimize the difficulties of
locating these improvements, given environmental and location constraints. Nor does the report
deal with rehabilitation or maintenance needs, which may overlap with capacity expansion needs
for specific road sections, or may compete for funds, thus increasing costs. All these issues warrant
attention, but do not obviate the need for a comprehensive assessment of capacity expansion
opportunities.

Instead, the report focuses on the capacity side of congestion relief, looking at what a concerted nationwide effort to address severe congestion through capacity relief might cost. It is therefore intended as input to the policy debate, not a conclusion.

4 Reason Foundation

P a r t 2

Methodology

e focus initially on severe congestion in 403 urbanized areas in the United States with

populations greater than 50,000 persons in 2003.¹⁰ Supplemental analysis is also added for

rural congestion outside of urbanized areas, and moderate congestion in urbanized areas. A more
detailed description of the steps is provided in Appendix B. The basic steps followed in the study
are:

1. Current Congestion. Estimates of current (2003) severe congestion,¹¹
sometimes called

'gridlock',¹² are prepared by consolidating statistics on urbanized area congestion indices for 86 larger regions,¹³ supplemented by estimates of congestion indices for 317 additional smaller regions. Special tabulations from the national Highway Performance Monitoring System are also used to estimate, for each urbanized area, the amount of mileage and lane-miles with
severe congestion, by functional class. ¹⁴ Along with current travel time index (TTI) values, this data is used as estimates of current congestion.

2. Future Congestion. Future congestion is estimated by relating current congestion indices to
urbanized area population and traffic density (daily vehicle-miles-of-travel¹⁵ per mile of road).
Forecasts of population and traffic density to 2030 are then made for each urbanized area, and
future TTIs are also estimated by trend. A separate forecast of lane-miles with severe
congestion (volume exceeds capacity) is then made using forecasts of population and traffic
congestion indices. These projections incorporate planned capacity additions under long range
transportation plans. This is the primary estimate of future congestion used for analytical
purposes.

3. Needed Capacity Increases. Capacity increases needed to relieve severe congestion are then
estimated from the future congestion estimates (lane-miles over: traffic exceeds capacity) for
each urbanized area, such that peak-hour capacity is sufficient to carry peak-hour traffic
volumes. These estimates are adjusted to account for diverted traffic (traffic moving to faster
roads as they are widened) using network-based traffic assignments or similar information for
32 urbanized areas participating in the study. The numbers from these contributing areas are
highlighted in
Appendix C. Results from these 32 urbanized areas assignments are then used to
scale up the preliminary estimates for the other urbanized areas.

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 5

4. Cost of Capacity Increases. To estimate costs of capacity increases, federal estimates of

construction costs per-lane-mile (engineering, construction, right-of-way, and mobilization) for roads of various types are multiplied by estimates of needed lane-miles from Step 3. These
costs are then adjusted for price increases between 1997 and 2005, different construction costs by state, the need for large-bridge widenings, induced travel,¹⁶ and the need for elevated-or-
tunnel construction. Risk analysis software (Crystal Ball®) is used to analyze the uncertainty in these assessments. This software considers a range of inputs rather than an average for
particular entries in the database. Ranges are used for the cost factors in the model, and the
resulting uncertainty in the overall estimate is then determined and shows that our estimate
falls in the middle of the most likely range for the likely costs.

5. Comparisons with Long-Range Plans. Cost estimates for relieving severe congestion are

then compared with the financial transportation plans of 43 urbanized areas to determine what portion of current plans might be implied by a significant congestion-relief effort.

6. Rural Congestion and Moderate Urban Congestion.
In a short follow-up analysis, the

additional cost of reducing congestion further in rural areas and moderate congestion in urban areas is also estimated using state-level data.

7. User Benefits of Congestion Reduction. Finally, savings in delay are estimated for each
region, and for the United States as a whole. These are compared with the cost of capacity
expansion. This estimate is conservative since it does not consider fuel savings, reduced
operating costs, or other benefits. Two more detailed examples (Atlanta and Detroit) are also
developed to show how travelers benefit from congestion removal, through higher speeds and
savings in travel time, operating costs, and lower accident rates.

6 Reason Foundation

P a r t 3

Eliminating Severe Congestion by
Increasing Capacity

A. Overall Findings

1. Urban Population Trends

The 403 largest urbanized areas of the United States, those over 50,000 in population in 2003,

together contain about 189 million people, or about 66 percent of the U.S. population. ¹⁷ Assuming recent economic and population trends continue, the population of these urbanized areas is
expected to grow to about 245 million by 2030, an increase of about 30 percent.

2. Trends in Congestion

The Texas Transportation Institute generates an annual survey on congestion. The Institute uses a
"Travel Time Index" (TTI), defined as the ratio of travel time in peak hours to the travel time in
off-peak hours.¹⁸ For instance, an index of 1.5 means that travel time in the peak hour is 50 percent
longer than in the off-peak. The 'delay' in the travel time is the portion over 1.0. This data was
used to chart trends in congestion in the nation's largest 86 cities, then extended to other smaller
urbanized areas, and then forecast to 2030 based on trends and on forecasts of population and
traffic density.

Table 1 shows recent trends and forecasts of travel time indices for urbanized areas. For large areas
over 3 million in population, congestion is predicted to increase from an average of 1.46 to 1.76
over the next 25 years; for smaller cities, the forecast is for less severe but faster growing
congestion.

However, if trends continue, by 2030 even small cities will be experiencing significant and
noticeable congestion. In very large regions, 'delay' over the next 25 years will increase 65
percent, from 46 percent over free-flow travel time to 76 percent over free-flow travel time. In
smaller regions, the 'delay' portion of peak-hour travel time will more than double. These results

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 7

suggest that traffic congestion is likely to widen geographically and deepen in intensity, affecting more mid-size and smaller urbanized areas and worsening throughout the nation. Residents of smaller areas may notice the effect more than residents of larger areas.

Table 1: Trends and Forecasts of Travel Time Indices

City Size 1982 1993 1995 2003 2030 Est. Relative Increase in

'Delay' 2003 to 2030

Ave 3+ M 1.15 1.35 1.36 1.46 1.76 65%

Ave 1-3 M 1.08 1.18 1.21 1.28 1.53 89%

Ave 500K-1M 1.05 1.11 1.13 1.18 1.36 100%

Ave 250-500 K 1.04 1.06 1.15 150%

Ave 50-250 K 1.03 1.04 1.09 125%

To put these in perspective, consider today's congestion levels. Present-day (2003) Los Angeles is the most congested city in the United States, with a travel time index of 1.75. But by 2030,
urbanized areas with over three million people will be averaging about the same travel time delay (1.76) as today's Los Angeles. Cities with travel time delays equal to today's Los Angeles will include Atlanta, Denver, and Minneapolis/St. Paul.

By 2030, regions in the 1-3 million range will be seeing congestion about as severe as present-day Chicago (1.56). These cities include Baltimore, Portland, Sacramento, and Tucson. By 2030, small regions will be seeing congestion about the same as areas with over one million in population saw in the early 1980s.

Table 2: Cities with 2030 Travel Time Delays Worse Than Today's Los Angeles

City Population in 2030 (000s) Congestion Index in 2030

Los Angeles-Long Beach 15,652 1.94

Chicago 9,522 1.88

Washington 5,973 1.87

San Francisco-Oakland 4,968 1.86

Atlanta 5,009 1.85

Miami 7,551 1.84

Denver-Aurora 3,210 1.80

Seattle-Tacoma, WA 3,963 1.79

Las Vegas 1,029 1.79

Minneapolis-St. Paul 3,370 1.76

Baltimore 2,437 1.75

Portland 2,513 1.75

8 Reason Foundation

Table 3: Additional Cities with Travel Time Delays Worse Than Today's Chicago

City Population in 2030 (000s) Congestion Index in 2030

New York-Newark 21,295 1.74

Sacramento 2,488 1.73

Dallas-Fort Worth 7,014 1.73

San Diego 3,720 1.70

San Jose 2,036 1.65

Phoenix-Mesa 5,313 1.64

Riverside-San Bernardino 2,629 1.64

Charlotte 1,185 1.62

Bridgeport-Stamford 1,018 1.62

Boston 4,636 1.62

Houston 3,987 1.61

Philadelphia 5,879 1.61

Tucson 1,094 1.60

Salt Lake City 1,251 1.59

Orlando 2,112 1.59

An alternate way of viewing traffic congestion is to look at overall
traffic density, the amount of
traffic per mile of road space. Table 4 shows recent trends and forecasts of traffic density for cities
by size.

Table 4: Changes in Traffic Density Forecast, 403 Urbanized Areas

City Size 1995 2003 2020 2030 Percent Increase 2003-30

Ave 3+ M 7,050 7,500 8,400 9,000 19.8

Ave 1-3 M 6,000 6,300 7,000 7,500 18.3

Ave 500K-1M 5,800 6,000 6,400 6,600 10.5

Ave 250-501K 4,700 4,900 5,500 5,800 17.6

Ave 50-251K 3,800 4,100 4,800 5,300 28.2

Traffic Density = Average annual daily traffic, per mile of road, rounded to 100s.

Traffic densities are rising, but not as fast as travel time factors, because as urban regions spread they add more road mileage, and because congestion goes up faster than just the increase in
driving. Nevertheless, street-level traffic is likely to rise about 20 percent in large regions, but even higher, almost 30 percent, in smaller regions over the next several decades.

A third way of looking at congestion is the number of lane-miles of severely congested facilities.
"Severe congestion" is defined as peak-hour traffic volume which exceeds the peak-hour capacity
of the facility to carry it. Recent data from the federal Highway Performance Monitoring System
(Table 5)
show that the Urban Interstate system has the highest proportion of severely congested
miles, 16.4 percent, followed by other freeways, 11.5 percent. About 39,500 lane-miles of road in
the 403 largest urbanized areas currently carry more traffic than their rated capacities. This number

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 9

is expected to grow about 50 percent to about 59,700 severely congested lane-miles by 2030,

assuming trends in traffic density and population continue.

Table 5:
Severely Congested Roadways, 403 Urbanized Areas

Roadway Type 2003 Miles 2003 Miles Percent 2003 Lane-Miles 2030 Lane-Miles Percent

Severely Severely Severely Severely Change

Congested* Congested Congested* Congested*

Urban Interstate 12,766 2,100 16.4 17,800 27,400 54.3

Urban Other 8,677 1,000 11.5

Freeway

Urban Other 44,351 2,200 4.9 9,000 12,400 37.5

Principal Arterial

Urban Minor Arterial 75,124 3,700 4.9 12,700 19,900 56.1

Urban Collector 75,894 2,700 3.6

Urban Local 553,822 - -

Total 770,634 11,700** - 39,500 59,700 51.1

Severely congested= facilities for which peak-hour traffic volumes exceed capacity * Rounded to nearest 100 for convenience.

** Excluded local mileage.

The spreading of urbanized areas may obviate congestion trends somewhat, and in some regions

congestion trends may be flattening a bit. On the other hand, rising auto use at either end of the age spectrum, longer and healthier living, and greater economic wealth suggest that travel will continue to increase and that congestion is likely to spread unless capacity is increased. Trends may change, but these look likely to continue for some time, one reason we have expanded the assessment
below to include rural and moderate congestion.

3. Costs to Relieve Severe Congestion

The amount of additional capacity needed to relieve predicted severe congestion is greater than the
lane-miles severely congested, because the additional capacity will be partially filled by cars
shifting from other roads or other travel times. Using the results of specialized studies by 32
participating cities (described in Appendix B), and expanding these findings to all 403 cities, we
estimate that about 104,000 lane-miles, or about 6.2 percent of the current urbanized area lane-
miles, would be needed to provide enough capacity to relieve severe congestion in urbanized areas.
The cost of this mileage in 2005 dollars is about $391 billion, or $3.8 million per lane-mile.

But since these capacity additions constitute major system expansions in many urban regions, their likely impact in creating some additional travel due to the convenience of uncongested roadways should also be considered. This effect, sometimes called 'induced travel' is estimated at about 10,900 lane-miles, and is estimated to add about $41 billion.¹⁹

10 Reason Foundation

Additions for major bridge widenings and for elevated or tunnel-design sections might also

significantly increase costs, because of high-cost construction in constrained space.²⁰ We estimate
that about $31 billion would be needed for major bridge widenings to accommodate increased
capacity requirements, and about $70 billion would be needed for elevated structures or tunnel
sections.

Thus, the total cost to provide the needed capacity to significantly relieve severe congestion is about $533 billion, in 2005 dollars. Over 25 years, this is about $21 billion per year.²¹

Table 6: Summary of Findings: Cost to Relieve Severe Congestion

Lane-Miles Increment, Cost,

Needed $Billion $Billion

Base Estimate 104,000 (6.2%) 391 391

Add-on for Induced Travel 41

Add-on for Large Bridge Widening 31

Add-on for Elevated or Tunnel Structures 70 533

These estimates compare reasonably well with other recent partial assessments. The USDOT

estimate of $49.9 billion annually includes costs of improving the physical condition of roads and
bridges as well as increased capacity, and so should be somewhat higher than our estimates.²² A
recent study for work needed to reduce Atlanta's future travel time index to 1.18 (about the same
as removing severe congestion), estimates the cost at $22.1 billion; when adjusted for a key
different assumption concerning arterial lane-miles, the number is about $13.4 billion, very close to
our estimate of $13.1 billion for Atlanta.²³ And a recent assessment of the needs for a group of
major Texas urbanized areas estimates that $54 billion would be needed to reduce the "Texas
Congestion Index" to 1.18, close to our Texas state total estimate, $49.1 billion.²⁴ Though some
factors could raise costs above our estimates, we are confident that the above national estimate is
reasonable at this time.²⁵

These cost estimates should be put in perspective. The present U.S. highway program, including

capital and maintenance, costs about $140 billion per year (the new federal transportation bill

increases that by about 40 percent); over 25 years, it might be expected to total at least $3.5 trillion,
in current dollars. ²⁶ Thus $533 billion, if fully added to the current program, would be an increase
of about 15 percent of highway capital and maintenance expenditures. But since the current
program includes some capital actions that will relieve some severe congestion, and the program
itself has been increasing, the percent increase over the current program is probably closer to 10
percent.

Another way to look at these costs is to compare them with other common expenses.


In 2005, Americans spent about $41 billion at Lowes Home Improvement, $81.5 billion at
Home Depot, and about $13.6 billion on pet food;


There are about $21 billion in student loans in default nationwide;

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 11


In 2003, the federal government spent nearly $25 billion on things it could not identify—
"unreconciled expenditures."

So, the $21 billion/year cost of removing severe congestion would be about one-fifth of what we spent at two major chains to fix up our homes, and about one and a half times what we spent for pet food. It is an amount our government has been willing to essentially write off in the past. The cost of relieving severe congestion may not be as high as many say.

4. Unit Costs per Capita and per Commuter Trip

From a per-commuter perspective, the projected cost of $533 billion over 25 years is in the range of $368 per commuter per year for large regions, declining to less than $50 per commuter per year in smaller regions.

Table 7: Individual Costs of Relieving Severe Congestion

City Size Cost per Commuter Cost per Resident Cost per Commuter Cost per

per Year ($) per Year ($) per Day ($) Commuter Trip ($)

Ave 3+ M 368.47 184.23 1.47 0.75

Ave 1-3 M 118.97 59.49 0.48 0.24

Ave 500K-1M 147.04 73.52 0.59 0.30

Ave 250-500K 151.73 75.87 0.61 0.31

Ave 50-250K 44.96 22.48 0.18 0.09

Average for All $196.21 $98.11 $0.78 $0.39

Put another way, the costs average about 39 cents per trip by each commuter—from 75 cents per commuter trip in large regions down to 9 cents per trip for commuters in small regions.

It would not seem unreasonable that commuters in large cities might be willing to pay, in various ways, 75 cents per trip for relief from severe congestion. It seems highly likely that commuters in smaller regions would be willing to pay 9 cents per trip, in various ways, for relief from severe
congestion. And these costs would be even smaller if spread out over all trips—commuter trips are only about 20-25 percent of all trips.

B. Costs by Road Type (Functional Class)

As might be expected, the additional capacity necessary to significantly reduce severe congestion is largely concentrated in the higher-level roads.²⁷ Table 8
shows that the Urban Interstate (UI) and Other Freeway/Expressway (OFE) system would need to be expanded in capacity by about onethird nationwide; urban Other Principal Arterials would need to be expanded 11.7 percent. For the lower systems about a 6.3 overall percent increase is implied. The unit cost per lane-mile averages about $9.4 million for Interstates, $3.6 million for principal arterials, and $2.2 million for lesser streets. Overall, the average unit cost is about $5.1 million per lane-mile.

12 Reason Foundation

Table 8:
Costs by Roadway Type

Roadway Type Current Current Est. Additional Percent of Cost of Cost per

2003 Lane-Miles* Lane-Miles Current Additional Added

Miles* Needed* System to Lane-Miles Lane-Mile

be added ($Billion) ($Million)

Urban Interstate 21,400 113,600 38,600 33.90 362 9.4

and OFE

Urb OPA 44,400 177,400 20,700 11.70 75 3.6

Urb Min Arterial 75,100 150,200 1,300 0.80 96 2.2

Urb Collector 75,900 151,800 41,900 27.50

Local 553,800 1,107,600 800 0.07

Other 900

Total 770,600 1,700,600 104,100 6.20 533 5.1

* Rounded to nearest 100 for convenience.

C. Lane-Miles Needed and Costs by Urbanized Area and Region

The additional capacity needed to eliminate severe congestion is not evenly distributed across the
403 urbanized areas. Of the $533 billion in total costs, 10 cities with over three million population
account for 61.0 percent (Table 9). Regions between one million and three million population
account for about 18.4 percent of the total, and those between 1 million and 500,000 population
account for about 9.0 percent. Smaller regions account for 11.6 percent of the total costs.

Table 9: Costs of Relieving Severe Congestion by City Size ($B)

City Size Interstate Other Minor Total Percent of Percent of Cost per Hour

and OFE Principal Arterials Total Cost Total US of Delay

Arterial Pop Saved

3+M $251.1 $36.0 $38.5 $325.6 61.0 23.7 $ 2.72

1-3 M 70.7 10.7 16.7 98.2 18.4 16.6 1.83

500k-1M 20.1 12.0 16.0 48.1 9.0 8.1 3.73

250-500K 14.7 10.5 15.2 40.4 7.6 6.3 10.44

50-250K 5.5 5.1 10.6 21.2 4.0 10.4 6.43

Total All $362.3 $74.4 $96.9 $533.5 100.0 65.2 $ 2.76

In terms of delay saved per dollar expended, the urbanized regions also show differences. The largest regions typically have lower cost per hour of delay saved because congestion is more pervasive even though expansion costs are higher. Overall, the cost per hour of delay saved is about $2.76, with the larger regions lower and the smaller ones higher.²⁸

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 13

1. The 10 Largest Urbanized Areas

As the next table (Table 10) indicates, the 10 largest regions have a third of the U.S. urban

population, a third of the urban lane miles with traffic that exceeds their capacity

(Volume/Capacity ratios above 1.0), and a fourth of the urban lane-miles needed to relieve

congestion, but almost 2/3 of the total costs. This is not surprising—the costs for capacity

additions in major cities are likely to be extremely high as available rights-of-way (ROW) are often
fully built out and potential ROW is often cost-prohibitive forcing creative (and costly) alternatives
to normal highway construction. Cost per hour saved is generally lower in the larger regions;
Detroit's cost per hour saved is higher because it has less congestion relative to size, so while it
needs a lot of capacity, the amount of congestion relieved is less, so the delay saved is less.

Table 10: Costs of Relieving Severe Congestion for the Top Ten Cities in Population ($B)

City 2003 Pop 2030 Pop 2030 Lane 2030 Lane Total Costs of Cost per Hour

(000s) (000s) Miles expected Miles Lane Miles of Delay Saved

to be congested Needed Needed ($Billion)

New York 17,717 21,295 3,827 2,446 $38.6 1.24

Los Angeles 12,520 15,652 3,594 3,695 67.7 2.62

Chicago 7,702 9,522 2,793 3,875 53.8 3.52

Philadelphia 5,287 5,879 1,475 1,929 19.6 3.75

Miami 5,104 7,551 1,919 3,400 30.0 3.39

Dallas 4,312 7,014 2,646 3,656 26.1 3.52

Washington 4,277 5,973 1,130 1,803 16.2 1.52

San Francisco 4,120 4,968 1,304 2,261 29.2 3.72

Boston 3,988 4,636 990 1,493 20.3 4.56

Detroit 3,939 4,277 1,136 2,301 24.1 9.05

Subtotals 68,966 86,767 20,813 26,858 325.6

National Totals 189,510 245,523 59,688 104,122 533.5

% of Total 36.4 35.3 34.9 25.8 61.0

A closer look at the urbanized areas across the U.S. reflects the trends noted above: the larger areas have the bulk of the costs, and the costs are more concentrated in the interstates and arterials
(Figure 2). The larger circles show the larger concentration of costs, with urban interstates,
freeways, and principal arterials making up most of the costs in major urbanized areas. The overall pattern reflects this concentration of higher costs in the larger areas, with heavier concentrations of costs on the east and west coasts and in states with large metropolitan regions. The five general regions of the United States will be more fully analyzed below. Appendix C
contains a complete listing of each state and its regions' congestion circumstances.

14 Reason Foundation

Figure 2: Urban Areas in the United States Requiring
Congestion Relief with Costs to Relieve Congestion

2. The Northeast/North-Central United States

Figure 3 and Table 11 show needs for the Northeast/North-Central United States. With 25 states,
this region is estimated to have 47.7 percent of the future U.S. urbanized area population and 46.6
percent of future severe congestion needs. The two states with the largest needs in this region,
Illinois and New York, account for 18.7 percent of the U.S. total. Within the region, metropolitan
New York, Chicago, Boston, Washington, Philadelphia, and Detroit dominate the requirements,
each with more than $16 billion in needs. Numerous mid-sized areas also require $1-5 billion each.
However, even small urbanized areas in rural states such as Maine and Vermont have congestion-
relief needs.
Maine's needs are concentrated in Portland ($130.8 million), Lewiston ($21.7 million) and Bangor ($24.6 million).

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 15

Figure 3: Urban Areas in the Northeast/North Central United States

Requiring Congestion Relief with Costs to Relieve Congestion ($Million)

20.76

1557.92

3. The Southeast United States

Figure 4 and Table 12 show the pattern of congestion needs for the Southeast region. Although the
region as a whole is less congested than the Northeast, the needs are still large. Texas and Florida,
the top two states in the region, together contain 16.4 percent of the national total needs. Dallas and
Miami, each needing more than $26 billion, are the top areas in the region, with Atlanta, Houston
and San Antonio each needing at least $5.6 billion. Many other urbanized areas need $1-2 billion
each. (New Orleans is estimated to need $598 million, assuming its recovery from Hurricane

Katrina puts it back on its trend.)

16 Reason Foundation

Table 11: Costs of Relieving Severe Urban Area (UA) Congestion for the 25 States in the

Northeast/North- Central United States ($Billion)

State 2003 UA 2030 UA 2030 UA Lane 2030 UA Total Costs of

Pop (000s) Pop (000s) Miles expected Lane Miles Lane Miles

to be congested Needed Needed ($B)*

Illinois 9,114 11,044 3,037 4,459 55.0

New York 21,089 24,573 4,735 4,511 45.0

Michigan 6,732 7,666 1,785 3,668 27.1

Pennsylvania 9,978 10,698 2,456 4,465 25.5

Massachusetts 5,575 6,493 1,214 1,961 21.9

Washington,

DC 4,277 5,973 1,130 1,803 16.2

North Carolina 3,507 5,257 1,537 4,361 12.4

Minnesota 2,803 3,756 1,427 2,531 7.7

Ohio 8,062 8,954 1,212 1,610 5.6

Tennessee 3,334 4,467 1,291 2,754 5.0

Kentucky 1,372 1,703 391 1,234 4.6

Missouri 3,930 4,757 1,163 1,972 4.6

Connecticut 2,837 3,234 585 1,618 3.4

Virginia 3,269 4,021 735 989 3.1

Indiana 2,167 2,691 762 2,269 3.1

Wisconsin 3,019 3,519 877 1,687 3.0

Maryland 2,689 3,299 546 580 2.3

Rhode Island 1,218 1,411 267 257 0.85

New Jersey 734 913 164 388 0.65

Iowa 961 1,184 165 304 0.57

New Hampshire 391 521 141 218 0.30

West Virginia 509 487 77 154 0.28

Maine 268 314 50 82 0.18

Vermont 133 168 28 61 0.13

Delaware 80 107 25 42 0.06

Subtotals 98,048 117,210 25,801 43,980 248.5

National Totals 189,510 245,523 59,688 104,122 533.5

% of Total 51.7 47.7 43.2 42.2 46.6

*Cities grouped into major state, so costs may include work in nearby states.

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 17

Figure 4: Urban Areas in the Southeast United States Requiring

Congestion Relief with Costs to Relieve Severe Congestion ($Million)

Costs to Relieve Congestion ($M)

50000 Urban Interstates and Freeways

25002.5 Other Principal Arterials

5 Minor Arterials, Collectors, Local Streets

0 100 200 300

Miles

Table 12: Costs of Relieving Severe Urban Area (UA)Congestion for the Southeast United States ($B)

State 2003 UA 2030 UA 2030 UA Lane Miles 2030 UA Lane Miles Total Costs of Lane

Pop 000s Pop 000s expected to be congested Needed Miles Needed ($B)*

Texas 13,244 19,951 7,986 12,929 $49.1

Florida 13,122 19,474 3,990 8,536 38.7

Georgia 4,311 6,716 1,516 3,220 14.3

South Carolina 1,720 2,285 726 1,934 4.9

Louisiana 2,534 2,829 846 1,248 3.3

Oklahoma 1,483 1,844 363 727 3.1

Alabama 1,906 2,239 458 967 2.5

Arkansas 692 938 271 1,207 2.5

Mississippi 743 953 139 254 0.72

Subtotals 39,755 57,229 16,296 31,024 119.2

National Totals 189,510 245,523 59,688 104,122 533.5

% of Total 21.0 23.3 27.3 29.8 22.3

*Cities grouped into major state, so costs may include work in nearby states.

18 Reason Foundation

4. The Southwest United States

Figure 5 and Table 13 reflect the Southwest Region, where California dominates the congestion needs picture, with 22.9 percent of the national total. Within California, Los Angeles and San Francisco comprise about 80 percent of the California total. These two urbanized areas together have almost 18 percent of the national need, almost as much as New York and Illinois combined. Other states also have important needs. Major urbanized areas with significant needs are Phoenix and Denver, each with close to $10 billion in needs.

Figure 5: Urban Areas in the Southwest Requiring Congestion
Relief with Costs to Relieve Congestion ($Million)

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 19

Table 13: Costs of Relieving Severe Urban Area (UA) Congestion for the 8 States in the Southwest

United States ($Billions)

City 2003 UA 2030 UA 2030 UA Lane Miles 2030 Lane Miles Total Costs of Lane

Pop (000s) Pop (000s) expected to be congested Needed Miles Needed ($B)*

California 30,487 39,874 8,730 13,132 121.9

Colorado 3,246 5,048 1,111 4,668 11.4

Arizona 3,909 6,888 4,082 3,813 11.3

Utah 1,830 2,797 505 948 2.3

Nevada 1,147 1,483 281 919 2.3

Nebraska 852 1,107 262 966 1.7

New Mexico 738 1,058 249 556 1.4

Kansas 601 730 148 578 0.81

Subtotals 42,810 58,985 15,368 25,579 153.2

National Totals 189,510 245,523 59,688 104,122 533.5

% of Total 22.6 24.0 25.7 24.6 28.7

*Cities grouped into major state, so costs may include work in nearby states.

5. The Northwest United States

The Northwest Region as a group has only about 2 percent of the national severe-congestion needs, but two states and urbanized areas dominate its requirement. Seattle and Portland, each with over $2.5 billion in needs, are the largest urbanized areas; these two have 70 percent of the region's
needs. However, even rural states like Montana and Idaho have some needs. In Idaho, for instance, Boise ($277 million), Idaho Falls ($14.5 million) and Pocatello ($9.6 million), have some localized congestion needs. In Montana, Missoula ($23.1 million), Great Falls ($10.2 million), and Billings ($23.7 million) constitute the state's modest need.

Table 14: Costs of Relieving Severe Congestion in the Northwest United States ($B)

State 2003 UA 2030 2030 UA Lane 2030 UA Lane Total Costs of Lane

Pop (000s) UAPop Miles expected to Miles Needed Miles Needed ($B)*

(000s) be congested

Washington 4,081 5,497 1,063 1,477 $6.9

Oregon 2,372 3,478 660 1,020 3.2

Idaho 536 841 180 278 0.37

North Dakota 276 322 55 108 0.15

South Dakota 190 260 26 51 0.06

Montana 245 293 24 31 0.06

Wyoming 129 154 25 22 0.05

Subtotals 7,829 10,845 2,033 2,987 10.7

National Totals 189,510 245,523 59,688 104,122 533.5

% of Total 4.1 4.4 3.4 2.9 2.0

*Cities grouped into major state, so costs may include work in nearby states

20 Reason Foundation

Figure 6: Urban Areas in the Northwest United States Requiring Congestion Relief ($M)

Costs to Relieve Congestion ($M )

50000 Urban Interstates and Freeways

25002.5 Other Principal Arterials

5 Minor Arterials, Collectors, Local Streets

0 100 200 300

Miles

6. Alaska and Hawaii

The severe congestion needs of Alaska and Hawaii are relatively modest. In Hawaii, Honolulu

($1023.7 million) and Kailua-Kaneohe ($50.1 million) constitute the state needs, while Anchorage ($815.3 million) and Fairbanks ($33.5 million) represent Alaska's need.

Table 15: Costs of Relieving Severe Congestion for Alaska and Hawaii ($B)

State 2003 UA 2030 UA 2030 UA Lane 2030 UA Total Costs of

Pop (000s) Pop (000s) Miles expected Lane Miles Lane Miles

to be congested Needed Needed ($B)

Hawaii 742 832 121 321 1.1

Alaska 326 423 68 230 0.85

Subtotals 1,068 1,255 189 551 1.9

National Totals 189,510 245,523 59,688 104,122 533.5

% of Total 0.6 0.5 0.3 0.5 0.4

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 21

Figure 7: Urban Areas in Alaska and Hawaii Requiring Congestion
Relief and Costs to Relieve Congestion ($M).

22 Reason Foundation

Table 16: States Ranked by Table 17: States Ranked by 2030 Table 18: States Ranked by Total

Congested Lane Miles in 2030 Urban Area Lane Miles Needed Costs of Lane Miles Needed

2030 Urbanized Area 2030 Urban Area Lane Total Costs of Lane

State Lane Miles Congested State Miles Needed State Miles Needed ($B)

1. California 8,730 1. California 13,132 1. California 121.90

2. Texas 7,986 2. Texas 12,929 2. Illinois 55.00

3. New York 4,735 3. Florida 8,536 3. Texas 49.10

4. Arizona 4,082 4. Colorado 4,668 4. New York 45.00

5. Florida 3,990 5. New York 4,512 5. Florida 38.70

6. Illinois 3,037 6. Pennsylvania 4,465 6. Michigan 27.10

7. Pennsylvania 2,456 7. Illinois 4,459 7. Pennsylvania 25.50

8. Michigan 1,785 8. North Carolina 4,361 8. Massachusetts 21.90

9. North Carolina 1,537 9. Arizona 3,813 9. D.C. 16.20

10. Georgia 1,516 10. Michigan 3,668 10. Georgia 14.30

11. Minnesota 1,427 11. Georgia 3,221 11. North Carolina 12.40

12. Tennessee 1,291 12. Tennessee 2,754 12. Colorado 11.40

13. Massachusetts 1,214 13. Minnesota 2,531 13. Arizona 11.30

14. Ohio 1,212 14. Indiana 2,269 14. Minnesota 7.70

15. Missouri 1,164 15. Missouri 1,972 15. Washington 6.90

16. D.C. 1,130 16. Massachusetts 1,961 16. Ohio 5.60

17. Colorado 1,111 17. South Carolina 1,934 17. Tennessee 5.00

18. Washington 1,063 18. D.C. 1,803 18. South Carolina 4.90

19. Wisconsin 877 19. Wisconsin 1,687 19. Kentucky 4.60

20. Louisiana 846 20. Connecticut 1,618 20. Missouri 4.60

21. Indiana 762 21. Ohio 1,610 21. Connecticut 3.40

22. Virginia 735 22. Washington 1,477 22. Louisiana 3.30

23. South Carolina 726 23. Louisiana 1,248 23. Oregon 3.20

24. Oregon 660 24. Kentucky 1,234 24. Oklahoma 3.10

25. Connecticut 585 25. Arkansas 1,207 25. Virginia 3.10

26. Maryland 546 26. Oregon 1,020 26. Indiana 3.10

27. Utah 505 27. Virginia 989 27. Wisconsin 3.00

28. Alabama 458 28. Alabama 967 28. Alabama 2.50

29. Kentucky 392 29. Nebraska 966 29. Arkansas 2.50

30. Oklahoma 363 30. Utah 948 30. Utah 2.30

31. Nevada 281 31. Nevada 919 31. Nevada 2.30

32. Arkansas 271 32. Oklahoma 727 32. Maryland 2.30

33. Rhode Island 267 33. Maryland 580 33. Nebraska 1.70

34. Nebraska 262 34. Kansas 578 34. New Mexico 1.40

35. New Mexico 249 35. New Mexico 556 35. Hawaii 1.10

36. Idaho 180 36. New Jersey 388 36. Alaska 0.85

37. Iowa 165 37. Hawaii 321 37. Rhode Island 0.85

38. New Jersey 164 38. Iowa 304 38. Kansas 0.81

39. Kansas 148 39. Idaho 278 39. Mississippi 0.72

40. New Hampshire 142 40. Rhode Island 257 40. New Jersey 0.65

41. Mississippi 139 41. Mississippi 254 41. Iowa 0.57

42. Hawaii 121 42. Alaska 230 42. Idaho 0.37

43. West Virginia 77 43. New Hampshire 218 43. New Hampshire 0.30

44. Alaska 68 44. West Virginia 154 44. West Virginia 0.28

45. North Dakota 55 45. North Dakota 108 45. Maine 0.18

46. Maine 50 46. Maine 82 46. North Dakota 0.15

47. Vermont 28 47. Vermont 61 47. Vermont 0.13

48. South Dakota 26 48. South Dakota 51 48. South Dakota 0.06

49. Wyoming 25 49. Delaware 42 49. Montana 0.06

50. Delaware 25 50. Montana 31 50. Delaware 0.06

51. Montana 24 51. Wyoming 22 51. Wyoming 0.05

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 23

D. Risk Analysis of the Cost to Relieve Severe Congestion

Costs to add capacity were calculated using estimates of construction costs, induced travel, bridgewidening construction, and elevated-tunnel construction costs (see Appendix B). But there is a fair amount of uncertainty in those numbers. To determine the likely range of costs, an uncertainty analysis of the cost was conducted using the Crystal Ball® software package. This software uses a range of numbers for each factor affecting the cost rather than just one, and produces a range of estimates rather than just one. (See Appendix H for the full risk analysis.)

The analysis found that there is a 95 percent certainty that the costs to relieve severe congestion
will fall between $454.6 billion and $623.0 billion, in today's dollars, an 80 percent probability
that the costs to relieve severe congestion are less than $573 billion, and a 90 percent probability
that they are less than $593 billion. Hence, our estimates are well within the most likely range.

Finally, we conducted a sensitivity analysis of the impact of the factors that were varied in Crystal
Ball® on the final cost distribution. These results show that the higher construction costs for the
interstates and freeways, and differences in state construction costs are the key factors influencing
overall cost. They account for about 60 percent of the uncertainty in the final cost. Although major
bridge-widening needs and elevated sections are individually expensive, they are a relatively small
part of total costs.

E. Additional Costs of Removing Moderate Urban Congestion

The analysis so far focuses on severe urban congestion, defined formally as congestion in which traffic exceeds the capacity of the carrying facility, "Level of Service (LOS) F" in transportation parlance. But congestion at lower levels, known as LOS levels D and E, is also common although not so deleterious to travel times. Since congestion is rising in all urban areas, our first focus was on the capacity needed to relieve severe congestion, and the costs of doing so.

Moderate congestion can also be reduced by the provision of more capacity and of course by other
means. To evaluate the costs of reducing moderate congestion by capacity improvements we
undertook an analysis similar to the detailed LOS F, with several important differences. Primarily
we used state-level data, not urbanized area data; this obviated the need for the 32 participating
urbanized areas to provide additional traffic assignment information. Otherwise, the methodology
(see Appendix B) was similar, using comparable unit costs of construction and criteria for
widening.

Data from 2003 show about 15,900 miles of road in urbanized areas with congestion levels

between 80 and 95 percent of capacity—approximately levels of service D and E, or moderate congestion. This is somewhat more than the 11,700 miles estimated to be severely congested. Using similar growth rates from 2003 to 2030, it is estimated that by 2030 this number will increase to 42,000 miles.

24 Reason Foundation

Since this congestion is moderate rather than severe, it would be imprudent to add more than just two lanes (one on each side) to provide additional capacity for this traffic. In addition, there is
unlikely to be significant pent-up demand for use of these facilities, as there would be for severely congested facilities. Therefore the additional lane-miles needed to deal with this congestion are just twice the mileage of moderately congested facilities, or 84,000 lane-miles. This compares with an estimated 104,000 lane-miles needed to deal with severe congestion.

Because this example requires less expansion, the costs of this expansion are significantly less too.
We estimate that the effort would cost about $270.5 billion, in 2004 dollars. This compares with
an estimate of $533.4 billion for dealing with severe congestion. Thus, if both severe and moderate
congestion were targeted, the cost of significant reduction would be in the range of $803.9 billion,
in 2005 dollars.

Table 19: Mileage and Costs of Removing Moderate Urban Congestion

2003 Miles 2030 Miles 2030 Lane-Miles Cost to Provide

0.80-0.95 0.80-0.95 Needed to Additional Capacity

(000s) (000s) Remove (000s) ($B)

Interstate and OFE 3,801 10,395 20,790 117.1

Other Primary Arterials 3,924 10,206 20,412 65.0

Minor Arterials 4,828 12,469 24,937 50.7

Major Collectors 3,300 8,966 17,931 37.6

Total 15,853 42,035 84,070 270.5

State-level details of this assessment are provided in Appendix D.

This assessment does not deal with the advisability of removing moderate congestion. While a case
could be made that any congestion, moderate or severe, should be removed since it is economically
inefficient, the specific benefits of widening a particular road to remove LOS F, or E, or D
congestion, would have to be evaluated on a case-by-case basis. Our assessment merely deals with
the overall magnitude of the problem and the likely scale of dealing with it nationally.

F. Rural Congestion

Not all traffic congestion is urban: anyone who has felt the frustration of sitting in a line on a rural road going to a ski area or a beach resort understands that rural congestion can and does occur, and can be significant.

Although this study focuses primarily on urban congestion, we did take a brief look at rural

congestion.²⁹ The methodology is similar to that used to review LOS D and E congestion: we

looked at state-level data only, not county or urban-fringe data. Costs for widening rural roads are
significantly cheaper than for urban roads, and the extent of the network that might have to be

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 25

treated is considerably less too. Therefore, our findings are modest, compared with urban congestion. Appendix B describes our methodology in more detail.

Tables 20 and 21 summarize the findings. Overall, about 2,800 miles of rural highway are

currently severely congested (have volume/capacity ratios of 0.95 or higher). Another 4,000 miles
are moderately congested. Using similar growth rates for our urbanized areas, but aggregating to
the state level, we estimate that about 8,200 miles are likely to be severely congested in 2030, and
11,700 miles are likely to be moderately congested; this is probably an overstatement since rural
traffic is not growing as fast as urban traffic. Details for each state are shown in Appendices E and

F.

Since most of these roads have little potential for significant pent-up demand, most could be

widened by just two lanes to deal with this severe congestion. Thus, about 16,400 lane-miles of additional capacity would be needed to deal with severe congestion, and another 23,400 miles for moderate congestion by 2030. However, most of this mileage is in lower-cost environments with flat or rolling terrains that are not nearly as expensive as urban environments. Therefore, the cost of this additional mileage is estimated at about $14.2 billion, in 2005 dollars, for severe congestion, and $19.7 billion for moderate congestion, over the next 25 years.

Table 20: Mileage and Costs of Relieving Severe Rural Congestion

Rural LOS F Analysis

2003 Miles 2030 Miles 2030 Cost to

congested congested Lane-Miles Provide

(000s) (000s) Needed (000s) ($B)

Interstate 607 1,558 3,117 2.5

Other Primary Arterials 948 2,823 5,646 5.5

Minor Arterials 729 2,268 4,535 4.0

Major Collectors 516 1,528 3,057 2.1

Total 2,800 8,177 16,354 14.2

Table 21: Mileage and Costs of Relieving Moderate Rural Congestion

Rural LOS D-E Analysis

2003 Miles 2030 Miles 2030 Cost to

0.80-0.95 0.80-0.95 Lane-Miles Provide

(000s) (000s) Needed (000s) ($B)

Interstate 1,392 3,745 7,490 5.8

Other Primary Arterials 954 2,793 5,586 5.4

Minor Arterials 1,065 3,315 6,630 5.8

Major Collectors 575 1,862 3,724 2.8

Total 3,986 11,715 23,429 19.7

26 Reason Foundation

E. Comparison with Long-Range Plans

To bring these findings into perspective, we have prepared an analysis of how the implied costs to relieve severe congestion compare with the planned expenditures in 43 selected urbanized areas. The sources of the costs shown here are the latest long-range plans, as described in the Websites of each urbanized area's MPO. Most plans are for 2030, although a few are for 2025. The urbanized areas are in order by 2003 population. Table 22
summarizes the findings.

According to the table, the urbanized areas reviewed plan to spend, in total, about $1.47 trillion
over the next 25 years on their transportation plans. Of this, 43.1 percent will be on highway-
related projects (capital and maintenance), 52.8 percent on transit (capital and maintenance), and

3.6 percent on other projects, primarily pedestrian/bike facilities and enhancements. With the

exceptions of New York City and Chicago, most of the transit commuting shares are under 10

percent, and many are in the range of 1-2 percent for smaller regions. Overall, the cost of removing
severe congestion in these urbanized areas ($413 billion) as estimated in this study, amounts to

28.0 percent of the total cost of the long-range plans as now formulated by the MPOs. Although in some areas these costs are substantially higher as a percentage of the total and in others lower, it is clear that in many areas, removal of severe congestion would be a relatively small portion of the plan focus if it were adopted as a policy. This is particularly true for smaller urbanized areas, where the costs of removing severe congestion are in the range of 10 percent of plan costs.

Of course, some of the MPO expenditures are for congestion relief and might overlap with our estimates. Most of the expenditures planned by MPOs, however, serve other purposes. To the extent that their goals include congestion relief, this would offset some costs and rising
construction prices. But in most cases, their plans would not eliminate severe congestion or even reduce congestion below current levels.

Table 23 aggregates the costs in Table 22 by city size and divides them out by commuter trips over 25 years. The table shows that the cost per commuter trip to relieve LOS F congestion ranges
between $0.65 and $0.13 per trip. Although this estimate is just for those cities for which we
reviewed the long-range plans, it gives a feel for the range of values.

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 27

Table 22: Costs to Relieve Congestion versus Present Plan Costs

Urbanized 2003 UA LRP Total Auto LRP Total Transit LRP Total LRP Cost to Percent of

Area Pop K Highway Commute Transit Commute Other Plan Cost Relieve LOS Total LRP

Costs Share** (%) Costs Share*** (%) Costs F Congestion

New York 17,170 $78.7 B 38.9 $249.0 B 48.7 $0 $327.8 B 38.5 B 11.8

Los Angeles 12,520 48.5 B 88.0 66.9 B 4.8 0 115.4 B 67.7 B 58.7

Chicago 7,702 33.5 B 81.3 27.5 B 11.9 0 61.0 B 53.9 B 88.3

Philadelphia 5,287 21.9 B 82.3, 22.8 B 10.0 12.7 B 57.4 B 19.6 B 34.1

Miami 5,104 6.0 B 90.7 13.3 B 3.9 0 19.3 B 30.0 B 155.6

Dallas-FW 4,312 30.6 B 92.2 13.5 B 2.1 1.0 B 45.1 B 26.1 B 58.0

Wash. D.C. 4,227 36.9 B 46.6 56.4 B 38.8 0 93.3 B 16.2 B 17.4

San Fran. 4,120 42.0 B 80.8 76.0 B 9.2 0 118.0 B 29.2 B 24.8

Boston 3,988 4.5 B 82.7 43.8 B 9.4 0 48.3 B 20.3 B 42.1

Detroit 3,939 31.5 B 93.4 9.3 B 1.6 0.2 B 41.0 B 24.1 B 58.7

Seattle 2,946 49.4 B 81.9 46.3 B 7.9 5.9 B 101.6 B 4.8 B 4.7

Atlanta 2,924 29.6 B 90.7 21.5 B 3.0 1.9 B 53.0 B 13.1 B 24.6

San Diego 2,872 8.1 B 88.1 15.9 B 3.9 8.3 B 32.2 B 10.1 B 31.5

Houston-Galv 2,620 46.7 B 91.2 17.9 B 3.3 12.7 B 77.3 B 9.2 B 11.9

Minneapolis 2,482 5.6 B 75.0 2.6 B 14.2 0.7 B 8.8 B 7.6 B 85.9

Baltimore 2,076 13.2 B 72.3 11.8 B 18.2 0.5 B 25.5 B 1.8 B 7.2

Denver 2,050 53.9 B 88.4 23.4 B 4.1 10.5 B 87.8 B 10.0 B 11.3

Portland OR 1,685 14.2 B 74.5 5.5 B 12.9 na 19.7 B 2.7 B 13.7

San Jose 1,664 1.1 B 90.3 6.9 B 2.8 0.6 B 8.5 B 1.3 B 15.0

Cincinnati 1,606 5.7 B 83.6 1.6 B 8.2 0.1 B 7.4 B 0.6 B 8.5

San Antonio 1,333 6.5 B 91.4 4.0 B 2.9 0 10.5 B 5.6 B 53.7

Columbus OH 1,195 4.0 B 92.2 1.3 B 2.7 0.2 B 5.4 B 1.5 B 27.1

Salt Lake 877 3.2 B 91.4 17.3 B 2.4 2.4 B 23.0 B 1.2 B 5.4

Austin 757 15.4 B 92.2 6.2 B 1.8 0.3 B 21.9 B 2.5 B 11.3

Charlotte 725 1.2 B* 91.1 6.3 B 2.6 5 M 7.6 B 2.9 B 38.3

Tucson 720 9.6 B 87.8 3.4 B 3.0 1.4 B 14.4 B 1.0 B 6.8

El Paso 629 4.4 B 93.6 1.8 B 1.9 0 6.2 B 1.4 B 23.0

Akron 614 2.3 B 91.1 0.3 B 3.4 62 M 2.7 B 0.3 B 9.7

Raleigh 528 5.7 B 93.1 2.2 B 1.9 3 M 7.9 B 3.3 B 41.2

Bakersfield 443 4.2 B 88.2 1.4 B 2.1 15 M 5.7 B 0.4 B 7.4

McAllen TX 376 3.9 B 93.1 72 M 0.0 42 M 4.0 B 0.6 B 16.2

Spokane 357 1.2 B 88.8 1.4 B 1.2 0.4 B 3.0 B 1.4 B 47.0

Little Rock 338 2.8 B 95.6 0.8 B 0.7 25 M 3.6 B 2.3 B 63.8

Corpus Christi 295 741 M 93.4 163 M 0.3 0 904 M 862 M 95.4

Boise 254 2.2 B 91.5 na 0.8 na 2.2 B 0.3 B 12.4

Eugene 239 1.2 B 83.2 0.8 B 2.1 28 M 2.0 B 0.2 B 11.4

Lincoln 227 1.5 B 90.9 na 1.2 na 1.5 B 0.1 B 7.8

Lubbock 206 0.9 B 95.5 0.2 B 0.1 0 1.1 B 0.2 B 13.7

Fredericksburg 168 2.1 B -- na -- na 2.1 B 0.3 B 13.8

Binghamton 137 690 M 87.2 130 M 1.4 0 820 M 132 M 16.1

Sioux City 108 609 M -- 121 M -- 0 730 M 21 M 2.8

Missoula 74 118 M -- 66 M -- 21 M 205 M 23 M 11.3

Elmira NY 57 658 M -- 130 M -- 4 M 792 M 26 M 3.2

Total 101,951 $636.7 B NA $780.0 B NA $60.0 B $1,476.7 B $413.4 B 28.0

*A $3.57 billion road plan has recently been proposed.

** Based on 2003 American Community Survey Data for metro area. Combines single occupancy driving and carpooling ***Based on 2003 American Community Survey Data for metro area. Includes taxi.

28 Reason Foundation

Table 23: Average Costs per Resident to Relieve Congestion versus Present Plan Costs,

Weighted by Population

City Size Average Average Average LRP Average Cost to Average Cost

Population LRP Plan Cost per Relieve LOS F to Relieve

(000) Cost ($ B) Commuter Congestion ($ B) Congestion, per

Trip ($) Commuter Trip

3+ Million 9,503.4 137.1 2.31 38.9 0.65

1-3 Million 2,294.7 42.6 2.97 6.5 0.45

500K-1M 708.8 12.9 2.90 1.8 0.40

250K-500K 354.3 3.5 1.56 1.0 0.44

50K-250K 179.7 1.4 1.21 0.2 0.14

*Based only on cities reporting in Table 22

Such funds might be available through the reallocation of expenditures within the plans, if

congestion relief were given higher priority. But the challenge of adding capacity to reduce

congestion is not the focus of most MPOs. Most appear to have adopted a policy that congestion will be addressed by other means and at a lower priority.

The following are typical of comments that propose to reduce congestion by non-capacity means:

Eugene, Oregon

"The intent is to defer motor vehicle capacity increasing transportation system improvements until existing constraints can be overcome or develop an alternative mix of strategies (such as land use measures, TDM, short-term safety improvements) to address the problem."

"Encouraging the use of transportation modes other than the single-occupant vehicle will become more important as the region grows and traffic congestion levels increase."

The most important goal is "identifying the means to reduce reliance on the automobile by increasing the transportation choices available in the region."

"The ability of the region to fund capacity-increasing roadway projects will be limited by other allocation decisions."³⁰

— Lane Council of Governments, Central Lane Regional Transportation Plan

Jacksonville, FL

"To give priority to improvements that do not require additional travel lanes." "To expand and enhance pedestrian and bicycle access to all areas."

"To coordinate with the region's congestion management system in relieving existing congestion and preventing congestion where it has not yet occurred."³¹

—First Coast MPO, 2030 Long-Range Transportation Plan

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 29

Cincinnati, Ohio

"Traditionally, the solution to congestion has been to expand roadway construction capacity. It has become apparent, however, that metropolitan areas cannot build their way out of
congestion."³²

—OKI Regional Council of Governments,
OKI 2030 Regional Transportation Plan 2004 Update

Austin, Texas

While Austin has a Congestion Management System, none of the transportation improvement

strategies include adding more lanes. The improvement strategies include upgrading traffic signals,
modifying bus routes, installing reversible travel lanes, and promoting alternative transportation
modes (ridesharing, transit, bicycling, walking, etc.).³³ Austin has $262M of bike projects in the
2030 LRP.

—Capital Area MPO, Capital Area MPO Mobility 2030 Plan

Other cities place low priority on congestion reduction:

Bakersfield, California

Does not specifically list congestion relief as a goal.

"Delay the need for future increases in highway capacity and congestion relief through implementation of Transportation Control Measures."

"Promote sustainable community design that supports transit use and increases non-

motorized transportation while still meeting the mobility needs of residents and employees."³⁴

—Kern Council of Governments, Long Range Transportation Plan

Charlotte, North Carolina

"The transportation industry is giving more attention to safeguarding the natural

environment, and construction practices have changed to improve the way transportation projects affect their surroundings."

"The increased demands on transportation funding continue to create a backlog of unfunded projects…."

"MUMPO's plan is to increase choices in transportation….."³⁵

—Mecklenburg-Union MPO, 2030 Long Range Transportation Plan

Columbia, South Carolina

The congestion management goals do not include adding lane capacity. The primary goal is "to develop a means to reduce traffic demand….by reducing the percentage of single occupancy vehicles and promoting public transit."³⁶

— Columbia Area Transportation Study, Long-Range Inter-modal Transportation Plan 2025

30 Reason Foundation

Raleigh, North Carolina

Out of four goals and several objectives within each goal, addressing congestion was goal #4 and the last objective: "Maximize the highway system efficiency using means other than adding
general-purpose traffic lanes."³⁷

—Capital Area MPO, Capital Area MPO Mobility 2030 Plan

Detroit, Michigan

Out of five strategies to reduce congestion, adding capacity is the last resort.

"Road widening to increase capacity when other strategies are not applicable or do not reduce congestion to an acceptable level."³⁸

—Southeast Michigan Council of Governments,
2030 Regional Transportation Plan for Southeast Michigan

The unmistakable impression one draws from reviewing these plans is that urbanized areas

transportation plans are not focused on congestion reduction, and to the extent that it is a priority, the approach is to increase choices for other modes rather than to provide additional highway
capacity.

A few cities appear to be assessing the implications of what congestion reduction might cost:

San Antonio, Texas

To reduce congestion from 1.47 Texas Congestion Index (TCI, similar to TTI—Travel Time Index) to 1.12, "the region must find an additional $8 billon in funding."³⁹

—San Antonio & Bexar County MPO,

Mobility 2030 San Antonio Mobility-Bexar County Metropolitan Transportation Plan

Atlanta

In perhaps the most stunning case of attention to congestion, Atlanta's Congestion Mitigation Task Force has passed resolutions calling for raising the weight placed on congestion relief from 11 percent to 70 percent in project
selection, and setting a TTI value of 1.35 (presently 1.44) as a performance index for planning and project selection.⁴⁰

—Congestion Mitigation Task Force, Final Report and Recommendations

Texas

As noted above, the Governor's Business Council has adopted a goal of 1.18 for the Texas Congestion Index (similar to the TTI) for Texas urbanized area, and estimated the needed additional funding at $54 billion to achieve it.⁴¹

—Governor's Business Council Transportation Task Force

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 31

Washington

"In Washington…the growth of travel demand has outpaced expansion of …system

capacity…leaving .a growing backlog of capacity needs. The …imbalance affects … daily lives and almost every sector of economic activity"⁴²

—WSDOT Transportation Commission,

Urban Areas Congestion Relief Analysis Work Progress Report

32 Reason Foundation

P a r t 4

User Benefits of Congestion Reduction

sers benefit from congestion reduction in many ways, but even looking at just a few of

them—reduced travel time, lower operating costs and lower accident costs—shows the

benefits can be substantial. The data below are taken from two sources. First, several of the 32

cities that provided detailed data for the fine estimates made in this study included additional

detailed information from their traffic assignments which permits calculation of user benefits. This calculation is relatively conservative and does not include the savings from wasted fuel (that extra fuel consumed during the slower speeds or the stop-and-go conditions of congested travel). These additional fuel costs are estimated to be about 8 percent of the delay costs in the 2005 Urban
Mobility Report, which is based on 2003 data.⁴³ These costs could be as high as 10 percent of
delay costs, if current gasoline prices are used. Second, a more general measure—cost per hour of delay saved—is calculated for each urbanized area by estimating total delay saved by commuters over 25 years, and comparing that to estimated costs of congestion relief.

We use this data to show in some detail the benefits, using Detroit and Atlanta, two of the cities
that provided additional traffic assignment data. These two cities are quite different—south v.
north, fast growing v. slow growing, high congestion v. less congested, etc—and so the benefit
results from them give an idea of the range possible. We also show the benefits for all cities by
size.

Table 24 shows the findings for Detroit. The expansion of capacity to eliminate LOS F congestion results in an 11.6 percent reduction in vehicle-hours of travel and a 12.2 percent increase in speed. Note that the total regional travel, about 155 million miles daily, is not changed significantly
(induced travel might increase this total somewhat). The travel time savings amount to about
531,000 hours daily.

User benefits are shown in the bottom of the table. At $10/hour (a conservative estimate), the value
of the time saved traveling without severe congestion over 20 years is about $26.6 billion.⁴⁴
Savings from fatal accidents and operating costs increase the total to about $30.3 billion.
Compared to the estimated cost of the added lane-miles for Detroit ($24.1 billion) the results
appear to be mildly positive (i.e., the benefit/cost ratio in greater than 1). Of course these benefits
do not include the other benefits, such as increased choices, or increased access to goods and
services. Detroit is not a particularly congested place, so the benefits may be understated.

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 33

Table 24: User Benefits Analysis for Detroit

Daily vehicle miles traveled Daily vehicle hours Average

(VMT) (Millions) traveled (VHT) (Millions) Speed

2030 LRP Assignment 157.774 4.593 34.35

2030 Unconstrained 156.575 4.062 38.55

Diff -1.199 -0.531 4.2

% Diff -0.76 -11.57 12.23

Savings: Miles 1.199

Savings: Hours 0.531

Life Savings Years 20

Days/year 250

Value of Time $10.00/hr

Operating cost/mile $0.60

Ave fatal accident cost 3.0 M

Ave acc rate/100 M VMT 1.5

Lives saved over 20 years 89.92

Lifetime value of travel time saved $26.6 billion

Lifetime value of operating cost saved $3.4 billion

Lifetime value of lives saved $269.8 M

Total savings over 20 years $30.3 billion

Cost to Relieve LOS F Congestion $24.1 B

Benefit/Cost Ratio 1.26

Table 25: User Benefits Analysis for Atlanta

AM and PM vehicle miles AM and PM vehicle hours Average

traveled (VMT) (Millions) traveled (VHT) (Millions) Speed

2030 LRP Assignment 103.899 5.942 17.49

2030 Unconstrained 103.268 4.328 23.86

Diff -0.63 -1.61 6.37

% Diff -0.61 -27.16 36.46

Savings Miles 0.63

Hours 1.61

Life Savings Years 20

Days/year 250

Value of Time $12.00/hr

Operating cost/mile $0.60

Ave fatal accident cost 3.0 M

Ave acc rate/100 M VMT 1.5

Lives saved over 20 years 47.25

Lifetime value of time saved $96.6 billion

Lifetime value of operating cost saved $1.9 billion

Lifetime value of lives saved $141.8 M

Total savings over 20 years $98.6 billion

Cost to Relieve LOS F Congestion $13.1 B

Benefit/Cost Ratio 7.53

Table 25 contrasts Detroit against a more congested city, Atlanta, showing results for peak-period
travel only. The expansion of that city's network to eliminate severe congestion would result in a

34 Reason Foundation

27.2 percent reduction in peak-period travel times and a 36.5 percent increase in peak-period

speed, saving regional commuters about 1.61 million hours a day in travel time. At $12 an hour (also conservative, given Atlanta's higher average earnings), this time savings would be valued at $96.6 billion over 20 years. Savings from reduced operating costs and reduced fatal accidents would bring the total savings to $98.6 billion. Compared to the implementation costs of $13.1 billion, this is a substantial benefit. Even if the benefits are over-stated and the costs understated, the benefit-cost ratio is likely to be highly positive.

These examples are probably extreme. Detroit represents a "low-case" situation for a region with relatively high construction costs and relatively low congestion. Atlanta represents lower
construction costs and higher congestion. Nevertheless they represent the range of results one
might expect from a more detailed assessment of the impacts of congestion removal on traffic flow and particularly on travel time savings.

To estimate benefits for each region we also developed an estimate of the cost per hour of delay saved for each urbanized area. Appendix C shows the detailed results for each region, which are summarized in the following table.

Table 26: Cost of Capacity Expansion Per Hour of Delay Saved

Urban Area Size Average Annual Delay Total Cost over Average Cost per Hour of

Saved, Hours 25 years, ($M) Delay Saved ($)

3+ M 4,780,230,762 325,599.10 2.72

1-3 M 2,151,708,742 98,185.99 1.83

500K-1M 515,696,950 48,123.03 3.73

250-500K 154,626,517 40,359.34 10.44

50-250K 131,988,660 21,229.19 6.43

Total 7,734,251,631 533,496.60 2.76

Nationwide, commuters would save about 7.7 billion hours of delay annually. The cost of capacity increases per hour saved, averages about $2.76, with larger regions showing lower cost per hour of delay saved. In Appendix C, only a handful of urban regions have costs per hour of delay saved greater than $20. These numbers compare favorably with the federal guidelines for transit 'New Starts,' currently $21/ hour of "transportation benefit."

Additional savings, not quantified here, would include lower fuel use, reduced accident rates and
vehicle operating costs, shipping costs and truck travel time reductions, and greater reliability.

In addition to these savings, which are real for commuters, communities also gain significantly by
more accessibility. Workers have more choices of employment within a given travel time of home,
shoppers have a greater range of markets and products, and employers have a larger pool of
workers for jobs. These real benefits result in lower costs of goods and services to urban societies.
Although some travel time savings are lost by traffic attracted to faster routes (thus lowering
speeds), for most capacity improvements, this effect is minor. Our work in North Carolina
studying over 300 road improvement projects estimated that about 92 percent of the additional
capacity provided by road improvements would be converted into travel time savings and greater
choices.⁴⁵

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 35

P a r t 5

Conclusions

A. The Magnitude of Present and Future Congestion

U.S. urban area population is expected to increase about 30 percent over the next 25 years. As

Table 27 shows, that means that the number of congested lane-miles of urban roads will increase 51 percent. Increases in congestion will come not because Americans are wasteful drivers or
extravagant, but simply because there will be more people living in urbanized areas, competing for already limited road space in peak travel times.

Urban congestion will continue to increase unless significant action is taken to reduce it. Traveler delay in urbanized areas will double over the next 25 years, with the greatest relative increases coming in the smaller regions.

Table 27:
Severely Congested Facilities, 403 Urbanized Areas

Roadway Type 2003 Lane-Miles 2030 Lane-Miles Percent

Severely Congested* Severely Congested* Change

Urban Interstate 17,800 27,400 54.3

Urban Other Freeway

Urban Other Principal Arterial 9,000 12,400 37.5

Urban Minor Arterial 12,700 19,900 56.1

Urban Collector
Urban Local

Total 39,500 59,700 51.1

Severely congested= facilities for which peak-hour traffic volumes exceed capacity * Rounded to nearest 100 for convenience.

B. Capacity Needs to Eliminate Severe Congestion

Increasing congestion cannot be confronted by just one strategy alone, but must be dealt with using
a variety of actions. Capacity increases, road pricing, information systems technologies, incident
and accident management, traffic operations, signal optimization, and—where justified—better

36 Reason Foundation

transit service, can all be part of the mix. Options should be carefully evaluated for effectiveness, based on cost per hour of delay saved.

However, the key longer-term strategy is likely to be increased highway capacity. This is because only significant increases in highway capacity, combining added infrastructure and more efficient operation, provide the means of keeping up with projected growth of population and traffic. It will require 104,000 additional lane-miles in our urban areas—about 6.2 percent of the current urban lane-miles—to eliminate severe congestion.

C. The Cost of Dealing with Congestion

In the 403 largest urbanized areas of the United States, about $533 billion in 2005 dollars will be needed over the next 25 years to deal with the most severe congestion.

These costs are $21 billion per year over 25 years. They are about 15 percent of likely government expenditures for highway transportation over the next 25 years. The cost is about 28 percent of
what the urbanized areas already plan to spend for their transportation plans. On a per-trip basis,
the costs range from as little as 9 cents/trip for smaller cities to about 75 cents/trip for large cities.
Some of these funds are already in those plans, since each plan has some congestion-reduction
funding in it already.

Severe congestion needs are spread throughout the United States. Although the 10 largest

urbanized regions account for about 61 percent of needs, congestion relief is important not just for
big cities. All states and all cities have legitimate interests in reducing congestion, because
congestion reduces their competitiveness and increases people's costs of daily travel. The cost per
hour of delay saved, averaging $2.76, is generally lower in larger regions because traffic and delay
is so much higher there. However, even in smaller regions the cost is quite low per hour of delay
saved.

If moderate levels of urban congestion are also addressed, another $270.5 billion will be needed. If rural congestion—a growing phenomenon—is also addressed, another $14.2 to $33.9 billion will be needed, depending on what magnitude of congestion is addressed. These findings are
summarized in the following table.

Table 28: Summary of Needs and Costs

Region Severe Congestion Moderate Congestion Totals

Urbanized Areas Lane-Miles 104,220 Lane-Miles 84,070 Lane-Miles 188,290

Cost $533.4 Billion Cost $270.5 Billion Cost $803.9 Billion

Rural Areas Lane-Miles 16,354 Lane-Miles 23,429 Lane-Miles 39,783

Cost $14.2 Billion Cost $19.7 Billion Cost $33.9 Billion

Totals Lane-Miles 120,574 Lane-Miles 107,499 Lane-Miles 228,073

Cost $547.6 Billion Cost $290.2 Billion Cost $837.8 Billion

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 37

Some might ask, "won't we need another $533 billion for the following 25 years, and so on, and so
on? Won't traffic just keep growing?" After more than two decades of delayed investment, we
must first deal with the accumulated backlog of under-investment. Making this expenditure now
helps the United States maintain competitiveness over the 25-year horizon, but it does not remove
the responsibility to make more improvement beyond then. It is likely that additional
improvements might be needed in the next quarter century, and no one should think of the
'transportation congestion problem' as solvable with this expenditure. But once we catch up,
keeping up should not be as difficult or anywhere near as expensive. As long as we value private
mobility and freedom of choice, personal travel will continue to grow, though probably less rapidly
in the future. For hundreds of years the United States basically kept up with its mobility needs by
improving its transportation systems, and must continue to do so if it is to remain economically
healthy.

D. These Costs Are Reasonable Compared to Planned Transportation Spending

It is possible for America to 'build out' of severe congestion, and it is relatively inexpensive to do
so. The $533 billion estimate for relieving severe congestion ($21 billion/year) is about one-quarter
of the total cost of the 25-year transportation plans we reviewed, and is about 15 percent of the
total highway budget over 25 years; it would be a maximum of about 30 percent increase in capital
expenditures, if all current capital work were not capacity-increasing. It is about three times what is
now being spent
annually on 'new starts.' Moreover, at $2.76 per hour of delay saved, the cost is
considerably lower than most comparable costs for other transportation improvements.

Looked at another way, on a per-trip basis, the costs range from as little as 9 cents/trip for smaller
cities to about 75 cents/trip for large cities. Some of these funds are already in urban areas' long
range transportation plans, since each plan has some congestion-reduction funding in it already.

A real concern is whether additional urban highway capacity can be provided, given current

environmental constraints and often strong community opposition. The locations for much needed capacity may be in corridors with major right-of-way constraints. While, of course, each project must pass environmental screens, failure to plan or to evaluate the possibility just because a
widening is deemed unwanted or expensive is not good practice. Communities are seeing that other options may not be effective and that progress must nevertheless be made if worse congestion is to be avoided. Newer environmentally friendly and so-called 'context-sensitive' designs, and
innovative designs for constricted locations, can permit construction.⁴⁶

Options for pricing, in conjunction with new capacity provision, should not be overlooked. In

many locations where additional capacity is needed, the options for priced lanes or tolled-HOV

combinations may be feasible. Pricing will also help pay for some needed capacity and thus reduce
the burden on traditional funding sources. However, for portions of systems that are isolated or
arterial in character, location-based pricing may not be feasible. No mileage-based pricing
mechanisms are yet available that would permit 'universal' time-of-day pricing, and that might not

38 Reason Foundation

be permissible, even if it were feasible. In our view, pricing options remain an important part of the demand-supply equation in combination with capacity increases, but will not substitute for the need for additional capacity.

E. The Likely Benefits

The primary benefits of investing in congestion relief will flow to both users and non-users. Specifically:

Benefits to users will be primarily in the form of savings in travel time, with smaller benefits in
reduced operating costs and reduced accident costs. The examples for Detroit and Atlanta and the
cost per hour of delay saved for each region show that significant savings are likely if projects are
targeted at congestion relief, and that these savings are likely to be cost-effective. These real
savings put time and money in consumers' pockets and can be reinvested in other goods and
services that directly help the economy. Important secondary user benefits include increased
reliability of travel times, reduced travel stress, and improved reliability of activity scheduling.
Although these benefits have been traditionally more difficult to quantify, they are nevertheless
substantial and real.

Other non-user benefits are also substantial. They include significantly increased choices of both labor supply and consumer purchases through greater reach of destinations within travel times. This lowers the cost of goods and services through competition.

Benefits to businesses include reduced delivery costs through reduced travel times for trucking operations, lower operating costs and lower accident rates. Important secondary business user benefits include improved just-in-time delivery, reliability of shipments, smoother supply-chain management, and more regular production operation. These savings in lower logistical costs are often passed on to consumers through competition. In the aggregate they help to maintain
America's competitive edge in the global economy and make domestic transportation costs very 'flat' relative to other nations, a significant advantage.

Cities that reduce congestion also benefit substantially, through increased economic

competitiveness and lower costs relative to neighbors. In fact, cities that don't improve their access through reduced congestion costs may find their competitive edges slipping away to more
favorable locations.⁴⁷

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 39

F. Summary

A concerted and focused effort will be required to address the challenge of congestion relief. By
enumerating the problem, this report hopes to shed light on its magnitude and costs, but also finds
that the problem is tractable. By challenging the transportation community to act, we hope to re-
establish the vision of transportation excellence that brought the nation to its present state of
transportation quality. We envision that congestion can be significantly reduced and system
reliability improved by a combination of new capacity and pricing (particularly on the higher
systems), in combination with aggressive actions to improve system efficiency and operations. We
look forward to the day when the transportation community proudly replaces the phrase "We can't
build our way out of congestion" with the phrase, "We significantly reduced congestion."

40 Reason Foundation

About the Authors

Authors

avid T. Hartgen, Ph.D., P.E. is Professor of Transportation Studies at the University of

North Carolina at Charlotte, where he established the Center for Interdisciplinary

Transportation Studies and now teaches and conducts research in transportation policy. He is the
author of about 329 papers and reports on transportation policy and planning, is U.S. Editor of the
international journal Transportation, and is active in professional organizations, particularly the
Transportation Research Board. He holds engineering degrees from Duke University and
Northwestern University, has taught at SUNY Albany, Union University, Syracuse University and
lectures widely. He can be contacted at dthartge@email.uncc.edu, or by telephone at 704-687-
5917. His Web site is http://www.geoearth.uncc.edu/Dhartgen.htm.

M. Gregory Fields is a graduate student at the University of North Carolina at Charlotte pursuing masters degrees in Geography (Transportation), Earth Sciences (Environmental Monitoring) and Sociology, with plans to graduate in December 2006. A retired U.S. Army officer, he holds a BS degree from West Point and a master of arts from Webster University in St. Louis. He has
contributed to several transportation studies including a review of the cost-effectiveness of North Carolina's highway projects, county-level road condition trends in North Carolina, and a review of South Carolina's traffic modeling systems.

Project Director

Robert W. Poole, Jr. is Director of Transportation Studies at the Reason Foundation in Los

Angeles. He received B.S. and M.S. degrees in engineering from MIT and did additional graduate work in operations research at NYU. He worked in aerospace and for several research firms before launching Reason Foundation in 1978. His 1988 policy study, "Private Tollways: Resolving
Gridlock in Southern California," directly inspired California's 1989 public/private toll roads law, which has been emulated in more than a dozen other states. He has advised the U.S., California,
and Florida departments of transportation, and served 18 months as a member of California's
Commission on Transportation Investment. He has also advised the last four White Houses on
various transportation policy issues.

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 41

Endnotes

1 David Shrank and Tim Lomax, The 2005 Urban Mobility Report, Texas Transportation

Institute, College Station, TX, May 2005. Available at http://mobility.tamu.edu.

2

Federal Highway Administration, Our Nation's Travel, Washington DC 2005, p. 16-17.

3

4

5

6

7

8

9

10

11

12

Federal Highway Administration, "Safe, Accountable, Flexible, Efficient Transportation

Equity Act: A Legacy for Users (SAFETEA-LU): A Summary of Highway Provisions", Washington DC 20590, August 10, 2005.

American Highway Users Alliance, Unclogging America's Arteries: Effective Relief for
Highway Bottlenecks, 1999-2004, One Thomas Circle NW, 10^th Floor, Washington DC 20005. February 2004.

US Department of Transportation, 2002 Status of the Nation's Highways, Bridges and Transit: Conditions and Performance, Washington DC 20590, 2004, p. 7-6.

Governor's Business Council, Shaping the Competitive Advantage of Texas Metropolitan Regions: The Role of Transportation, Housing and Aesthetics, Governor's Business Council Transportation Task Force, Austin, TX, Draft October 2005.

TDA Inc., End Gridlock Now, Our Transportation Mess: Truth and Facts versus Myths and
Baloney, Seattle WA May 2, 2003; and WSDOT Transportation Commission, Urban Areas
Congestion Relief Analysis Work Progress Report Briefing Paper, Olympia WA, Feb. 2005.

Governor's Congestion Mitigation Task Force, Final Report and Recommendations, Atlanta, GA Dec. 6, 2005.

Transportation Research Board, Critical Issues in Transportation, National Academy of Sciences, Washington, DC, December 2005.

Urbanized areas are used as the 'geography' in the study because they form the basis of formal urban transportation planning, published historical traffic and network data is readily available, and the Texas Transportation Institute uses them for congestion monitoring. Although this data (primarily the HPMS database) has some problems, primarily an expanding geography over time, it is significantly better as a data set than others.

Severe congestion is defined as that occurring when the peak-hour volume of traffic on a road
is greater than the carrying capacity of the road, that is, the volume/capacity ratio is greater
than 1.0. Moderate congestion is defined as occurring when the peak-hour volume of traffic on
a road is between 80 percent and 95 percent of the carrying capacity of the road. These
definitions correspond closely to Levels of Service F and E-D, respectively, in the widely used
Highway Capacity Manual. See Appendix B for a more complete discussion.
Technically, 'gridlock' refers to cases in which connected streets are jammed by traffic.

42 Reason Foundation

¹³ 85 regions for which congestion data is available from the Texas Transportation Institute, plus

Durham, NC, which had been consolidated with Raleigh NC in the TTI reports.

¹⁴ The Highway Performance Monitoring System is a federally-managed data reporting system in

which the states annually report conditions and performance of about 116,700 sections of

roadway nationwide. The data is used in a wide range of state and national reports and is the basis of many tables in Highway Statistics, FHWA's annual report. HPMS is the only state-to-
state consistent report of highway conditions.

¹⁵ A 'vehicle-mile-of-travel' (VMT) is a commonly used measure of travel, defined as one

vehicle traveling one mile.

¹⁶ 'Induced travel' represents travel shifted from other modes, time periods, days, and possibly

additional land uses increases as a result of the increased access of major road improvements. It is distinguished from 'diverted travel' which is traffic diverted from other routes to faster routes after widening. Both forms of additional travel are accounted for in this study.

¹⁷ Numbers in the text are rounded for convenience. See Appendix C for specific numbers for

urbanized areas.

¹⁸ The 'Travel Time Index' is computed from the sample of HPMS road sections in the region.

For each section, data is obtained annually on average daily traffic, section length, posted

speed, peak hour traffic and roadway characteristics. Based on capacity estimates, operating
speeds and running travel times for peak and off peak are then estimated for each section. The overall index is then computed by weighting these travel times by the traffic on each section. The Index can be thought of as how long, relatively, it would take to travel a typical road
section in the peak hour versus the off-peak. Other statistics (delay, speed, fuel) are then
estimated from this data.

¹⁹ The amount and cost of 'induced travel' is controversial, ranging from estimates near zero to

near 100 percent in some studies. Anthony Downs refers to 'triple convergence' to describe diversion to widened roads from other routes, time periods and modes; these are generally handled quite well by current forecasting methods. New traffic caused by regional growth is not handled so well. It is this form of change that most planners mean by 'induced travel'. To be (conservatively) high, we estimated increases of about 15 percent for large regions and smaller amount for smaller regions. See Appendix B for details.

²⁰ Samuel, Peter, Innovative Highway Design, Report for the Mobility Project, Spring 2006.

²¹ If 'year of construction' estimates are used, nominal costs would be higher. We estimate that

about $692 billion (in 'year of construction' dollars) would be needed.

²² USDOT, 2002 Conditions and Performance Report, op. cit.

²³ W. Cox and A. Pisarski, Blueprint 2030: Affordable Mobility and Access for all of Atlanta and

Georgia, Report to the Governor's Mitigation Task Force, Atlanta GA, June 21, 2004, p. 6.

²⁴ Governor's Business Council Transportation Task Force, o. cit., p. vii.

²⁵ Recent work by FHWA looking at unit costs suggests that the previous numbers used in HERS

double-count the individual state adjustments and that overall costs in some categories are

substantially higher. Alam M Timothy D and Sissel S., New Capital Cost Table for Investment Economic Analysis, Record 1932, Transportation Research Board, Washington, DC 20590, p. 33-42. And recent cost increases for construction components may have raised costs
substantially at least temporarily.

BUILDING ROADS TO REDUCE TRAFFIC CONGESTION 43

²⁶ Federal Highway Administration, Highway Statistics, 2003, Table HF10, Disbursements for

highways, all levels of government, Washington DC 2004.

²⁷ Traffic planners describe roads by 'functional class' referring to the intended purpose that

various roads are intended to serve. For instance, traffic on 'urban interstates' are more likely to be traveling longer distances for commuting between regions or states; and traffic on 'urban principal arterials' would be more likely to be a mixture of land-use-oriented and short-
distance intra-urban traffic.

²⁸ These might be compared with the criteria that the Federal Transit Administration uses in

evaluating 'New Starts' projects, currently $21.00 per hour of transportation benefit.

²⁹ The Road Information Program (www.trip.org) has recently reviewed the extent of

'recreational congestion' and identified some of the nation's worst 'recreational congestion'

hotspots.

³⁰ Lane Council of Governments, Central Lane Regional Transportation Plan, Chapters 1, pages

1,4,& 5; Chapter 2, pages 4, December 2004, www.thempo.org

³¹ First Coast MPO, 2030 Long-Range Transportation Plan, Goals & Objectives, Goal C, pages

1-2, 2004, www.firstcoast2030/com.

³² OKI Regional Council of Governments, OKI 2030 Regional Transportation Plan 2004

Update, Congestion Management Strategies, page 6-8, 2004, www.oki.org

³³ Capital Area MPO, Capital Area MPO Mobility 2030 Plan, Comparison of 2025 AMATP &

Recommended CAMPO Mobility 2030 Summary Presentation, 2005, www.campotexas.org.

³⁴ Kern Council of Governments, Long Range Transportation Plan, Chapter 2, Transportation

Planning Policies, page 6, 2004, available at www.kerncog.org.

³⁵ Mecklenburg-Union MPO, 2030 Long Range Transportation Plan, 2005, Chapter 7.0,

Financial Plan, page 7-1, www.charmeck.org

³⁶ Columbia Area Transportation Study, Long-Range Inter-modal Transportation Plan 2025,

2003, Chapter 1, page 4.

³⁷ Capital Area MPO 2030 Long Range Transportation Plan, Raleigh, NC. September 15, 2004.

www.campo-nc.us/lrt/2030.

³⁸ Southeast Michigan Council of Governments, 2030 Regional Transportation Plan for

Southeast Michigan, November 2004, page 13, www.semcop.org

³⁹ San Antonio & Bexar County MPO, Mobility 2030 San Antonio Mobility-Bexar County

Metropolitan Transportation Plan, 2004, Chapter 2, page 6.

⁴⁰ Congestion Mitigation Task Force, Final Report and Recommendations, GDOT, December 6,

2005.

⁴¹ Governor's Business Council Transportation Task Force, op.cit.

⁴² WSDOT Transportation Commission, Urban Areas Congestion Relief Analysis Work Progress

Report, Olympia WA: February 2005.

⁴³ Schrank, David and Tim Lomax. The 2005 Urban Mobility Report. College Station, TX: The

Texas Transportation Institute, Texas A&M University System, May 2005. Available at:
http://mobility.tamu.edu/ums/report/.

44 Reason Foundation

⁴⁴ Traditional transportation benefit-cost assessments often value time at about ½ the prevailing

wage rate, assuming a mix of trip purposes. Since these are primarily congestion-related

benefits, a higher value, $10-12/hour, is more appropriate.

⁴⁵ Hartgen, David T., Highways and Sprawl in North Carolina, Charlotte, NC: University of

North Carolina at Charlotte, 2003. Available at:

http://www.johnlocke.org/policy_reports/display_story.html?id=41.

⁴⁶ Peter Samuel, Innovative Roadway Design: Making Highways More Likable, Reason

Foundation, forthcoming, 2006.

⁴⁷ Contrary to popular belief, taxes that governments receive from increased economic activity

are not a benefit since they represent a transfer of resources from the private sector to the

public sector; if left in the private sector they might have been equally or more productive to the economy.

Mobility Project Advisory Board

The Mobility Project Advisory Board provides overall program guidance, suggestions on research,
and feedback on studies. It does not necessarily endorse the conclusions of individual studies.

n Rob Atkinson, Vice President, Progressive

Policy Institute

n Al Appleton, Senior Fellow, Regional Plan

Association

n Peggy Catlin, Director, Colorado Tolling

Enterprise and Deputy Executive Director, Colorado Department of Transportation

n Professor Robert Cervero, College of Envi-

ronmental Design, University of California, Berkeley

n Professor Randall Crane, Urban Planning,

University of California, Los Angeles

n Professor Elizabeth Deakin, City and

Regional Planning, University of California, Berkeley, and Director, University of Cali-
fornia Transportation Research Center

n Professor Max Donath, Director, Intelligent

Transportation Systems Institute, Univer-
sity of Minnesota

n Robert Dunphy, Senior Resident Fellow

Transportation, Infrastructure, Urban Land Institute

n James Ely, Director, Florida Turnpike

Enterprise and Vice President, International Bridge, Tunnel & Turnpike Assn

n Professor David Gillen, Transportation

Policy, University of British Columbia

n Professor Genevieve Giuliano, Director,

METRANS Transportation Center, Univer-
sity of Southern California

n Professor Peter Gordon, School of Policy,

Planning, and Development, University of Southern California

n Gary Groat, Director of Project Develop-

ment, Fluor

n Professor David Hartgen, Transportation

Studies, University of North Carolina, Char-
lotte

n Tony Kane, Director, Engineering and

Technical Services, American Association of State and Highway Transportation Officials (AASHTO)

n Steve Lockwood, Principal Consultant,

PBConsult

n Jim March, Industry and Economic Analysis

Team Leader, Office of Policy and Govern-
mental Affairs, Federal Highway Adminis-
tration

n Joel Marcuson, PE, Traffic Engineering/

Intelligent Transportation Systems Expert, Jacobs Engineering

n Nancy McGuckin, Travel Behavior Analyst,

Consultant

n Professor Michael D. Meyer, School of Civil

& Environmental Engineering, Georgia Institute of Technology

n Professor James Moore, Department of

Industrial and Systems Engineering, Uni-
versity of Southern California

n John Njord, Executive Director, Utah DOT

n Ken Orski, Principal, Urban Mobility Corpora-

tion

n Mary Peters, Former Administrator, Federal

Highway Administration

n Alan Pisarski, Transportation Consultant

n Steve PonTell, President, LaJolla Institute

n Pete Rahn, Director, Missouri Department

of Transportation

n Jon Ramirez, Senior Vice President, Cofir-

oute USA

n Darrin Roth, Director of Highway Opera-

tions, American Trucking Associations

n Gabriel Roth, Civil Engineer and Transport

Economist

n Tom Rubin, Transportation Consultant

n Phillip Russell, Director, Texas Turnpike

Authority Division

n Peter Samuel, Transportation Consultant

n Professor Kenneth Small, Department of

Economics, University of California, Irvine

n Professor Pravin Varaiya, Department of

Electrical Engineering and Computer Sci-
ences, University of California, Berkeley

n Professor Chelsea C. White III, Schneider

National Chair of Transportation and Logis-
tics, Georgia Institute of Technology

Reason Foundation

3415 S. Sepulveda Blvd., Suite 400
Los Angeles, CA 90034

310/391-2245

310/391- 4395 (fax)
www.reason.org