A Model to Support Stakeholder Evaluation of Transportation Policy Options in Las
Vegas, Nevada
Krystyna A. Stave and Richard E. Little!
Abstract
This paper describes a group model building process conducted to help a stakeholder
advisory committee develop policy recommendations for a regional transportation agency. Over
seven months, we worked with committee members and the transportation authority’s technical
staff to develop a strategic-level model. The model reflects the group’s definition of the problem
and collective understanding of the system. The group’s goal was to improve traffic congestion,
flow and air quality over a 25-year planning horizon at the lowest system cost. Their final
recommendation included a mix of policies with an estimated cost of $2.5 billion that achieved
the greatest air quality benefit and significantly reduced congestion and increased flow.
Participants were very positive about the model-based process, saying it was faster and “less
painful” than similar processes they had participated in. This paper discusses several issues
including the role group model building played in the stakeholder involvement process,
limitations of the model, constraints on the process and factors that were critical to the success of
the process.
Keywords: transportation policy; air quality; traffic congestion; group model-building, public
policy; public participation
Introduction and Problem Statement
Traffic congestion, transportation problems and air quality issues are concerns for
metropolitan areas all over the world. Las Vegas, Nevada, one of the fastest growing
metropolitan areas in the U.S., is no exception. In the last decade, the population has grown by
approximately 5% per year, from 780,000 in 1990 to over 1.3 million in 2000 (CCA, 2000).
Traffic congestion has increased significantly and traffic-related air quality has worsened. Air
quality degradation affects human health in the region and also feeds back to affect the
transportation system because of the connection between compliance with federal air quality
standards and federal funding for transportation projects. If air quality does not meet federal
standards, the Las Vegas region’s $80 million per year of transportation funding is in jeopardy.
In December 2000, the Board of Directors of the Regional Transportation Commission
(RTC) convened an advisory group of 30 system stakeholders. The group was to meet once per
month for a year. Their charge was to make recommendations to the Board at the end of the year
Assistant Professor and Graduate Research Assistant, respectively. Department of Environ mental Studies,
University of Nevada Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154-4030 (E-mail for corresponding
author Stave: kstave@ccmail.nevada.edu)
about how the RTC should address the region’s transportation problem. The advisory group
included elected officials, representatives of the business community and tourism industry,
environmentalists, bus riders, other public agencies, and community residents. Participants had
no particular knowledge of the transportation system other than their observations as system
users.
After initial meetings discussing the way the transportation system worked and the
group’s perception of what constituted the transportation problem, we were brought in to
facilitate a group model building process. We had a team of three people. One, with
considerable public participation experience, served as a liaison with the advisory group; the
other two were the modelers. One of the modelers facilitated the group model-building process
with the advisory committee members and the other worked closely with the RTC technical staff.
The group model-building process is discussed in greater detail in a companion paper in the
System Dynamics Review (Stave 2002). This paper focuses on the model itself and the group’s
use of the model for policy analysis.
Figures 1, 2, and 3 below show the problematic trends in traffic congestion, system-wide
average speed (an indicator of traffic flow) and average daily carbon monoxide generated by the
transportation system. These provide the reference modes that define the transportation problem.
Figure | shows congestion, measured as system-wide traffic volume divided by system-wide
capacity steadily rising until it reaches approximately .8 in 2025. System-wide average speed
(Figure 2) decreases from about 35 miles per hour in 2000 to 30 miles per hour in 2025. Carbon
monoxide production also increases. The preferred trend, or goal, for congestion is to at least
maintain congestion at its 2000 level of 0.5, what the RTC estimates is a “free-flow” condition,
and if possible, find ways to decrease congestion below its present level. For system-wide
average speed, the preferred trend would be to maintain the current speed, or if possible, increase
it. Congestion and average system speed are measures of the ease of mobility in the region.
Worsening congestion and flow are undesirable. They affect not only the quality of life of the
region’s residents, but also affect the local economy. Air quality has a somewhat different role in
that worsening air quality can have direct consequences on the region through the loss of federal
funding for transportation projects and a loss of autonomy over population growth and land use
management. The region must maintain CO below the federal standard. These three trends both
define the problem and set the criteria for measuring the effectiveness of potential solutions to be
tested by the model.
Model Development
System Conceptualization
The group defined seven main problem dimensions (shown as circles in Figure 4) from
their initial list of concerns and several key measures of each dimension (hexagons in Figure 4).
We worked with a subgroup of eight participants to develop and refine a causal map of the
system from which the modelers created a basic stock and flow model. At this stage, we worked
=
oo
Congestion
Figure I. Reference
| mode for traffic
| ner congestion as
o
a
BR DD
a indicated by system-
| wide average volume
Average
a
ie}
over capacity.
System-Wide Daily
Volume/Capacity
o
1990 2000 2005 2010 2015 2020 2025
N
YEAR
Flow
vo
& Figure 2. Reference
5 — mode for traffic flow,
> a measured by system-
< wide avera ge speed.
» Fs
3 Data and projections
= by RTC.
=
gv
2
wo
a
1990 2000 2005 2010 2015 2020 2025
YEAR
Air Quality
Figure 3. Reference
— mode for air quality
as indicated by carbon
—— monoxide emissions.
3 | _$g-} ——
Emissions
Tons/Day
Carbon Monoxide
oos3d
ooo3nd
1990 2000 2005 2010 2015 2020 2025
YEAR
improve
Air
Quality
optimize
Costs
increase
use of
reduce The Alternative
Congestion |» r ortation Modes
Challenge
improve
improve eer
How improve Experience
Safety leva of
driver
frustration
Figure 4. Initial definition of problem dimensions and measures (Stave 2002).
closely with the technical staff of the RTC to confirm the model structure and to determine
values for system parameters.
In the course of discussions with the subgroup, full group and RTC technical staff, the
seven dimensions were simplified to only four: congestion, flow, air quality, and cost. Figure 5
shows an overview of the model’s causal structure.
System Structure
The model has seven subsectors containing 17 stocks as shown in Table 1. It was
developed using Vensim® DSS (Ventana Systems 1988-2000a), and runs in Vensim® Model
Reader (Ventana Systems 1988-2000b). Each subsector is represented on a different model
distance per trip
aN total travel
ric
erceived population _— “ \: ewe oat iy
attractiveness volume of
oflas Ve +. use ” mass personal
é: transit and — Vehicles
alternative -
modes
attractiveness of +
- ——— ih
caperity street and
+, alternative medifier
wal
fo + pe | >» ‘panty
number of
number of rail lane sales
miles, buses and i
routes
P munber 9 Pa maniber of lane
neyele routes s under
a construction CO per
number of vehicle mule
lane miles
ordered difference between
CO budget and co
amount generated “f——
4 uo No
+" Cost budget
Figure 5. Overview of causal structure. Variables in italics represent policy options. Shaded boxes are output
display variables (Stave 2002).
view. Figures 6 and 7 show the Demand and Capacity subsectors, respectively, for illustration.
The model runs from 1990 to 2025, by year, with a time step of one day.
Table 1. Summary of RTC3 Transportation model structure.
Subsector Stocks or Major Variables
Demand Las Vegas Population (1 stock)
Capacity Lane miles (6 stocks)
Mass Transit Buses and Bus routes (4 stocks)
Rail miles (2 stocks)
Alternative modes Bicycle routes (1 stock)
Traffic management Infrastructure (2 stocks)
Air Quality Carbon Monoxide (no stock, calculated)
Cost Cumulative cost (1 stock)
perceived attractiveness
of LY as a function of travel distance increase
population due to population increase
a IOKUP
effect of perceived 1990 to 2001 vehicle
Lvpepiahaon —«Svariaanas of = aivete ™ average occupancy rete new Vehicle
affects migration? onal vate distance per trip cae Coup ENC rate
of decrease actual vehicle
births occupancy ree
normal rate “Time?
of increase actual rate average # trips per
of decrease day per capita total effectiv
Las ¥i e:mile
actual rate“ ¢-—— Valley J —— ame Volume/ E an a
of increase people | Population! people ON Gapanty”
sowing xt moving out total travel i=
deaths
demand A 4
volume of. essai
vehicles on roads
average
lifetime avg. speed
volume of riders on
public transportation volume of use of
other modes system-wide avg.
speed LOOKUP
percent of travel satisfied percent of travel using
1 of travel satisfed b by other altemative modes___pioVan,truck,motorcycle
percent of travel satisfied by (walking, biking, ete)
mass transportation
Figure 6. RTC3 Model — Demand Subsector
maxinmum HS lane miles
max HS lane miles inc 2 time to construc
onstruction at one time ‘at can —e = + HS lane mile
Backlog of Hig} High Speed High Speed
<2 Speed Lane Mile > Ue Hil Lane Miles in
munber es ase tobe Constmcta} munber of HS lane | Consiniction] | naw HS lane miles Service
ae ae mules put into completed
construction HS lane mile
7 or ——" design capacity
. ime to begin .
time to add, mile constrecti HS lane mile <capacity mi
new miles munber already planned eee” emmaxirmam HS lane ep —* capacity modifier adding siz
to backlog HS lane mules added) RTC HS lane mul that can mpleted P VACTS
“7 plan LOOKUP 2 HS lane mile :
panbero? apctyeoaneaa Ne total effective
additional H pacity
dine gly tobe year to begin to installing ITS mules ost mile capacity
HS lane miles args
— Pe ae ae
no actions switch LS lane mile ctive capacity
capacity modifier
«time to begin Lane LS lane mile
LS lame wales ig Tal? COMStRUEHION> design capacity ie soa eae - to
max 3 inco population grow
nstruction at one time ¥.
Low Speed Low Speed
Backlog of Low] : ow Spee
DS Speed Lane Mile Lame Miles —— Lane Miles i | capacity modifier due to
muiper of L$ lane | to be Constmuctel amumber of L$ lane Construction} #7 low speed lane Service population population growth
mules added to miles put into miles completed increase LOOKUP
backlog pcopstiuction yatio
Sew a
er already p! max LS lane miles
add new LS lane miles added “+ RTC LS lane mil that can be
e plan LOOKUP. completed im 1 year
Figure 7. RTC3 Model — Capacity Subsector
Model Subsectors
Demand
The Demand subsector represents total travel demand as a function of population and
average number of trips per capita. It calculates the volume of personal vehicles on
roads, in vehicle-miles traveled, as the daily trips per capita times distance per trip,
vehicle occupancy rate, and percent of travel using personal vehicles. Distance per trip is
assumed to increase with population, representing the increasing spread of the
metropolitan area.
Capacity
The Capacity subsector calculates the effective number of lane miles in the system.
Effective system capacity is the total number of lane miles in service modified by friction
factors due to population growth (which decreases capacity) and traffic management
system infrastructure (which increases capacity). The model includes two kinds of lane
miles — high speed lane miles such as freeways and major arterials, and low speed lane
miles such as minor arterials and residential streets. The delay between the time that new
lane-miles are ordered and put into service is represented by a chain of three stocks:
backlog of ordered lane-miles, lane-miles under construction, and lane-miles in service.
Mass Transit
The Mass Transit subsector determines the total percentage of travel served by Citizens
Area Transit (CAT) buses, Rapid Transit buses and rail. The amount of travel served by
each of these modes is a function of their availability, which is the number of service
hours for buses and number of miles in service for rail. The desirability of mass transit is
limited in the model, based on studies done by the Texas Transportation Institute (TTI
2001) and communication with the RTC technical staff. Maximum possible mode share
for regular buses is assumed to be 6%, for rapid transit buses, 5%, for rail, 6.5%, and for
bicycle travel, 1.5% of total daily trips. The relationship between desirability and
availability of mass transit was modeled with a lookup graph developed from RTC
technical staff perceptions.
Alternative Modes
The Alternative Modes subsector accounts for the amount of travel served by bicycle
routes.
Traffic Management
The traffic management subsector represents infrastructure that increases the capacity of
the system by improving traffic flow. This model includes two kinds of traffic
management infrastructure: intelligent transportation systems (ITS) and traffic signal
coordination. This sector keeps track of the number of ITS signs in operation and the
number of traffic signals connected to the Las Vegas Area Computer Traffic System
(LVACTS) system. Traffic management infrastructure is assumed to increase lane-mile
capacity.
Air Quality
Carbon monoxide is used as an indicator of overall Air Quality in this model. Carbon
monoxide was chosen because it is the largest contribution of air pollution from the
transportation system. The amount of CO generated each year by vehicles is calculated as
a function of vehicle miles traveled in that year and the average CO contribution per
vehicle mile, which fluctuates according to the average system-wide traffic speed. CO
contributions per mile change every five years to reflect expected changes in vehicle fleet
composition and improvements in emissions reduction technologies. CO was used as a
proxy for all transportation-related air quality parameters. It was assumed that other
vehicle-related parameters such as volatile organic carbons and nitrous oxides would
follow similar trends and be affected similarly by policies.
Cost
The model calculates the total additional cost of the scenario each year and keeps track of
the cumulative cost over the course of the simulation run. Costs in any given year are the
total of capital costs incurred that year (for roads constructed, buses bought, or rail miles
completed, for example) plus operational costs (per bus route, or for road maintenance),
plus the cost of lost federal funding ($80 million) if CO produced by vehicle travel in that
year exceeded the CO budget in that year. The output of the cost sector represents only
the additional system costs. The Southern Nevada Regional Transportation Commission
has established and budgeted for all capital and maintenance costs associated with all
transportation expenditures through the year 2025.
Input/Output Interface
The model’s custom input/output “dashboard” is shown in Figure 8.
Model Parameters
The data for this model came from the Regional Transportation Commission’s technical
staff. The Texas Transportation Institute’s 2001 Urban Mobility Report (TTI 2001) provided
background information about transportation issues in other parts of the U.S. that helped
determine model equations. Specific data about Las Vegas came from several RTC about
transportation demands of the resident and tourist communities (RTC 2000, RTC 1996 RTC
1995). In addition, the RTC collects data about transportation behavior to verify survey data and
the predictions of modeling tools used by planning staff. Model parameters are based on data
compiled from these RTC sources and personal communication with RTC technical staff.
Model Validation
Figure 6 shows the model output for the base run, a “do nothing beyond what is currently
planned and funded” scenario. The output shows the same trends as shown in the reference
mode graphs (Figures 1, 2, and 3). The first 11 years of the reference modes represent the
historical trends and the next 14 years are the calculated projections. The output from the
validation run shows a volume/capacity ratio that increases from .4 to .8 during the 25 year
simulation run. This output matches the historical and predicted data for Southern Nevada. It
supports the contention that the model contains the essential system structure responsible for
generating the problematic behavior.
Eugemen VE) ares Ree DH
High speed lane
miles
Low speed lane
miles
CAT buses
CAT
routes
RT buses
RT bus
routes co
—
‘2 Quay [CD matey) accel Cues)
Rail miles
Bike routes
a J
—
Carpool (|_ -———————————}],,, ia
effect co .
Miles of | ——————] NY
i
_
i
Irs bl i 7 sane
Traffic
=| signals
Sear - Unie TC idea Ceemtate Kr-rgn eae E01
Figure 8. Inpu¥output screen showing results of the base run scenario. The red line on the Air Quality graph shows
the carbon monoxide budget, the daily average transportation-related amount Las Vegas is allowed to produce
according to federal regu lations.
Policy Analysis
The group first asked to see model output for each policy option separately , then ran a
series of policy scenarios combining different policy levers. The key scenarios that were
compared are shown in Table 2. The first scenario they ran was the validation scenario. This
scenario is referred to by the RTC3 as the “current plan” because it represents the transportation
planning and budgeting for the Las Vegas Valley through the year 2025. This plan includes an
increase in roads, bus service, rail service, traffic management systems, and vehicle travel
alternatives such as bike paths and car pooling. Because all of these improvements and additions
are incorporated into the model, this scenario is run without any changes to the input variables.
The additional costs of this scenario are approximately | billion dollars which result from the
loss of federal funding tied to air quality violations.
Maximum Scenario — Needs Assessment
Throughout the RTC3 process the full group was presented with information about the
transportation system as well as the planning efforts in place to address the system. One concept
presented to the group was known as the “unfunded needs assessment,” which presents all of the
system upgrades that could be made, but which are not funded. In other words, the needs
assessment is a wish list that would have to be paid for by local money that does not presently
exist. This scenario (represented as Scenario 2 in Table 2) provided for system upgrades of all
types, from additional lane miles to an increase in bus and rail service as well as traffic
10
management systems and bike lanes. The only parameter that is not altered in this scenario is the
vehicle occupancy rate or carpool effect, which is maintained at its current level.
The results of this scenario show significant improvements in congestion and air quality
but with a cost in excess of 6 billion dollars, which the group considered to be politically
prohibitive and unattainable. As a result, the RTC3 felt that the cost made the maximum scenario
unacceptable.
Minimum Scenario — “Should Do
With such a high cost associated with the wish list of the needs assessment, the RTC3
wanted to generate a scenario that would provide the system with the basic necessities for
improvement while maintaining the lowest cost. The variables and values for the “should do”
scenario were the product of repeated model runs and associated deliberations, which provided
new suggestion for model inputs and continued investigation of system structure and function. It
was decided by the group that a “should do” scenario would involve the lowest cost system
upgrades that would be considered as obvious necessities by the average voter or system user.
The parameters selected, listed as Scenario 3 in Table 2, consisted of traffic management options,
bike lane additions and an increase in vehicle occupancy, which represents an increase in car
pooling.
The minimum scenario showed a cost of one billion dollars, the bulk of which comes
from air quality violations. While the minimum scenario did not have as great an effect on
congestion as the maximum scenario, the two scenarios were actually quite similar in the trends
as well as the final outcomes. The maximum scenario ended with a system wide volume/capacity
ratio of approximately .5 and the minimum scenario ended with a .55 volume/capacity ratio. Air
quality was also similar, but the maximum scenario stayed below the violation limits for all but
two years, while the minimum scenario exceeded the limits for 10 consecutive years.
The results of the minimum scenario provided the RTC3 with a considerable amount of
information and feedback about the system as well as additional questions. After seeing the
results of the minimum scenario and comparing them to the results of the maximum scenario the
group wanted to understand why the results were so similar and yet the costs were so disparate.
The group wanted to understand and develop a mid-level scenario that could achieve an
improvement in the traffic system as well as an improvement in air quality compliance.
Mid level Scenario
Based on the comparison of the minimum and maximum scenarios, the group felt that a
mid-level scenario should be developed, which addresses both traffic con gestion and air quality,
while still costing far less than the maximum scenario (Scenario 4 in Table 2). Additional roads,
bus service and rail service were included in this scenario at reduced levels from the maximum
scenario. Traffic management options, bike lanes and carpooling were all included at the
maximum levels thought to be possible for the Las Vegas Valley.
The mid-level scenario demonstrated a volume/capacity trend that improved over the
maximum scenario during the middle seven years of the run, and then slightly exceeds it towards
the end. In the area of air quality both the mid-level and maximum scenarios show compliance
for most of the model run, with almost identical time periods of violation near the end of the run,
and then returning to compliance for the remainder of the run. The cost of the mid-level scenario
is approximately 2.5 billion dollars.
11
Table 2. RTC3 Scenarios
SCENARIOS
1 2 3 4 5
RTP Needs x Y Zz
Net Net 2000- “Shoul | Mid- | Needs
Capital O&M 2025 d-do” level +
Policy Cost per Cost per Max Carpo
Lever Unit Unit Value ol
per year
High speed
Lane miles $ 2M/mi $ | 1,000 608 850 425 850
35,000/mi
Low speed
lane miles $ 1M/mi $ | 1,000 2243 0 0 i)
35,000/mi
CAT buses $ ne 500 199 455 225 455
80,000/bus
CAT routes — | $480,000/ 100 15
route 24 15 7
RT buses $ = 100 25 24 12 24
180,000/bus
RT routes — $ 480,000/ 30 5 4 2 4
route
Rail miles $ 24 Mimi $ 250 6 80 19 80
600,000/m
i
Bike routes $105,350/mi $7,000/mi 1,000 462 400 300 400 400
Carpool effect — $35.4 M/1 1.65 1.32 1.32 1.4 1.4 1.4
person
incr
Miles of ITS $425,000/mi $3,000/mi 200 130 20 70 70 20
Traffic $ 250,000/ $ 3,000/ 350 0 350 350 350 350
signals signal signal
cost $1.5B $6B $1B $2.5 $5.5B
B
Policy Analysis Summary
The workgroup presented the five scenarios shown in Table 2 to the full group and
suggested that the group recommend the Mid-level scenario. Members of the full group were
concerned about the cost, so they made several suggestions for modifying the mid-level package.
When none of their ideas decreased the cost without making air quality or congestion worse, they
agreed to recommend the mid-level scenario. Figure 9 shows the output for all five scenarios.
12,
Ceermen 0) ‘aera ee PH
High speed lane
miles
Low speed lane
miles —_(—]____}},,
———————— —=
Lend
—————————— in
CAT buses
CAT
Toutes | co
——— Tae a
RTbuses
l i &
RT bus . ie
Touies co
‘HF Dalie ICD ewer) ‘na oeeal Co 5)
Figure 9. Results of policy scenario tests.
Discussion
This model is relatively simple. It reflects the charge of the advisory group and the
limited time frame. The purpose of building and using the model was strategic problem-solving
and education, not optimization or operational-level planning. The model served several
purposes: to increase participant understanding of how the transportation system works, to
organize a large amount of information, to build a shared basis for policy identification and
evaluation among advisory group members, and to provide a tool for comparing the relative
merits of suggested policy scenarios. Thus, although some system feedback is represented in the
model, the most valuable feedback in this case was in the process, not the model. As the group
used the model to test policies, they would respond to the output with discussions of how the
result was generated and what effect different policies had. The members of the group were
refining their knowledge of the system at the same time they were working with each other to
develop a common idea about what to recommend. Stave (2002) describes in greater depth the
role the model played in group discussions.
We had strong support from the RTC technical staff for developing this simplified model.
Several members of the staff remarked they were very pleased to have a simplified tool for
communicating the complexities of the system to others. They use models extensively for
13
planning, but their models are much more detailed and are not useful for general communication.
Even with the support of the technical staff, however, determining parameters for the simplified
model was a challenge because the RTC works at a much more disaggregated level. For
example, the RTC defines 12 different kinds of lane-miles. Summarizing these categories into
the two we used in the model required decisions about how to combine the RTC categories in a
way that was accurate, yet maintained important distinctions such as design capacity. In
addition, some of the parameters we needed are uncertain. There is little information about the
relationship between mode share and availability of mass transportation, for example. We had to
rely on the perceptions of the RTC staff to generate this relationship. If we had had more time to
develop the model, we would have confirmed the assumptions embedded in these lookup graphs
with the advisory group participants.
Two major constraints on the level of sophistication of the model were the nature of the
advisory group members and the limited time available for the entire process. The group
members were all volunteer participants with other full-time jobs. Some participants were
familiar and comfortable with computer models; others were familiar with but skeptical of
models. Some had no familiarity with models but were curious, and others were suspicious or
wary of any technology. Since it was important that the model be accepted by as many of the
participants as possible, we had to keep the model relatively simple. In addition, the advisory
group process was constrained to 12 months. Since we did not start working with the group until
the fourth month, we had only seven months, meeting only 24 hours per month with the
workgroup to develop a working model. The first several months were spent identifying the
problem and developing the causal structure, which left only about ten weeks for development of
the structural model and parameterization.
In spite of the model’s simplicity and level of aggregation, the model and model
development process served several key functions for the group. As described in Stave (2002), it
provided a structure for organizing and connecting a large amount of seemingly unrelated chunks
of information that were new to most of the group members, keeping track of where the process
was going, setting boundaries for the types of policy options that were possible, documenting the
activity of the group, and identifying and evaluating policy scenarios. The model provided a
neutral framework for discussion and generated insight through behavior that was surprising to
the group. One of the surprising findings was the importance of the vehicle occupancy rate,
which can be influenced by carpool incentive programs. When the model showed that increasing
the occupancy rate alone could have as much of an effect on the problem as the most expensive
road-building and mass transit enhancement program, participants decided that even if
carpooling might not be a popular idea, carpool incentives should be part of any policy scenario
recommended. Perhaps the most useful insight was that, as one participant put it: “it is clear the
best we are going to do is keep the system from getting worse.” An RTC staff member said
“people seem disappointed that they can’t find the silver bullet, but ’'m not. What this shows is
there is no silver bullet” (Stave 2002). He felt that this made the group take the problem
seriously and that it made them more committed to convincing others to take it seriously.
14
References
Clark County Assessor (CCA). 2000. url:http://www.co.clark.nv.us/assessor/Census.htm#Clark
County Population.
Regional Transportation Commission (RTC). 1995. Las Vegas Regional Travel Demand Model
Documentation Report Update 1995. Regional Transportation Commission of Southern
Nevada. Government Center, Las Vegas, NV .
RTC. 1996. Las Vegas Valley Household Travel Survey.
RTC. 2000. Transportation Modeling Summary. Regional Transportation Commission of
Southern Nevada, December 2000.
Stave, K.A. 2002. Using system dynamics to improve public participation in environmental
decisions. System Dynamics Review 18 (2).
Texas Transportation Institute (TTI). 2001. The 2001 Urban Mobility Study. Texas
Transportation Institute. Texas A&M University System. College Station, Texas.
Ventana Systems, Inc. 1988-2000a. Vensim® DSS32 version 4.2a. ©1988-2000a. Ventana
Systems, Inc., 60 Jacob Gates Road, Harvard, MA 01451
Ventana Systems, Inc. 1988-2000b. Vensim® Model32 Reader version 4.2a. ©1988-2000b.
Ventana Systems, Inc., 60 Jacob Gates Road, Harvard, MA 01451. Available from
www.vensim.com,
Back to
15
