Towards Sustainable Tourism

Transportation Systems and

Services in Tennessee
Research Final Report from the University of Memphis | Sabyasachee Mishra, Mihalis M. Golias,
Kakan Chandra Dey, Diwas Thapa, Md. Tanvir Ashraf | September 12, 2022

Sponsored by Tennessee Department of Transportation Long Range Planning

Research Office & Federal Highway Administration
 DISCLAIMER
This research was funded through the State Planning and Research (SPR) Program by the
Tennessee Department of Transportation and the Federal Highway Administration under RES
2021-02: Towards Sustainable Tourism Transportation Systems and Services in Tennessee.
This document is disseminated under the sponsorship of the Tennessee Department of
Transportation and the United States Department of Transportation in the interest of
information exchange. The State of Tennessee and the United States Government assume no
liability of its contents or use thereof.
The contents of this report reflect the views of the author(s) who are solely responsible for the
facts and accuracy of the material presented. The contents do not necessarily reflect the official
views of the Tennessee Department of Transportation or the United States Department of
Transportation.

i
 Technical Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.
RES2021-02
4. Title and Subtitle 5. Report Date
September 2022
RES 2021-02: Towards Sustainable Tourism Transportation Systems
and Services in Tennessee 6. Performing Organization Code

7. Author(s) 8. Performing Organization Report No.

Sabyasachee Mishra, Mihalis M. Golias, Kakan Chandra Dey, Diwas
Thapa, Md. Tanvir Ashraf
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
The University of Memphis
Department of Civil Engineering 11. Contract or Grant No.
3815 Central Avenue Grant RES2021-02
Memphis, Tennessee, 38111

West Virginia University

Department of Civil and Environmental Engineering
1306 Evansdale Drive
Morgantown, West Virginia, 26506
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
Tennessee Department of Transportation Final Report
505 Deaderick Street, Suite 900 September 2020 – August 2022
Nashville, TN 37243 14. Sponsoring Agency Code

15. Supplementary Notes

Conducted in cooperation with the U.S. Department of Transportation, Federal Highway Administration

16. Abstract

Tennessee is home to some of the most popular tourist destinations in the country and the tourism industry
contributes considerably to Tennessee’s economy. The industry in the state has grown considerably in the previous
years and there is a need to identify and address issues related to transportation services to accommodate needs
and requirements of tourists. This study identifies current deficiencies in the transportation system dedicated to
tourists, and popular tourist destinations and origin markets. The findings from the study are used to provide policy
level recommendations for improving tourism focused transportation system in the state. Results show that
continued involvement of local and private tourism agencies in transportation project planning is needed, and the
collaboration between the state DOT, state tourism office and tourism agencies should be encouraged for a tourism
focused transportation system. Analysis of origin markets for major destinations show that most travels to
Tennessee are attracted from bordering states. This can help agencies identify routes of interest for future
improvement and expansion. Finally, based on our findings, key policy recommendations are presented to improve
the current state of transportation system and services for tourism in the state.
17. Key Words 18. Distribution Statement

SUSTAINABLE TOURISM, TRAVEL No restriction. This document is available to the public

from the sponsoring agency at the website
DEMAND, COVID-19, LONG DISTANCE http://www.tn.gov/.
TRAVEL
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 89

ii
Executive Summary
Research need
Tennessee has a growing tourism industry, that contributes significantly to the state’s economy,
but also faces a growing transportation issue that needs to be addressed; congestion in major
cities and tourist routes is a growing concern. Recent data from INRIX shows five of Tennessee’s
largest cities are ranked among the most congested in the US. Major cities in Tennessee are also
home to some of the most visited and popular attractions in the country. While congestion can
be addressed through infrastructure expansion such as the addition and widening of lanes, these
alternatives are expensive. Introduction of a tourism-focused multimodal transportation system
can offer a cheaper and more effective alternative to infrastructure expansion. Introduction and
expansion of multimodal transportation services require identification of popular tourists’ routes
and attractions, transportation services available to tourists and their current deficiencies, and
initiatives to address transportation issues faced by tourists. This research study undertook a
detailed analysis of these aspects of the transportation system in Tennessee to provide the
Tennessee Department of Transportation (TDOT) with policy level recommendations that would
improve current transportation services available to tourists.

Research approach
A detailed review of the literature was completed to gather a synopsis of practices followed by
states across the county related to planning tourism transportation projects, inter-agency
collaboration, and tourism data collection and analysis. This helped to identify existing and
recently adopted practices across the country.
An online survey was administered to state Departments of Transportation and transportation
agencies across the country to gather better insight on their practices. Their approach to
selection of tourism inclusive transportation projects, inter-agency collaboration, and use of
tourism related data was explored. Using the survey data and tourism related statistics, an
analysis was undertaken to classify states by tourism impact. This was done to identify practices
that were common in high-tourism impact states versus other states. The results showed that
states with high tourism impact (in terms of economy) had better inter-agency collaboration and
tourism inclusive project selection process.
Tourists’ trip characteristics are an integral part of long-distance travel demand modeling.
Therefore, an online survey was conducted to obtain trip characteristics of tourists visiting the
state. Additionally, their mode choice, travel distance and trip timings were also obtained using
revealed preference questions. Findings from the survey suggested the most popular time of the
year to visit the state is between April to August and people mostly preferred driving when
traveling to the state. However, when the travel distance is over 500 miles, people’s preference
for air travel increased considerably. The findings from the survey were also used to analyze the
impact of different scenarios on tourism. Specifically, the travel intention of tourists during the
COVID-19 pandemic was studied using structural equation modeling and the collected survey
data. Results suggested that better travel incentives and dissemination of pandemic related
information would encourage people to travel more.

iii
Using the national long-distance travel demand model, tourism travel was forecasted for four
scenarios. These scenarios were: improved transit access, reduced air fare, reduced congestion
on popular tourist routes, and increase in household income. Results from scenario analysis
suggested that all these scenarios would have a positive impact on tourism. Notably, there would
be a considerable reduction in use of cars with improved transit access and reduced air fare. On
the contrary, people would drive more with an increase in household income.
Investigation of origin state from where people travel to major destinations in Tennessee was
done using INRIX Trip Analytics to origin markets. Memphis, Nashville, Chattanooga, Knoxville,
and Gatlinburg were chosen as destinations in our analysis since they are home to numerous
attractions that are popular among out of state tourists. States closest to the destination were
found to attract more trips to these destinations.
To identify current deficiencies in transportation services available to tourists, a survey was
designed and administered to local and regional tourism agencies in the state. These agencies
included chambers of commerce and tourism departments. The survey responses suggested
that agencies desired dedicated budget to improve transportation services in their area and also
preferred to collaborate with other agencies on project planning and selection. Additionally, the
need for improved transit services, well maintained highways, tourism signage, and state level
legislature and guidelines for inter-agency collaboration were identified as the most prevalent
deficiencies experienced by local agencies.
The findings from these tasks were used to derive recommendations to TDOT which are
summarized in the following subsection.

Key Findings
The key findings based on this study are concluded as follows:
• There is better inter-agency collaboration in states with higher tourism impact.
• Scenarios that encourage use of alternate transportation modes such as transit and air
travel are expected to considerably reduce the use of cars.
• Local agencies opine that dedicated budget would better current state of transportation
systems and services.
• A better collaboration between agencies at the state, regional and local level is needed to
improve tourism inclusive project planning and improve related transportation services.

Key Recommendations
Transportation systems play a significant role in tourism development by connecting tourism-
generating regions to destinations. The distribution, capacity, efficiency, and accessibility of
transport services can not only affect how a destination develops but also visitors' mobility, and
the connectivity of tourist experiences within destinations. However, an increased number of
tourists can create challenges in terms of the sustainability of the tourism transportation system.
Thus, proper planning and policy development are necessary to maintain the sustainability of the
transportation system and destinations. This section presents policy and guidelines for
sustainable tourism transportation services.

iv
• TDOT and tourism agencies should collaborate more and share their projects and
findings to identify future needs and current trends. Better collaboration is needed for
collection and utilization of tourism travel data and tourism-based transportation project
prioritization.
• TDOT should continue to engage the private sector in tourism transportation service
planning and development. The private sector can play a key role in tourism-related
transportation planning. The primary areas for private sector collaboration with TDOT
and tourism agencies are private sector funding, marketing, data collection, and
dissemination of tourism information.
• Tourism agencies at local and regional level should be provided more opportunities to
participate in transportation project selection and planning process. This can be achieved
by providing additional funds and grant to agencies to improve tourism transportation
services at their level.
• A more detailed analysis of current state of transportation infrastructure in popular
tourist areas and routes is warranted, particularly for attractions in Nashville, Gatlinburg,
and Pigeon Forge. Tourists accessing attractions in these destinations have limited
options; better access through public transportation is needed in these cities.
• Most out-of-state tourists vising destinations in Tennessee are from nearby states. Major
routes that serve tourists traveling from bordering states should be given preference in
terms of timely repair and maintenance.

v
 Table of Contents
DISCLAIMER .................................................................................................................................................. i
Technical Report Documentation Page................................................................................................... ii
Executive Summary................................................................................................................................... iii
Research need ....................................................................................................................................... iii
Research approach ............................................................................................................................... iii
Key Findings ........................................................................................................................................... iv
Key Recommendations ......................................................................................................................... iv
List of Tables .............................................................................................................................................. ix
Chapter 1 Introduction......................................................................................................................... 1
1.1 Problem statement .................................................................................................................... 1
1.2 Project objectives ....................................................................................................................... 1
1.3 Methodological approach ......................................................................................................... 2
1.4 Report organization ......................................................................................................................... 2
Chapter 2 Literature Review................................................................................................................ 3
2.1 Tourism transportation in the U.S. ................................................................................................ 3
2.2 Long Distance Travel Demand Models for Tourism trips........................................................... 3
2.3 Current Tourism Development Strategies and Guidelines ........................................................ 5
2.4 Recreational travel intentions ........................................................................................................ 5
2.4.1 Theory of Planned Behavior .................................................................................................... 6
2.4.2 Necessary Condition Analysis ................................................................................................. 6
Chapter 3 Methodology ....................................................................................................................... 7
3.1 Survey of state DOTs and STOs...................................................................................................... 7
3.1.1 Analysis of survey responses using clustering technique................................................... 7
3.2 Analysis of tourist trip characteristics ........................................................................................... 9
3.2.1 Summary of trip characteristics.............................................................................................. 9
3.2.2 Antecedents to recreational travel intentions ...................................................................... 9
3.3 Analysis using national long-distance travel model .................................................................. 11
3.3.1 Scenario development ........................................................................................................... 11
3.3.2 Model implementation .......................................................................................................... 12
3.3.3 Scenario analysis..................................................................................................................... 13
3.4 Analysis of popular attractions .................................................................................................... 15
3.4.1 Access to destinations ............................................................................................................ 16

vi
 3.4.2 Analysis of trips for identification of origin markets ......................................................... 17
3.5 Transportation services for tourism in Tennessee ................................................................... 18
Chapter 4 Results and Discussion .................................................................................................... 19
4.1 Survey of state DOTs and STOs.................................................................................................... 19
4.1.1 Discussion of the Survey result............................................................................................. 19
4.1.1.1 Interagency collaboration in tourism transportation planning processes ......................... 19
4.1.1.2 Private sector involvement .................................................................................................... 20
4.2 Analysis of tourist trip characteristics ......................................................................................... 21
4.2.1 Summary of trip characteristics............................................................................................ 21
4.2.1.2 Household characteristics ..................................................................................................... 22
4.2.1.3 Travel mode and duration of stay ........................................................................................ 23
4.2.2 Antecedents to recreational travel intentions .................................................................... 25
4.2.3 Conclusion and implications ................................................................................................. 27
4.3 Analysis using national long-distance travel model .................................................................. 28
4.3.1 Scenario 1: Impacts of transit access improvement .......................................................... 28
4.3.2 Scenario 2: Impacts of air fare policy on tourism travel ................................................... 29
4.3.3 Scenario 3: Impacts of lower congestion along major tourist routes ............................. 30
4.3.4 Scenario 4: Impacts of increase on household Income .................................................... 30
4.3.5 Conclusion and implications ................................................................................................. 31
4.4 Analysis of popular destinations.................................................................................................. 32
4.4.1 Access to popular attractions................................................................................................ 32
4.4.2 Analysis of trips for identification of origin markets ......................................................... 35
4.4.3 Identification of popular routes ............................................................................................ 36
4.4.4 Conclusion and implications ................................................................................................. 37
4.5 Current state of tourism-related transportation system in Tennessee ................................. 38
4.5.1 Performance of available transportation systems and services ...................................... 38
4.5.2 Need for dedicated budget ................................................................................................... 38
4.5.3 Tourism transportation planning and its role in Tennessee ............................................ 39
4.5.4 Issues in transportation services .......................................................................................... 40
4.5.5 Ranking of initiatives to improve transportation system and services ........................... 41
4.5.6 Collaboration between agencies .......................................................................................... 42
4.5.7 Conclusion and implications ................................................................................................. 43

vii
 4.6 Policy recommendations .............................................................................................................. 43
Chapter 5 Conclusion ......................................................................................................................... 45
Recommendations ............................................................................................................................... 46
References ................................................................................................................................................. 48
Appendices ................................................................................................................................................ 53

viii
 List of Tables
Table 3-1 Descriptive statistics of land use variables (N=4,566) ........................................................ 12
Table 3-2 Parameter values used in scenario analysis ....................................................................... 13
Table 3-3 Summary of the proposed scenarios ................................................................................... 14
Table 3-4 Urbanized Areas in Tennessee based on the 2020 census............................................... 16
Table 4-1 Summary of results for hypotheses used in investigating travel intentions .................. 26
Table 4-2 Policy recommendations to address current concerns..................................................... 44
Table A-1 Summary statistics of collected variables ........................................................................... 53
Table A-2 Summary of response obtained from the survey .............................................................. 54
Table B-1 Descriptive statistics of participants’ response to attitudinal questions ........................ 59
Table B-2 Results from factor analysis .................................................................................................. 62
Table B-3 Summary of results for the hypotheses .............................................................................. 65
Table C-1 List of popular destinations................................................................................................... 68
Table E-1 List of survey recipients who were invited to undertake the survey of agencies .......... 75

ix
 List of Figures
Figure 3-1 Framework used to analyze survey responses............................................................................ 8
Figure 3-2 Theoretical model used in SEM analysis.....................................................................................10
Figure 3-3 Long distance travel simulation framework ..............................................................................11
Figure 3-4 Long distance trips per year in Tennessee for base scenario .................................................13
Figure 3-5 Popular attractions in Tennessee ................................................................................................15
Figure 3-6 Identification of O-D pairs for analysis route analysis ..............................................................18
Figure 4-1 Collaboration manner among state DOTs and STOs. ...............................................................20
Figure 4-2 Role of private sector stakeholders in tourism transportation service development .........21
Figure 4-3 Socio-demographic characteristics of tourists visiting Tennessee .........................................22
Figure 4-4 Household characteristics of tourists visiting Tennessee ........................................................23
Figure 4-5 Travel mode by distance traveled ...............................................................................................24
Figure 4-6 Duration of stay at destination ....................................................................................................24
Figure 4-7 Satisfaction ratings from tourists traveling to Tennessee .......................................................24
Figure 4-8 Distribution of travel across a year .............................................................................................25
Figure 4-9 Recreational travel frequency of tourists ...................................................................................25
Figure 4-10 NCA effect sizes (All effect sizes significant at p=0.01) ...........................................................26
Figure 4-11 Results from Scenario 1: Change in mode share in Tennessee ............................................28
Figure 4-12 Top 20 O-D pair under increased transit access scenario .....................................................29
Figure 4-13 Result from Scenario 2: Change in mode share due to reduced air fare ............................30
Figure 4-14 Result from Scenario 3: Change in long-distance trips under reduced congestion ...........30
Figure 4-15 Result from Scenario 4: Changes in mode share with increased household income ........31
Figure 4-16 Result from Scenario 4: Changes in trip purpose with increased household income .......31
Figure 4-17 Access to destinations from its nearest UA .............................................................................32
Figure 4-18 Access from nearest primary service commercial airport .....................................................33
Figure 4-19 Access from nearest interstate exit ..........................................................................................34
Figure 4-20 Access from nearest transit stop to the destination ..............................................................35
Figure 4-21 Priority markets for tourism across Tennessee ......................................................................36
Figure 4-22 Current state of transportation systems in the area ..............................................................38
Figure 4-23 Need for dedicated budget ........................................................................................................39
Figure 4-24 Opinions on current state of tourism related transportation services ................................40
Figure 4-25 Current issues in transportation services ................................................................................41
Figure 4-26 Solutions to address current issues in transportation services ............................................41
Figure 4-27 Ranking of initiatives to improve transportation system and services ................................42
Figure 4-28 Obstacles in collaboration between agencies .........................................................................42
Figure 4-29 Solutions to foster collaboration between agencies ..............................................................43
Figure A-1 Map of the survey respondents by state and agency types ....................................................54
Figure A-2 Multivariate clustering boxplot for identified state clusters....................................................55
Figure A-3 Map showing clusters of states based on tourism travel characteristics ..............................56
Figure A-4 Cluster-wise distribution of tourism data collection and forecasting practices ...................57
Figure A-5 Summary of responses on current deficiencies and its solutions .........................................58
Figure B-1 Predictor bottlenecks for the outcome to manifest .................................................................66
Figure D-1 Origin and destinations of trips made through I-55 E .............................................................70
Figure D-2 Origin and destinations of trips made through I-40 E .............................................................70
Figure D-3 Origin and destinations of trips made through I-40 W ............................................................71
Figure D-4 Origin and destinations of trips made through I-40 E .............................................................71
Figure D-5 Origin and destinations of trips made through I-24 E .............................................................72

x
Figure D-6 Origin and destinations of trips made through I-75 S .............................................................72
Figure D-7 Origin and destinations of trips made through I-40 W ............................................................73
Figure D-8 Origin and destinations of trips made through I-40 E .............................................................73
Figure D-9 Origin and destinations of trips made through I-75 S .............................................................74
Figure D-10 Origin and destinations of trips made through US-321 S to Gatlinburg .............................74

xi
Chapter 1 Introduction
1.1 Problem statement
Tennessee is home to several iconic tourist destinations and attracts millions of tourists every
year. The tourism industry in Tennessee outperforms many states in key indices such as travel
expenditure, payroll, jobs, and tax revenue which highlights the state’s potential for the tourism
industry. According to Tennessee Department of Tourist Development’s (TDTD) 2019 annual
report, tourism industry in Tennessee supports about 190,000 jobs in the state, generates a
payroll of approximately $5 billion with over $22 billion in travel expenditures, and a tax revenue
of about $2 billion. Analyses have shown that Tennessee (domestic travel) is the largest origin
market for tourists followed by neighboring states Alabama, North Carolina, Georgia, and
Kentucky (TDTD, 2018). The transportation system and services facilitating travel from outside as
well as within state is necessary to promote tourism in Tennessee.
Research has shown that availability of transportation services in recreational areas determines
trip characteristics of tourists and the recreational activities they engage in (Anderson et al.,
2011). For example, the availability of bicycle facilities and infrastructure encourages tourists to
use bicycles for transportation and recreational purposes. According to the recent data, the five
major cities in Tennessee are among the 250 most congested in the US with Nashville ranked at
54, Cleveland at 138, Memphis at 144, Chattanooga at 166, and Knoxville at 198 (INRIX 2021 Global
Traffic Scorecard, 2021). Therefore, an effective transportation system that offers alternative
services and contributes towards reduction of congestion in major cities and destinations is
necessary. Developing policies that encourage the use of multimodal transportation systems and
enhance tourists’ experience by reducing traffic congestion on major tourist routes and
destinations is essential for tourism.
States with large tourism industries have invested considerably in transportation systems and
services. Similarly, Tennessee needs to develop and maintain a robust and sustainable
transportation system that can cater to the needs of the fast-growing tourism industry. Many
recreational destinations in the state see a seasonal influx of tourists that exceeds the current
capacity of transportation services. Therefore, development of multimodal inter-city and intra-
city transportation services accompanied by effective operational policies is necessary to
promote the quality of tourism travel and establish the state as a primary attraction among
tourists.
This study investigates measures taken by states across the country to establish an effective
transportation system that supports tourism. An assessment of the current state of
transportation systems in the state using surveys and analysis of long-distance travel is
undertaken to present Tennessee Department of Transportation (TDOT) with planning policies
and recommendations to establish a sustainable transportation system that can serve the
growing needs of tourists traveling to the state.

1.2 Project objectives

The overarching goal of the research project is to develop planning and policy guidelines for
sustainable transportation systems and services in promoting the tourism industry in Tennessee.
The specific objectives of the project are:

1
 1. Undertake a comprehensive review of the literature on transportation planning adopted by
other states in the US and elsewhere.
2. Identify the current state of transportation systems and services available to tourists, their
deficiencies, and potential measures for their improvement.
3. Identify key destinations and corridors across the state that are primarily used by tourists
or long-distance travelers.
4. Develop policies and guidelines for a sustainable transportation system capable of
promoting the tourism industry.

1.3 Methodological approach

The methodological approach undertaken by the research team can be summarized under six
steps as follows. A detailed description of the methodology is presented in Chapter 3.
1. Review of the literature to gather an understanding of state-of-the-art practices in terms of
planning for tourism related sustainable transportation systems and services.
2. Survey of state Departments of Transportation (DOTs) and State Tourism Offices (STOs) to
understand state specific tourism related travel demand modeling practices, tourism data
sources, data analysis methods, tourism inclusive project selection practices, consideration
of sustainable transportation for tourism, and the collaboration between diverse tourism
stakeholders.
3. Analysis of tourism trip characteristics, which is fundamental to long-distance passenger
travel demand models, using online survey of travelers.
4. Analysis of current and forecasted tourist travel patterns using the national long-distance
travel demand model and scenario analysis.
5. Survey of local tourism agencies on the present state of tourism transportation system and
services in their jurisdiction, current deficiencies, and preferred improvements.
6. Make policy level recommendations based on findings.

1.4 Report organization

The rest of the report is organized as follows. Chapter 2 includes a brief review of the literature
encompassing various aspects of transportation systems and services associated with tourism.
This includes economic impacts of tourism, methodological approaches used for data collection
and analysis of long-distance trips. Chapter 3 presents a detailed description of the stepwise
methodology applied in the current research project. The findings from applied methods are
included in Chapter 4. Chapter 5 presents a detailed conclusion of the research project along
with policy recommendations.

2
Chapter 2 Literature Review
An extensive review and synthesis of published research and pilot projects related to sustainable
tourism transportation was undertaken to get an understanding of state-of-the-art practices
across agencies. Our review included all aspects of planning and development of sustainable
transportation systems and services for tourism. The literature is summarized as follows.

2.1 Tourism transportation in the U.S.

Tourism travel is defined as the temporary movement of people to various destinations outside
their normal travel patterns (e.g., commuting to work or going shopping for groceries) (Pincus et
al., 1999). Tourism specifically focuses on travel comprising of a night away from home or a day
trip which is 50 or more miles one way, with the basic unit of measurement is a “person-day”
(Pincus et al., 1999). This definition allows transportation planners to separate tourism trips from
commuter trips and assess their economic impacts on communities. Tourism has positive
impacts on a local community but also has negative impacts. The positive or negative impacts of
tourism can be categorized into four main types: economic, social, cultural, and environmental
(Holden, 2016). Unplanned growth of tourism industries and related infrastructures and services
can lead to a variety of detrimental impacts on communities, economic bases, and natural
resources. Sustainable tourism includes planned transportation systems and services that can
be used to control and manage the travel demand for tourist destinations (i.e., right number of
tourists, at the right destination of choice, at the right times) which could balance the
sustainability of destination ecology, economic interest of local communities and transportation
systems capacity. Sustainable transportation systems ensure a viable relationship between local
population and tourists by minimizing the negative impacts of tourism-related transportation.
According to the National Advisory Committee on Travel and Tourism Infrastructure (NACTTI),
the tourism industry accounts for 2.7% of GDP and seventh-largest employment sector in the
U.S. (NACTTI, 2016). In 2017, the U.S travel and tourism industry generated over $1.6 trillion in
economic output while supporting 7.8 million American jobs and one in eighteen U.S. jobs,
directly and indirectly, relied on the travel and tourism industry (SelectUSA, 2020). Tennessee is
among the fastest-growing travel destinations in the U.S. for international tourists. In 2019,
Tennessee tourism hit a record-high $23.27 billion in annual economic impact (The Soundtrack of
America: Made in Tennessee (FY 2019 Annual Report), 2018). Tourism supported 195,000 jobs and
produced $1.92 billion in state and local sales tax revenue in 2019. Travel to Tennessee topped
126.18 million person stays in 2019, up 5.7 percent from the previous year. In 2019, 92 counties
in the state saw an increase in domestic travel spending.

2.2 Long Distance Travel Demand Models for Tourism trips

A significant portion of long-distance trips are recreational trips. Previous studies defined ‘‘long-
distance’’ trips based on the travel distance (e.g., longer than 50 miles) and travel times (e.g.,
greater than 40 minutes one way travel time) (Bierce & Kurth, 2014). Several studies defined long-
distance trips that were greater than 50 miles (Bierce & Kurth, 2014; Erhardt et al., 2007; Rohr et
al., 2013) and few other studies considered trips longer than 100 miles as long-distance trips (Frei
et al., 2010; Frick & Grimm, 2014; Kuhnimhof et al., 2014). Though long-distance tourism trips
represent only 1% of the total trips, they share 15% of the total vehicle miles traveled (VMT) in

3
the U.S. (Schiffer, 2012). Long-distance tourism trips have not been an integral part of the state
or regional travel demand models in the US for many years because of the lack of modeling data.
In the last few years, several initiatives and new data sources have enabled the inclusion of long-
distance trips in travel demand modeling (e.g., (Davis et al., 2018; Llorca et al., 2018; Yang et al.,
2019)).
Erhardt et al. (2007) developed a long-distance travel model as a part of the Ohio statewide travel
demand model (Erhardt et al., 2007). In this study, the authors used the Long-Distance Travel
(LDT) survey to collect information on all trips greater than 40 mi, which were not regular work
commute trips. The LDT model was developed using 3,660 traffic analysis zones (TAZs) in Ohio
and a ring of 588 TAZs outside Ohio. Households with more automobiles and higher incomes
were more likely to make long-distance trips due to a higher level of mobility and a higher income
level to absorb cost. The study also found that hotel employment was a strong indicator of long-
distance travel as travelers stay near tourist attractions and business districts. Rohr et al. (2013)
developed an LDT model for Great Britain. This study reported a positive relationship between
car ownership and long-distance travel. Similar to the Ohio LDT model, this study also found that
income had a strong positive effect on long-distance travel. Men were more likely to make long-
distance recreational trips compared to women, whereas families with children were less likely
to make long-distance trips.
Another study integrated a national long-distance trip model to the statewide travel demand
model in Tennessee (Bernardin Jr et al., 2017). The Federal Highway Administration’s new
national long-distance passenger travel model (rJourney) was used in the Tennessee statewide
travel demand model and calibrated the model using cell phone data. In the study, the
researchers used long-distance origin-destination (O-D) data derived from cell phone data to
model the statewide long-distance trips. Another study in California used long-distance tour data
collected through statewide household travel surveys to explore non-commute long-distance
trips' behavioral factors (Davis et al., 2018). This study also used social media data (i.e.,
Foursquare) to describe destination characteristics and their significance in explaining long-
distance tour behavior. Foursquare data was used to identify the participants’ long-distance by
activity type. Long-distance trips were often chaining of trips with different purposes. Path
analysis of long-distance tours found that high-income households were likely to travel by air and
often combined work trips and leisure trips. On the other hand, low-income households were
most likely to rely on vehicular travel. The miles driven by car was positively associated with the
number of employed household members. Larger households were less likely to make long-
distance tours with many miles driven and households not living in single-family homes were less
likely to drive far. Households with lower income, with lower number of cars, and living in city
centers were more likely to make tours using public transportation services.
Similarly, a long-distance travel demand model was developed for the province of Ontario,
Canada using trip survey data, location-based big data, and trip planning services (Llorca et al.,
2018). According to the Travel Survey for Residents in Canada, long-distance trips were defined
as non-recurrent overnight trips and day trips longer than 40 km. Foursquare and Rome2rio data
were combined with trip survey data in developing a microscopic discrete choice long-distance
travel demand model. Combining these two data provided a comprehensive view of the long-
distance travel characteristics. Foursquare data was statistically significant on the number of
check-ins at destinations, especially for leisure trips, and improved the goodness-of-fit compared

4
with models that only used population and employment. Day trips were more likely to have a
closer destination, while long tours were mostly overnight trips. The chances of choosing air
travel compared to personal vehicle increased with the increase in trip distance and an overnight
trip. Similar to other studies, this study also found that higher income groups had a lower
probability of selecting bus and rail.

2.3 Current Tourism Development Strategies and Guidelines

The National Cooperative Highway Research Program (NCHRP) Report 419: Tourism Travel and
Transportation System Development surveyed state travel/tourism offices and DOTs to determine
their policies and institutional practices supporting tourism growth (Frechtling et al., 1998).
Survey findings indicated that interagency coordination was key to effective planning and
implementing transportation projects that support tourism. Usually, state DOT and state travel
offices collect different types of tourist data, which can be shared to facilitate more integrated
decisions on tourism-related transportation projects in an informed way. This study identified
eleven principles that emphasized the need for collaboration between different public and
private tourism stakeholders, involvement of the private sector, and performance measures for
tourism transportation services. In addition, this study proposed guidelines to be considered in
the development of integrated planning and project development (Frechtling et al., 1998).
Recently, several state DOTs have developed plans to integrate tourism into transportation
planning. In 2015, the California Department of Transportation developed the Interregional
Transportation Strategic Plan that includes two criteria related to tourism in individual
transportation project funding decision-making. The first criterion focused on the project’s
impact on improving corridor access to/from major generators of economic activity (e.g.,
passenger and/or freight gateways, business centers) and tourism destinations. The second
criterion focused on the possible reduction in travel time to and from freight gateways, centers
of significant economic activity, jobs, or tourism destinations. Similarly, Travel Michigan
developed a strategic tourism plan for the year 2012 to 2017 by analyzing the deficiencies in the
tourism industry and potential ways to overcome the shortcomings. Informal and formal
collaboration and cooperation between public and private entities were vital for the success of
both individual tourism businesses and the entire tourism industry in Michigan (Nicholls, 2012).
To improve the quality, connectivity, and diversity of tourist transportation service options
throughout Michigan, the strategic tourism plan outlined collaboration plans with Michigan DOT.
The study documented strategies for enhancing the visitor’s in-state travel experience by working
with Michigan DOT to standardize and improve the Welcome Center experience and keep
selected rest areas open year-round. The study also outlined plans to work with other
appropriate authorities to establish tourism information kiosks in high visitor traffic areas (e.g.,
welcome centers, airports, train stations, convention centers).

2.4 Recreational travel intentions

Compared to the previous year, there was a 59% and 70% decrease in domestic and international
air travel respectively in 2020 because of the COVID-19 pandemic (Heeb, 2021). As of Summer
2022, various strains of the virus are still circulating, and the travel industry should remain
prepared for future outbreaks. Identifying quick economic recovery measures in advance can be
helpful in this regard. Understanding the public’s opinions, psychological constructs, and travel

5
intentions can be beneficial to this. Therefore, research was undertaken, as a part of the project,
to investigate the travel intentions of tourists.
Travel intentions can be assessed using the Theory of Planned Behavior (TPB) that is based on
Structural Equation Modeling (SEM) (Ajzen, 1991). Applying SEM requires survey data with
attitudinal questions administered to tourists (in case of this study, those who traveled to
Tennessee in the past). Furthermore, TPB can be improved with other analysis techniques such
as Necessary Condition Analysis (NCA) (Dul, 2016). Unlike commonly employed regression
techniques that test whether a predictor is sufficient to affect the outcome variable, NCA can test
whether the predictors are necessary to manifest the outcome thus addressing questions like,
“Is a psychological construct, which is an antecedent to recreational travel intention during the
pandemic, both sufficient and necessary to affect it? If yes, what is the minimum condition needed for
the psychological construct to manifest travel intentions?” The upcoming subsections present a brief
review of the literature on TPB and NCA.

2.4.1 Theory of Planned Behavior

Prediction of behavior as an outcome of intentions is based on the TPB (Ajzen, 1991). TPB asserts
that behavioral intentions, defined as factors that motivate certain behavior without necessarily
the behavior being performed, are affected by attitude, subjective norm, and perceived
behavioral control. Attitude refers to an individual's evaluation of the behavior in question.
Subjective norm is an individual’s approval or disapproval of people’s views and opinions
regarding the behavior. Perceived behavioral control is people's perceptions of their ability to
perform the behavior. Favorable attitude, subjective norm, and greater perceived behavioral
control usually strengthen behavioral intentions (Ajzen, 1991). More recent studies have
provided substantial evidence of this (Kim & Stepchenkova, 2020; Pooley & O’Connor, 2000).
Extension of traditional TPB is necessary when studying human decisions and behavior in
different settings. The Extended TPB (ETPB) incorporates various latent variables depending on
the research context to gain meaningful behavioral insights. Some of the variables that have been
included in ETPB and relevant to this study are socio-demographic attributes, past travel behavior,
public trust, anxiety, self-composure, travel concern, perceived knowledge, and perceived benefits.

2.4.2 Necessary Condition Analysis

The literature on NCA is limited since it is relatively new. The framework was first introduced to
identify the necessity of predictors in achieving favorable outcomes in organizational settings
(Dul, 2016). NCA is designed to complement the traditional sufficiency logic (i.e., a predictor is
sufficient to affect the outcome) imposed by regression techniques by applying the necessity
logic (i.e., the mere presence of a predictor might not be sufficient to affect the outcome, but a
certain level of the predictor might be necessary). Its application on SEM was demonstrated using
the traditional technology acceptance model where technology use is predicted by the various
latent psychological constructs (Richter et al., 2020). More recently, it was used to investigate
holiday intentions during the COVID-19 pandemic (Pappas, 2021). In their study, researchers
reported perceived travel, destination, and hospitality risks were necessary for predicting travel
intentions.

6
Chapter 3 Methodology
Following review of the literature, the study followed a stepwise methodology to accomplish the
project objectives. These steps are detailed in the upcoming sections and briefly introduced as
follows.
1. Survey of state DOTs and STOs on tourism related planning and managing of transportation
projects and systems, and its analysis using clustering.
2. Analysis of tourists’ trip characteristics and travel intentions using responses collected from
an online survey.
3. Use of national long-distance travel model and scenario analysis to ascertain potential
impacts of various initiatives on tourism.
4. Identification of popular tourist attractions and an assessment of accessibility to them.
5. Investigation of available transportation services, their deficiencies and measures for
improvement based on survey administered on local tourism agencies.
6. Policy implications based on findings from applied methods and analyses.

3.1 Survey of state DOTs and STOs

A survey of state DOTs and STOs across the country was undertaken to understand state specific
travel demand modeling practices, sources of tourism data and methods used in its analysis,
project selection practices, and collaboration between agencies. The survey was conducted
online using Qualtrics between June-July 2021. The single survey questionnaire was developed
for both state DOTs and STOs. Based on the selection of the respondent (whether they work for
a state transportation department or a tourism agency), they were directed to questions relevant
to them. On average, the survey took about 10 minutes to complete.
In total, 19 DOTs and 20 STOs responded to the survey. Only 33 responses were complete and
used in our analysis. The responses were analyzed to get a distinct idea of practices followed
across the country using the k-means clustering method. Summary of responses to key questions
in the survey is presented and briefly discussed in Chapter 4: Results and Discussion of the
report. The analysis framework used for clustering is as follows. Additional details on the
framework, and its data is presented in Appendix A.

3.1.1 Analysis of survey responses using clustering technique

The overall analysis framework used to analyze the survey responses is presented and outlined
in Figure 3-1. In steps 1 and 2, a search engine database was created for tourism travel-related
keywords. Based on the findings of past studies, five major aspects of tourism travel (i.e., lodging,
dining, tour, traffic, and recreation) were considered in this study to extract the search engine
data. These five categories represent travel demand and capture the supply side of the tourism
industry. Twenty-five keywords were selected for the search engine data collection. Google
search engine data was considered in this study. Once the keywords associated with tourism
travel were selected, they were manually checked using Google Trends website to ensure their
availability. Google presents the keyword data relative to the highest point on the chart for the
given spatial and temporal extent (GoogleTrends, 2021). They provide the value of a keyword
using a scale of 0 to 100. Next, several variables related to tourism travel and its impact on the
local economy, such as travel spending by domestic and international visitors, jobs created by

7
tourism travel, and tax collected from the travel industry, were selected to determine tourism-
related travel characteristics of the states.

Figure 3-1 Framework used to analyze survey responses

Tourism travel-related data were collected and processed for the clustering analysis in step 3.
Several variables related to tourism travel and its impact on the local economy, such as travel
spending by domestic and international visitors, jobs created by tourism travel, and tax collected
from the travel industry, were selected to determine tourism-related travel characteristics of the
states. Search engine data related to travel and tourism mentioned in steps 1 and 2 were
collected and used as a proxy variable to include state-wise tourism volume and travel pattern in
the clustering process. Then, fifty U.S. states were clustered based on their tourism and travel
characteristics. For the classification technique, the k-means clustering method was used which
is one of the most popular unsupervised classification algorithms used in classification problems
(Mather & Tso, 2016). The clustering method divided the states into clusters based on uniform
travel and tourism characteristics.
Step 5 of the framework was synthetic data preparation to generate survey data for states that
did not respond to the survey based on recorded responses. Iterative Proportional Fitting (IPF),
which was introduced by Deming and Stephan (1940), is an efficient and popular method for
synthetic population data generation from sample data (Tibshirani et al., 2001). Once the
clustering process was completed, the IPF algorithm was used to generate synthetic survey data
for all states within each cluster using the recorded survey responses. For this study, only
questions related to the collaboration manner, performance measures, the role of the private
sector, data collection, and forecasting method were synthesized. As the characteristics of the
states within each cluster were homogeneous, the synthetic data for each category should
represent tourism characteristics reasonably well.

8
3.2 Analysis of tourist trip characteristics
A second online survey was administered to people ages 18 and above who had travelled to a
recreational destination in Tennessee. The goal of this survey was to gather information on trip
characteristics of tourists which is a critical component of long-distance passenger travel demand
models (Outwater et al., 2015). This survey was administered on an online panel from Centiment
(Centiment, 2021) using Qualtrics between May-June 2021. The survey was comprised of
questions on socio-demographics, household characteristics, trip characteristics, and
recreational travel attitude in relation to the COVID-19 pandemic and took about 13 minutes to
complete. A quota sampling approach was followed to ensure the validity of responses. Age and
gender of the respondents was used as quota variables. The panel generated over 2,000
participants of which complete responses were obtained from 1,259 individuals.
The responses collected from the survey were analyzed in two steps which included i) descriptive
statistics of trip characteristics, and ii) investigation of variables predicting travel intention during
the COVID-19 pandemic.

3.2.1 Summary of trip characteristics

In the first step, descriptive statistics for major variables associated with socio-demographics,
household characteristics, trip characteristics, mode of travel, etc. were obtained to get a better
understanding of tourists and trip characteristics. While responses were also collected on travel
attitudes, particularly, considering the COVID-19 pandemic, they were utilized to investigate
travel intentions.

3.2.2 Antecedents to recreational travel intentions

As mentioned, in the second step, a research study was also undertaken to analyze responses
collected on recreational travel attitude in relation to the COVID-19 pandemic. The goal of the
study was to identify policy interventions that could be helpful in encouraging tourists to travel
during and after the pandemic. Our analysis used SEM based ETPB, and NCA to investigate
predictors of travel intentions and the necessary conditions, respectively.
3.2.2.1 Structural Equation Modeling
Psychological constructs represent an individual’s beliefs that affect their behavior. Unlike
observable variables they cannot be directly measured. They are derived from indicators in a
survey. The indicators ask respondents to rate certain attributes on a scale. These attributes are
designed to represent the beliefs and values of the respondents. Indicators are then used to
extract underlying latent psychological constructs based on correlation within one another using
factor analysis. SEM utilizes these latent constructs to investigate causal relationships using
hypotheses derived based on theoretical reasoning, logic, and prior research. SEM comprises two
components, i) structural equations and ii) measurement equations. Structural equations define
the relationship between the latent constructs and exogenous explanatory variables. The
measurement equations establish the relationship between indicators and the latent constructs.
The SEM model proposed for this study is shown in Figure 3-2 which is a representation of the
hypotheses used to guide our analyses. Moderating effects of public trust and subjective norms
were also introduced in our analysis. These hypotheses and results are presented in detail in
Appendix B. It is worth mentioning that the moderation effects in our SEM were introduced using
the indicator product approach (Steinmetz et al., 2011).

9
3.2.2.2 Necessary Condition Analysis
NCA is based on the premise of necessity logic that states certain observed outcomes manifest
only when a certain value of the predictor is present. NCA uses ceiling lines that are drawn either
using non-decreasing piecewise linear functions called Ceiling Envelopment or continuous
function Ceiling Regression (Dul, 2016) to find these values. These lines separate the smallest
rectangle that can enclose all the observations plotted in the Cartesian coordinate system
(outcome on the y-axis, predictor on the x-axis) into two zones, scope (lower zone) that
encapsulates all the observations and ceiling (upper zone) above the observations. Using effect
size, d the necessity of a variable in manifesting the outcome can then be evaluated. The effect
size is always positive and less than 1 and its Interpretation depends on the context (Dul, 2016).
For the study, we considered an effect size, d ≥ 0.1 as an indication of the necessary condition
since d < 0.1 is considered a small effect (Dul et al., 2020). NCA also facilitates calculation of
bottlenecks which determines the minimum “level” of the predictor variable necessary to
manifest a certain “level” of outcome in the response variable.
There are several ceiling functions. In this study, we employ the CR-Free Disposal Hull (CR-FDH)
ceiling function due to the continuous nature of the composite latent variable scores obtained
from SEM (Dul et al., 2020). CR-FDH function is drawn by applying ordinary least squares
regression on the leftmost edges of the CE-FDH function.
For our investigation, first, SEM analysis was undertaken and the statistically significant
exogenous predictors (socio-demographics and travel behavior) of latent constructs were then
identified. Then the relationship between the latent factors was investigated using the developed
hypotheses. NCA was then used to investigate the necessity of the latent predictors in
manifesting their respective outcomes. All predictor latent factors regardless of their significance
in SEM were included in NCA. This was done because certain variables insignificant in SEM could
still be significant in NCA (Richter et al., 2020).

Figure 3-2 Theoretical model used in SEM analysis

10
3.3 Analysis using national long-distance travel model
The objective of this task was to apply the national long-distance model to identify and forecast
tourism travel characteristics in Tennessee using scenario analysis. The national long-distance
model is a tour-based simulation model for long distance travel behavior and patterns. It is
implemented through a software platform known as rJourney (Outwater et al., 2015). The
passenger travel demand model operates as follows. First, the national long-distance model
estimates tour generation, scheduling, duration, and party-size models by purpose. Next, it
incorporates mode and destination choice models for different purposes, which include leisure
and vacation, visits to friends or relatives, personal business, commuting, and employer’s
business. Four modes can be modeled in accomplishing the estimated trips by purpose, which
are personal cars, intercity bus, intercity rail, and commercial air travel. The rJourney tool allows
the evaluation of different policy scenarios including transportation system improvement, fare
or service changes for various modes including highway, commercial air travel, intercity bus, and
intercity rail.

3.3.1 Scenario development

Figure 3-3 Long distance travel simulation framework

The steps followed for scenario analysis in this study is presented in Figure 3-3. The national long-
distance travel demand model adopts National Use Model Area (NUMA) based zone system.
NUMA-level zone system is a composite representation of counties and Census Bureau Public
Use Microdata Areas (PUMA) across the U.S. (Outwater et al., 2015). Adopting different zone
systems increase computational requirements and complexity in the data preparation process.
As the rJourney tool can model a maximum of 4,700 TAZs, the default NUMA zone system was
retained for the long-distance travel modeling in Tennessee. In addition, default synthetic
population data which includes around 115 million households, was used for the long-distance
model.

11
For the long-distance modeling, the rJourney simulation tool utilizes national land use data
generated using 2010 census data. To update the model for 2019, Census Tract (CT) level land
use data was collected from Census Bureau, US National Park Service, and National Center for
Education website. Then, the CT level data was aggregated to the NUMA level using ArcGIS Pro
software. Table 3-1 presents the summary statistics of the land use variables used in the long-
distance travel demand model.
Table 3-1 Descriptive statistics of land use variables (N=4,566)

Variable Min Max Mean Std. Dev

The land area in public parks (square miles) 0 62,940.05 150.54 1310.96
The number of households living 0 883,434.00 26,908.54 28,365.71
The number of university students enrolled 0 17,0061.00 4,469.98 11,054.64
The total number of jobs in the zone 1 938,014.00 28,193.98 41,616.61
The number of agricultural jobs in the zone 0 7,710.00 248.56 413.20
The number of mining jobs in the zone 0 11,210 130.02 404.24
The number of utility jobs in the zone 0 5,913 202.46 232.35
The number of construction jobs in the zone 0 76,465 1,575.56 2,018.38
The number of manufacturing jobs in the zone 0 31,245 2,187.65 2,469.25
The number of wholesale trade jobs in the zone 0 28,270 540.64 727.39
The number of retail trade jobs 0 53,458 1,314.02 1,559.76
The number of transportation services jobs 0 59,856 918.80 1,346.47
The number of information services jobs 0 13,122 332.87 557.90
The number of financial services jobs 0 24,572 612.85 1,077.72
The number of real estate service jobs 0 19,163 274.63 454.37
The number of professional services jobs 0 40,753 1,162.99 1,969.62
The number of managerial jobs 0 469 18.86 38.31
The number of administrative jobs 0 32,587 615.26 848.28
The number of education jobs 0 21,350 671.13 848.68
The number of medical jobs 0 37,000 805.58 1,067.84
The number of entertainment jobs 0 10,657 225.74 379.44
The number of accommodation jobs 0 44,921 654.43 1,081.57
The number of other service category jobs 0 50,949 996.68 1,789.32
The number of public administration jobs 0 22,553 799.63 1,046.15
Bus stations within 40 miles 0 130 14.02 21.63
Rail Stations within 50 miles 0 40.83 4.83 7.03
Distance from zone centroid to nearest rail station 0 48 12.11 12.23
Airports within 100 miles 0 11.12 3.54 2.24

3.3.2 Model implementation

The rJourney software tool allows several system parameters to manage the simulation process
in terms of runtime and computational power. The simulation control parameters and their
values used in this study are listed in Table 3-2. For the simulation process, the household
sampling rate was set to 100 (i.e., 1% of the households were sampled with the expansion factor
of 100). The long-distance travels are defined based on the distance between the TAZs. Past
studies used different distance thresholds in defining the long-distance trip. The most common

12
trip distance threshold used in past studies is 100 miles in one direction of travel (Bierce & Kurth,
2014). In this research, the long-distance travel cutoff value was set to 100 miles.
Table 3-2 Parameter values used in scenario analysis

Parameters Values
Household Sampling Rate 100
Months Simulated All
Each Day of Month Separately? Yes
Use Probabilities in Trip Matrix? No
Expansion Factors 100
HH Records (Million) 1.1
Long Distance Travel cutoff value (miles) 100

The rJourney can be used to study system-wide changes (e.g., impact of higher travel cost) and
their impacts on long distance travel distribution. System-wide changes that can be modeled are
household income, auto travel cost, auto travel time, air fare, and rail travel time. Four future
scenarios based on transit accessibility improvement, household income, auto travel cost, and
auto travel time were developed to investigate associated impacts on long-distance travel
demand in Tennessee.

3.3.3 Scenario analysis

Long-distance travel pattern in the state for base case scenario is presented in Figure 3-4 with
the updated 2019 land use data. As evident from the figure, the highest number of long-distance
trips associated to Nashville, Kingsport, and Knoxville. After modeling base scenario (no change
scenario), the long-distance travel demand models were simulated for four future scenarios
which are summarized in Table 3-3. These four future scenarios were developed to assess the
effect of relevant policy sensitive variables on long-distance travel patterns in Tennessee. The
result from the sensitivity analysis and key findings are discussed in the following subsections.

Figure 3-4 Long distance trips per year in Tennessee for base scenario

13
Table 3-3 Summary of the proposed scenarios

Min Max
Scenarios Description Properties
Change Change
1 Transit access improvement # of bus stops 30% 50%
# of rail station 15% 25%
# of airport 20% 30%
2 Air fare policy Air fare -30% -50%
Lower congestion along major
3 Auto travel time -25% -50%
tourist routes
4 Increase in household Income Household income 15% 30%

3.3.3.1 Scenario 1: Impacts of transit access improvement

High-quality and easily accessible transit services are critical for encouraging people to switch
from personal vehicles to public transportation (Liu et al., 2019). Several states developed
tourism strategic plans by focusing on transportation services as a key element in overcoming
mobility challenges and promoting tourism. According to Virginia Tourism Corporation, increased
air access, rail, and mass transit for strategic locations can promote tourism travels (Virginia
Tourism Corportation, 2013). In this scenario, impact of higher transit access to major tourist
destination in Tennessee was assessed in terms of increase in long-distance trip volume. For this
purpose, the number of bus stops within 40 miles of the NUMA centroid was increased by 30%
to 50%. To estimate the impact of increased rail access, the number of rail stations within 50
miles of the NUMA centroid was increased by 15% to 25%. Similarly, the number of airports within
100 miles radius of NUMA centroid was increased by 20% to 30% to estimate impact of increased
accessibility to air travel. This transit access growth forecast is based on the proportional and
modest expansion of the transit ridership for the next 20 years in the state (Tennessee
Department of Transportation, 2016).
3.3.3.2 Scenario 2: Impacts of air fare policy on tourism travel
Past studies reported that reduced transit fare targeted to specific groups of people resulted in
transit ridership increase (Darling et al., 2021). Furthermore, the survey (conducted in this study)
among tourism stakeholders revealed that reduced fares for the tourism transit services can
increase tourist volume and reduce auto dependency among visitors. It has been reported that
doubling the air fare would decrease the long-distance trips by 8% and increase modal shift from
air to auto by 3.2% (Outwater et al., 2015). In this scenario, air fares will be reduced by 25 to 50%
to estimate the impacts of reduced air fare policy on tourism trips.
3.3.3.3 Scenario 3: Impacts of lower congestion along major tourist routes
Several tourism agencies mentioned traffic congestion as one of the limiting factors of tourism
growth (Gonzalez-Rivera, 2018; Virginia Tourism Corportation, 2013). According to Virginia
Tourism Corporation, traffic congestion negatively impacted tourist volume and reduced cross-
visitation trips among tourist destinations (Virginia Tourism Corportation, 2013). Alleviating traffic
congestion can decrease the time people spend traveling, increase time spent at tourist
attractions and encourage people to visit additional tourist attractions. In this scenario, we
simulate the traffic congestion reduction by decreasing the auto and bus travel time for major
tourist zones in Tennessee. The auto travel time is reduced by 25 to 50% to estimate associated
improvement in tourism trips, based on past findings (Outwater et al., 2015).

14
3.3.3.4 Scenario 4: Impacts of increase on household Income
Previous study suggests a direct relationship between the frequency of recreational long-
distance and higher household incomes (Outwater et al., 2015). Household incomes also affect
mode share for long-distance trips as people with higher household incomes tend to use
personal vehicles more frequently than public transport. In this scenario, the household income
was increased by 15 to 30% (Outwater et al., 2015).

3.4 Analysis of popular attractions

In the fourth step, an analysis of popular attractions in the state was undertaken to gather a
synopsis of accessibility to tourist destinations. A list of 50 most popular attractions was compiled
based on their popularity in Google reviews (see Appendix C for the list). These attractions were
shortlisted from destinations included in the 2022 Tennessee Vacation Guide (The Soundtrack of
America: Made in Tennessee (2022 Vacation Guide), 2022). Finally, only those with at least 250
reviews (4.5 stars and above) were enlisted.

Figure 3-5 Popular attractions in Tennessee

These destinations were analyzed in two steps. In the first step, access to these destinations was
evaluated based on travel time and distance from the nearest i) Urbanized Areas (UAs), ii) primary
service commercial airport, iii) interstate, or transit stop. Figure 3-5 shows the locations of the
destinations along with the UAs, the five primary commercial airports, and statewide interstate
system. In the second step, trips made to major destinations were analyzed using INRIX Trip
Analytics. Since it was impractical and cumbersome to analyze trips made to all these
destinations five cities: Memphis, Nashville, Chattanooga, Knoxville, and Gatlinburg were
considered for analysis as they are home to many several attractions. Trips made to these
locations through major highways were used to extract origins, destinations, and number of trips.
The primary goal of this task was to identify major origin markets from where people traveled.
In addition, the national long distance passenger travel demand model was used to identify the
10 Origin-Destination (O-D) pairs within the state with the most trips. To get an idea of current
traffic flow conditions on the route serving these O-D pairs, values of Travel Time Index (TTI) for
weekdays and weekends were calculated using data from INRIX Trip Analytics.

15
3.4.1 Access to destinations
3.4.1.1 Access from nearest urbanized area
According to the urban-rural classification presented by the Census Bureau, UAs are those with
a population greater than 50,000 people. According to the 2020 census data, there are 15 UAs in
Tennessee (2020 PL 94-171 Redistricting Data Summary File, 2022). These cities are outlined in
Table 3-4. As an indicator of access, the travel distance and travel time between an attraction and
the nearest UA was derived from Google Map searches. Of all available routes, the route with the
shortest travel time during PM peak was considered as the measure of access.
Table 3-4 Urbanized Areas in Tennessee based on the 2020 census

City Population City Population City Population

Nashville 715,884 Murfreesboro 152,769 Bartlett 57,786
Memphis 633,104 Franklin 83,454 Kingsport 55,442
Knoxville 190,740 Johnson City 71,046 Smyrna 53,070
Chattanooga 181,099 Jackson 68,205 Collierville 51,324
Clarksville 166,722 Hendersonville 61,753 Spring Hill 50,005

3.4.1.2 Access from nearest primary service commercial airport

Access from nearest primary service commercial airports to the tourist destinations was also
determined using travel distance and travel time. Primary service commercial airports are
publicly owned airports that have more than 10,000 passengers boarding each year. These
airports serve as major hubs for passengers flying to and from destinations. There are five such
airports in Tennessee: Memphis International airport, Nashville International Airport, McGhee
Tyson Airport, Chattanooga Metropolitan Airport, and Tri-Cities Regional Airport. The distance
and travel time from the nearest airport to the attractions when using the fastest route during
the PM peak hours was obtained using Google Maps (Google Maps, 2021).
3.4.1.3 Access from nearest interstate and transit stops
Figure 3-5 suggests that the identified attractions are located within and outside the UAs. To
ascertain access to the destinations, two metrics were used. For the destination inside UAs, the
travel distance and travel time from nearest interstate exits and transit stops was determined.
Access to destinations outside UAs was determined based only on the distance from nearest
interstate exit since they are not accessible using transit services.
The travel time and distance from transit stops to the destinations was determined based on
either walking or driving depending on the distance between the stop and the destination. An
assumption was made that passengers traveling more than 0.5 miles from transit stops would
refrain from walking and use cars or other modes (usually micro-mobility options) for their travel
(Nabors et al., 2008). Nearest transit stops to the attractions within Memphis, Nashville,
Chattanooga, and Knoxville were determined using stops detailed in General Transit Feed
Specification (GTFS) data. For certain cities, Pigeon Forge, Gatlinburg, and Kingsport, maps posted
online by respective transit agencies were used to identify the stops. These stops were then
manually located on Google Maps of further analysis.

16
3.4.2 Analysis of trips for identification of origin markets
In the second step of analyzing tourist attractions, an analysis of origin and destinations was done
using INRIX Trip Analytics (INRIX, 2022). The goal of this task was to determine traffic volumes
traveling from origins to major destinations through major highways. This enabled identification
of major tourism routes and the priority origin markets. Only major highways were included in
our analysis because trip and traffic data in INRIX Trip Analytics are limited to major roads. The
analysis of trips was done at the county level since most of the attractions were clustered
together at this geographical resolution. It is worth noting that INRIX supports analysis at smaller
geographical resolutions such as sub-county, TAZ, and ZIP levels. This, however, is
computationally demanding with the resulting visualizations difficult to follow particularly when
origins and destinations are scattered over larger geographical regions, e.g., out of state.
Additionally, results from survey of tourists indicated that about 64% of recreational travels were
undertaken between April and August (Figure 4-8). Therefore, to reduce the computational times
our analysis included trips made between April 2021 to August 2021. Only light and medium
vehicle trips were included in our analysis.

3.4.3 Analysis of popular tourist routes

3.4.3.1 Identification of O-D pairs for analysis
As mentioned before, the national long distance passenger travel demand is a tour-based
simulation model for generating long distance travel behavior and patterns. The long-distance
model first estimates tour generation, scheduling, duration, and party-size models by tour
purposes. After that, the model generates individual and origin-destination matrix for different
modes and trip purposes. Tour purposes for national long-distance travel includes leisure and
vacation, visit to friends or relatives, personal business, commuting, and employer’s business.
Four modes can be modeled in accomplishing the estimated trips by purpose, which are personal
cars, intercity bus, intercity rail, and commercial air travel. Figure 3-6 represents the structure of
the tourism long-distance travel demand modeling system. As seen from the figure, final models
from rJourney produce individual tours with purpose, duration, and month. Also, the mode and
destination choice model produce individual tours with mode and destinations all over the US.
In addition to that, the microsimulation model produces number of trips for auto, bus, rail, and
air mode for different O-D pairs based on the NUMA zones.
Using the long-distance trips O-D matrix for four different modes, total trips made between the
NUMA zones are quantified. After that, the O-D pairs were filtered by selecting the pairs that had
both origins and destinations within the state. Once the O-D pairs within Tennessee are selected,
these pairs are sorted in a descending order based on the total number of trips made between
them. Based on the total trips, the top 10 O-D pairs within the state were selected.

17
 Figure 3-6 Identification of O-D pairs for analysis route analysis

3.4.3.2 Analysis of traffic conditions on popular routes

After identification of O-D pairs, major routes used for traveling between the O-D pairs on
Weekdays and Weekends was identified using INRIX Trip Analytics. Trips were analyzed for the
months April to August in 2019. TTI was then calculated for the most used route to check traffic
conditions on the route on weekdays and weekends.

3.5 Transportation services for tourism in Tennessee

In the fifth step, an online survey was administered to tourism agencies across the state to gather
information on the current state of transportation services in the state. These agencies included
chambers of commerce and local agencies that are responsible for developing tourism in their
area/jurisdiction. The list of agencies for the survey was compiled from the TDTD website and
2022 Tennessee Vacation Guide (The Soundtrack of America: Made in Tennessee (2022 Vacation
Guide), 2022). The list of agencies is provided in Appendix E. The survey administered on the
agencies consisted of questions associated with the agencies’ perception of current state of
transportation services in their area and the state, collaboration between agencies, issues in
transportation systems, and measures to address them. The survey was developed in Qualtrics
and emailed to the agencies along with a brief description of its objectives. The online survey
took about 8 minutes to complete. Of the 50 agencies that responded to the survey, 33 were
complete. Summary of these complete responses is presented in Chapter 4: Results and
Discussion.

18
Chapter 4 Results and Discussion
4.1 Survey of state DOTs and STOs
The survey of state DOTs and STOs included questions on specific tourism related travel demand
modeling practices, tourism data sources and data analysis methods, tourism inclusive project
selection practices, and collaboration manner between diverse tourism stakeholders. A brief
discussion of responses obtained on interagency collaboration and private sector involvement in
tourism transportation planning based on the survey is presented in the following two
subsections. As outlined in the Methodology section, analysis of survey responses was done
using k-means clustering which is detailed in Appendix A and summarized here.

4.1.1 Discussion of the Survey result

After application of the clustering technique, IPF was used to generate data for states that did
not respond to the survey using the collected survey data. IPF method was applied to each cluster
so that the generated synthetic data was more representative within each cluster. The clusters
were compared based on the collected and synthetic survey data related to agency collaboration
and private sector involvement, which are provided in the following sections (Ashraf et al., 2022).
This discussion focuses mainly on medium tourism impact states and low tourism impact states.
The reader may refer to Appendix A for a detailed description of classification of states based on
tourism impact.
4.1.1.1 Interagency collaboration in tourism transportation planning processes
Collaboration between state DOTs and STOs is critical in developing successful tourism
supporting transportation infrastructure and services (Petraglia & Weisbrod, 2004). State DOTs
and STOs were asked to select the nature of their current interagency collaborations. In general,
the survey results showed a lack of effective collaboration. DOT responses showed that
collaboration with STOs in transportation project planning and implementation was close to the
average for low and medium tourism impact states (Figure 4-1 (a)). It should be noted that in
Figure 1, the average responses are scaled from Never (1) to Always (5). Among the STO
respondents, medium tourism impact states had better transportation infrastructure and
services for the tourism industry than low tourism impact states (Figure 4-1 (b)). This implies that
better transportation infrastructure contributed to the economic impacts of tourism in medium
tourism states compared to low tourism impact states. States in the medium tourism impact
cluster had a better collaboration with state DOTs, neighboring state’s STOs, and public and
private sector stakeholders. Collaboration with state DOTs was close to the average for medium
tourism impact states and below average for low tourism impact states. Regarding collaboration
with other public and private sectors in tourism-related transportation project development,
states in the lower tourism impact cluster had slightly better practices than other states.
However, the frequency of collaboration with neighboring state DOTs and STOs in planning and
implementing tourism transportation projects was still less than “half of the time” for medium
and low tourism impact clusters. In collaboration practices, states with medium tourism impact
were better than those within low tourism impact clusters.
State DOTs and STOs were also asked about the primary obstacles to effective interagency
collaboration and actions/policies that can be taken to improve it. About 28% of the state DOTs
and 33% of the STOs stated identified lack of established policies as the primary obstacle. Lack

19
of funding was another obstacle identified by the responding agencies. The agencies mentioned
that state-level policies focused on collaboration need to be developed and set a requirement in
the project selection process to increase collaboration among tourism stakeholders.

(a) State DOT perspective

(b) STO perspective

Figure 4-1 Collaboration manner among state DOTs and STOs.

4.1.1.2 Private sector involvement

NCHRP Report 419: Tourism Travel and Transportation System Development identified eleven
principles to promote tourism and associated economic activity in the traditional transportation
planning and project development process (Frechtling et al., 1998). The involvement of private
sector stakeholders in tourism-related transportation planning was one of the critical elements
of these principles. 15 states (33% of states) in the medium tourism impact cluster reported that
the private sector plays an active role in tourism-related planning. On the other hand, only 9% of
state DOTs (3 out of 32 states) in the low tourism impact cluster reported the same for tourism-
related transportation infrastructure and service planning (Table 4-2 (a)). In addition, 42% (6
states out of 15) of the STOs in the medium tourism impact cluster and 12.5% (4 states out of 32)
of the state tourism offices of low tourism impact cluster occasionally engage in transportation
project selection (Table 4-2 (b)). The responses also provided insights into their agency's role in

20
tourism-focused transportation systems and service-related investment decision-making.
85.71% and 71.4% of tourism agencies reported they play a supporting role in tourism-related
transportation project selection and implementation in medium and low tourism impact clusters.
Overall, there is a lack of private sector involvement in the tourism and transportation-related
project selection process, and it should be addressed to realize the economic impacts of the
tourism sector.

(a) State DOT perspective

(b) STO perspective

Figure 4-2 Role of private sector stakeholders in tourism transportation service development

4.2 Analysis of tourist trip characteristics

4.2.1 Summary of trip characteristics
4.2.1.1 Sociodemographic characteristics of tourists
This section presents a summary of results from a survey administered on tourists who visited
recreational destinations in Tennessee in the past.

21
Figure 4-3 presents the distribution of tourists’ socio-demographic attributes. Clearly, almost half
of tourists visiting the state are young individuals less than 25 years old. Almost 32% of visitors
are full-time employees. Notably, the state is more popular among retired individuals who
account for more than 33% of all tourists. The distribution of the tourists’ personal annual income
suggests that tourists from all income classes travel to Tennessee. The marital status of tourists
suggests that there were more married and single individuals visiting Tennessee.

Figure 4-3 Socio-demographic characteristics of tourists visiting Tennessee

4.2.1.2 Household characteristics

Household characteristics of travelers presented in Figure 4-4 indicates that household income,
as with personal income, has no considerable influence on travelers. People living in rural or sub-
urban areas constitute about 68% of tourists. People belonging to smaller households are more
likely to visit Tennessee. While it seems that households with 3 more vehicles are less likely to
visit the state, it could be that these households were less in number since this variable was not
controlled using quota sampling when conducting the survey.

22
 Figure 4-4 Household characteristics of tourists visiting Tennessee

4.2.1.3 Travel mode and duration of stay

Figure 4-5 shows primary mode used to travel to Tennessee by distance traveled. Almost 40% of
visits are more than 500 miles away from their origin. Regardless of the distance, majority of trips
are made using household or owned vehicles. About 8% of trips more than 500 miles are made
using private or commercial airplanes. Notably, with an increase in travel distance there is a
consistent increase in air travel.
Figure 4-6 shows distribution of duration of stay. We can see about 60% of visitors stay from 2-4
nights at the destination.

23
 Figure 4-5 Travel mode by distance traveled

Figure 4-6 Duration of stay at destination

4.2.1.4 Satisfaction of tourists

Satisfaction of tourists for various aspects of tourism was collected using a 10-point Likert scale
from 1-Poor to 10-Excellent. In general, high satisfaction ratings were observed across all aspects.
Notably, satisfaction for value for money was comparatively less than other aspects.

Figure 4-7 Satisfaction ratings from tourists traveling to Tennessee

24
4.2.1.5 Trip timing
Figure 4-8 shows how tourism travel is distributed across a year. Typically, tourism travel peaks
in June. April, May, June, July, and August together account for about 64% of all travels made to
Tennessee. Summer breaks could be a reason for this.

Figure 4-8 Distribution of travel across a year

4.2.1.6 Trip frequency

Figure 4-9 presents recreational frequency of tourists who responded to the survey. More than
three fourths of all respondents indicated that they traveled at least once a year.

Figure 4-9 Recreational travel frequency of tourists

4.2.2 Antecedents to recreational travel intentions

4.2.2.1 SEM Analysis
The analysis of survey responses related to recreational travel attitudes during the COVID-19
pandemic was done using SEM (Thapa et al., 2022). The survey data was first tested for sampling
adequacy using Kaiser-Meyer-Olkin's measure and found to be excellent with an overall value of
0.91. Bartlett's test of sphericity was also found to be excellent with p < 0.001. Exploratory Factor
Analysis (EFA) was conducted on the attitudinal questions (also called indicators in SEM) using
varimax rotation and a cutoff of 0.4 for factor loadings (Guadagnoli & Velicer, 1988). This resulted
in ten factors (also called constructs or latent variables) that captured a cumulative variance of
63%. Cronbach’s alpha for all the extracted factors was greater than 0.75 and therefore
satisfactory (Faul et al., 2009). Results from factor analysis is outlined in Appendix B. The
extracted factors were used to create SEM model outlined in Figure 3-2. Model fit indices for the
model were also satisfactory with χ2 = 7334.07; df = 2066; p = 0.00; Root Mean Square Error of
Approximation (RMSEA) = 0.045, 95% CI = [0.044, 0.046], PCLOSE = 1.00; Tucker-Lewis Index (TLI)
= 0.94; Comparative Fit Index (CFI) = 0.95. The consensus regarding the cutoff values for good fit
indices are: RMSEA < 0.05, PCLOSE > 0.05 (higher the better); TLI > 0.9; and CFI > 0.9.

25
The measurement and structural equation models for our SEM are also provided in Appendix B.
A summary of decisions for our hypotheses is presented in Table 4-1. Note that for hypothesis
H2a, the relationship was statistically significant but with an inverse relationship of what was
proposed, therefore it is not supported. The decision on the remaining hypotheses is based on
the statistical significance of their path coefficients.
Table 4-1 Summary of results for hypotheses used in investigating travel intentions

Path
Hypotheses Relationships t-stat Decision
coefficient
H1a Public trust → Travel attitude 0.54* 12.14 Supported
H1b Public trust → Behavioral intention 0.27* 8.56 Supported
H1c Public trust × Travel concern → Behavioral -0.07* -2.21 Supported
intention
H2a Subjective norm → Travel attitude -0.36* -10.48 Not supported
H2b Subjective norm → Behavioral intention 0.39# 9.03 Supported
H2c Subjective norm × Travel anxiety → Behavioral -0.02 -0.60 Not supported
intention
H2d Subjective norm → Travel composure 0.45# 11.91 Supported
H3 Travel attitude → Behavioral intention -0.02 -0.57 Not supported
H4a Perceived behavioral control → Travel composure 0.26# 7.08 Supported
H4b Perceived behavioral control → Behavioral 0.45# 10.03 Supported
intention
H5 Travel composure → Behavioral intention 0.04 1.08 Not supported
H6a Perceived knowledge → Perceived benefits 0.24# 7.36 Supported
H6b Perceived knowledge → Behavioral intention 0.07 1.65 Not supported
H7 Perceived benefits → Behavioral intention 0.54# 12.17 Supported
Significance levels: *0.05, #0.001. Exceptions are statistically insignificant at 5% level of significance.

4.2.2.2 Necessary Condition Analysis

Figure 4-10 NCA effect sizes (All effect sizes significant at p=0.01)

Figure 4-10 shows the effect sizes of the predictor variables. The stacked bars represent the
outcomes with the effect from each predictor endogenous latent variables within them. Note
that the height of the bars represents the cumulative effect of all its predictors. Following our
rationale for determining necessary conditions, only the predictors with d ≥ 0.1 at a 5% level of
significance are considered to have a measurable effect and therefore presented in the figure.

26
The effect of endogenous latent variables with 0.1< d < 0.3 suggests medium effects. Notably, the
effect sizes of trust on attitude; subjective norm on comfort; perceived benefits, and perceived
behavioral control on behavioral intention is greater than 0.2. No significant effect sizes were
found for attitude, travel concern, travel anxiety for behavioral intention at a 5% level of
significance. The moderating effect of trust on travel concerns was found to be statistically
insignificant. Similarly, perceived behavioral control and subjective norm are necessary in
defining travel composure. Finally, public trust, subjective norm, perceived benefits, perceived
behavioral control and perceived knowledge are necessary to manifest intention.
The bottlenecks for predictors with medium effects are presented in the Appendix B and may be
interpreted as follows. A certain value of the predictor variable is necessary for a certain effect to
be manifested in the response variable. This amount is presented as a percentage of the
maximum value of the variable.
The interpretation of results from SEM and NCA can be based on three scenarios which are stated
below. Within the parenthesis, we also present the predictors and outcomes for which each
scenario was observed. In our analyses, only the latent constructs shown in Figure 4-10 were
significant in NCA (Scenarios 1 and 2 of possible scenarios for interpretation of the necessary
condition discussed below. A more detailed description is included in Appendix B). This suggests
a certain level of these variables is necessary for the outcome to manifest. These levels are shown
as bottlenecks in Figure B-1 in Appendix B.
1. Predictor is significant in SEM and NCA [Scenario-1]: A change in the predictor variable will
change the outcome but a certain level of the predictor variable is necessary for the outcome
to manifest. (Subjective norm → travel composure, perceived behavioral control → travel
composure; public trust → behavioral intention, subjective norm → behavioral intention, perceived
benefits → behavioral intention, perceived behavioral control → behavioral intention)
2. Predictor is not significant in SEM but significant in NCA [Scenario-2]: A certain level of the
predictor variable is necessary for the outcome to manifest but a change will not affect the
outcome. (Public trust → travel attitude; perceived knowledge → behavioral intention)
3. Predictor is significant in SEM but not in NCA [Scenario-3]: Change in the predictor variable
will change the outcome and no necessary condition exists for the predictor variable to
manifest the outcome. (Subjective norm → travel attitude; perceived knowledge → perceived
benefits)
For this scenario, we find that no minimum levels of travel anxiety and concerns are necessary
to manifest travel intention, but they are sufficient to manifest the outcome. A similar effect
of subjective norm and perceived knowledge on travel attitude and perceived benefits
respectively is also observed.

4.2.3 Conclusion and implications

Findings from investigation of travel intentions have far-reaching implications in terms of policy
considerations. The relationship between public trust and travel attitude in NCA suggests that
public trust in government is essential in developing a positive travel attitude (Q. Han et al., 2021).
Also, travel intentions can be promoted if the public’s knowledge of the pandemic is increased.
Better dissemination of COVID-19 related knowledge is also crucial to keep people informed so
they have higher perceived knowledge to develop travel intentions. Dissemination of correct

27
information through government channels can also increase travel intentions by developing trust
in the public. Campaigns to keep the public informed can be helpful in this regard. Secondly, the
public’s perceived benefits also have a measurable effect on intentions. Better incentives with
flexible booking and cancellations can also potentially increase perceived behavioral control and
therefore travel intention. Although compared to public trust, the necessary condition for the
subjective norm is lower, favorable subjective norm could follow with better knowledge of the
pandemic. Our finding suggests that subjective norms can reinforce travel intentions. It is logical
to assume that subjective could improve with public trust and perceived benefits. Therefore, we
recommend better information dissemination and providing travel incentives as the two most
crucial measures that can be adopted at the policy level for the quick economic recovery of the
travel industry.

4.3 Analysis using national long-distance travel model

The results from the four scenarios analyzed using the national long-distance passenger travel
demand model are as follows.
4.3.1 Scenario 1: Impacts of transit access improvement
Transit system accessibility measures physical access to public transit stops by evaluating the
distance, time, and convenience of accessing transit stops (Malekzadeh & Chung, 2020). In this
scenario, transit accessibility was improved by increasing bus and rail stops, and higher number
of airports within certain distance from each TAZ (discussed in the next section). The resulting
mode share under the increased transit accessibility scenario was compared with the base case
scenario. Figure 4-11 shows the change in mode share due to change in transit accessibility. As
expected, when there is better access to transit services, people would use them (rail and bus)
and air travel more frequently for long-distance trips. On the contrary, the use of personal cars
would decrease. Specifically, use of personal car could decrease by 0.93%-1.98%. The
corresponding increase in mode share for bus, rail, and air could be 0.11%-0.35% and 0.18%-
0.36%, 0.64%-1.34%, respectively.

Figure 4-11 Results from Scenario 1: Change in mode share in Tennessee

These potential impacts of transit accessibility increase can be used in tourism policy
development for promoting sustainable transportation services for tourism in Tennessee. The
change in mode share under increased transit access indicates that people could use more public

28
transit services if the service is widely available. Figure 4-12 shows the top twenty O-D pairs for
long-distance travel for trips that had both origin and destination within the state, and most trips
were centered around Knoxville. Policy makers can use the O-D pairs to prioritize transit service
improvement focusing on long-distance trips. With limited resources, O-D pairs with more long-
distance trips could be prioritized for new services or any existing transit services between those
zones can be improved. In addition, new routes and more frequent transit services can be
introduced along the most frequent O-D pairs.

Figure 4-12 Top 20 O-D pair under increased transit access scenario

4.3.2 Scenario 2: Impacts of air fare policy on tourism travel

One potential policy to promote more frequent tourism trips is to reduce fares for travel modes
serving major tourist destinations. In this scenario, air fare was decreased by 30% to 50% to
estimate associated impact on mode share and volume of long-distance trips. The change in
mode share is shown in Figure 4-13. The increase in long-distance trips using air mode could
mostly replace long-distance trips using personal cars. The estimated increase in mode share for
air is 1.98% and 1.45% for 50% and 30% air fare reduction, respectively. On the other hand, the
estimated decrease in personal car trips is 1.8% and 1.3% for 50% and 30% air fare reduction,
respectively. Mode share for bus and rail can also be expected to decrease under this scenario
although the estimated change is not as large as compared to personal car and air travel. Under
this scenario, the distribution of trip purposes could also change.
According to the tourist survey undertaken during this project and also past studies, the two most
widely used modes of tourism travel are air and personal cars (Speakman, 2005). The findings of
this research indicate that reducing air fares can increase tourism trips substantially. People
could be more encouraged to perform tourism trips if statewide tourism policy reduces air and
transit fares. A previous study has reported an increase in transit ridership when transit fares are
reduced and targeted to serve specific user groups (Darling et al., 2021). Thus, the transit fare
reduction can be targeted towards tourism trips via promotional initiatives and policy levels to
assess the proper balance between tourism development and revenue generation.

29
 Figure 4-13 Result from Scenario 2: Change in mode share due to reduced air fare

4.3.3 Scenario 3: Impacts of lower congestion along major tourist routes

Traffic congestion along major tourist corridor degrades tourism travel experience and reduces
tourism trips (Virginia Tourism Corportation, 2013). In this context, it is necessary to simulate the
effect of congestion on long-distance trips. The reduced congestion scenario was simulated by
reducing the auto travel time by 25% to 50% for the long-distance trips based on past research
(Outwater et al., 2015). The resulting change in long-distance trip generation pattern in
Tennessee is shown in Figure 4-14. It is observed that most of the NUMA zones could experience
an increase in long-distance trips under reduced congestion scenario. However, several areas
could also experience no change in auto trips as shown in the figure.

Figure 4-14 Result from Scenario 3: Change in long-distance trips under reduced congestion

4.3.4 Scenario 4: Impacts of increase on household Income

In this scenario, the household income was raised by 15%-30% to investigate its statewide impact
on long-distance trips. With the increase in household income, people are expected to make
more long-distance trips using personal cars (Virginia Tourism Corportation, 2013). With an
increase in household income people reduce long distance trips using public transit services.
Comparison of this scenario with base scenario shows that, for a 15% to 30% increase in
household income 5.8%-6.2% increase in long-distance trips using personal cars can be observed
as shown in Figure 4-15. Subsequently, there were 1.25%-1.46%, 0.14%-0.16% and 4.4%-4.5%
decrease in the share of bus, rail, and air modes, respectively.

30
 Figure 4-15 Result from Scenario 4: Changes in mode share with increased household income

An increase in household income is expected to influence tour purposes for long-distance trips,
as shown in Figure 4-15. Higher household income could increase leisure commute, and
employer’s business trip shares. Conversely, visiting friends and relatives, and personal trips
could decrease with higher household income. From our analysis, we observe the leisure trips
increases by 0.93% for 30% increase in income as shown in Figure 4-16. Policymakers should take
future transportation improvement projects considering the increased intra-state demand from
tourism travel resulting from changes in household income.

Figure 4-16 Result from Scenario 4: Changes in trip purpose with increased household income

4.3.5 Conclusion and implications

In this task, different scenarios based on transit accessibility improvement, reduced congestion,
reduced air fare, and increase in household income were developed to study future long-distance
tourism travel in Tennessee. Scenario analysis indicates that improving transit accessibility could
promote sustainable tourism transportation services. Further, tourists’ experience can be
improved by reducing congestion near the popular tourist destinations and along popular
tourism corridors. Policies that promote reduced transit fares can also promote tourism by

31
diverting people away from using personal cars and relieving congestion. In addition, increase in
household income could generate more tourism trips mainly from the use of personal cars. This
should be given due consideration when planning for future transportation system and services.
It is worth mentioning here that O-D analysis can be performed to identify patterns in long-
distance travel and prioritize areas and corridors for transit service improvement. Due to
limitations of the rJourney software (can model up to 4,700 zones), long-distance travel models
that consider smaller TAZs could be employed for in-depth analysis of policy impacts that are
associated with sustainable tourism system and services.

4.4 Analysis of popular destinations

4.4.1 Access to popular attractions
4.4.1.1 Access from nearest urbanized area
Analysis of access to tourist attractions was assessed based on distance and travel time. Travel
distance and travel times from nearest UA to the destinations are shown in Figure 4-17. The origin
UA and the destination is presented in the axis label. Destinations located within a UA are not
shown in the plot. Of the 28 attractions outside UAs, 18 are within 40 miles or an hour drive from
the nearest UA.

Figure 4-17 Access to destinations from its nearest UA

32
3.4.1.2 Access from nearest primary service commercial airport
Thirty nine of the fifty attractions are within 40 miles or an hour drive from the nearest primary
commercial airport (see Figure 4-18). The travel time and distance are proportionate for all
attractions except those located in Gatlinburg (Gatlinburg SkyLift Park, Anakeesta Theme Park,
Ober Gatlinburg Aerial Tramway). The travel time to these attractions is notably larger than the
travel distance.

Figure 4-18 Access from nearest primary service commercial airport

3.4.1.3 Access from nearest interstate and transit stops

Results showed that almost all attractions are within 40 miles of an interstate exit. A notable
observation was the large discrepancy between travel distance and travel time when traveling
from nearest interstate exit on I-40 to attractions in Gatlinburg. Similar observation was made
for access from nearest primary commercial airport to the attractions. Compared to others, the
travel times to these attractions were notably longer considering the distance traveled (Figure 4-
19).

33
 Figure 4-19 Access from nearest interstate exit

Figure 4-20 shows the distance and time taken for tourists to reach an attraction from nearest
transit stop by walking. The attractions in Memphis are mostly located close to the city downtown
which has good access to the transit system. While attractions in Nashville have access to transit,
it takes longer to travel when riding the transit and walking. The same is observed for attractions
in Pigeon Forge. Most attractions in Gatlinburg are also close to a transit stop.

34
 Figure 4-20 Access from nearest transit stop to the destination

4.4.2 Analysis of trips for identification of origin markets

Trips counts from INRIX Trip Analytics was used to identify major tourism markets by identifying
origin of long-distance trips. While INRIX Trip Analytics cannot provide actual trip volume, trips
aggregated over longer periods can provide a reasonably good estimate of traffic volumes for
comparison (e.g., qualitative information such as routes with more traffic, origin with more trips,
etc. can be used for relative comparison since quantitative data such as the actual trip counts is
not available). Notably, most of the attractions were located within Memphis, Nashville,
Chattanooga, Knoxville, and Gatlinburg. The major highways used to access these cities were
chosen to identify trips origins. The results from analysis of trips are included in Appendix D. The
results show the interstate segments, its length and the direction of travel considered when
executing data request in INRIX Trip Analytics. A summary of our findings is presented in Figure
4-21. Note that the figure highlights net trips and therefore only those states that produced more
trips than it attracted are shaded.

35
 (b) Nashville
(a) Memphis

(c) Chattanooga (d) Knoxville

(d) Gatliburg

Figure 4-21 Priority markets for tourism across Tennessee

4.4.3 Identification of popular routes

The ten intra-state O-D pairs with highest frequency trips identified using the national long
distance passenger travel demand model are presented in Table 4-2 along with the route, travel
distance, travel time, and Travel Time Index (TTI) for Weekday and Weekend trips. TTI is a metric
used to quantify congestion based on the travel times during congestion and free flow
conditions. It is calculated by dividing the average travel time by the free flow travel time.
Therefore, it shows the travel times in comparison to the free flow travel times. For example, for
the trips made from Hamilton to Knox County, a TTI of 1.035 suggests that during normal traffic
flow conditions, the travel times are 3.5% higher than in free flow conditions. TTI values in the

36
table are coded using the superscripts s, e, and g to highlight TTI values which are smaller, equal,
and greater for Weekday compared to Weekend trips, respectively.
Table 4-2 Travel Time Index for O-D pairs with most trips

Counties Weekday trips Weekend trips

Length Avg Avg
Rank Route Travel Travel
(mi) Travel Travel
Origin Dest. Time TTI Time TTI
time time
(mins) (mins)
(mins) (mins)
I Hamilton Knox I 75; I 40, I 75 73.29 60.33 62.47 1.035e 57.93 59.93 1.035 e
II Knox Hamilton I 40, I 75; I 75 70.8 58.65 60.45 1.031s 56.95 58.97 1.035g
I 40; I 40, SR
III Putnam Knox 67.85 57.05 59.02 1.034 s 55.93 57.95 1.036 g
299; I 40, I 75
IV Knox Sullivan I 40; I 81 72.08 59.00 60.62 1.027 e 57.15 58.72 1.027 e
I 40, I 75; I 40;
V Knox Putnam I 40, SR 58; I 67.68 54.58 56.73 1.039 s 53.23 55.57 1.044 g
40, SR 299
VI Sullivan Knox I 81; I 40 72.02 57.92 59.27 1.023 s 55.62 57.68 1.037 g
I 65; Four-
Forty
Parkway, I
VII William Knox 440; I 24; I 40; 165.74 142.05 146.72 1.033 g 139.17 143.18 1.029 s
I 40, SR 56; I
40, SR 299; I
40, I 75
I 40, I 75; I 40;
I 40, SR 58; I
40, SR 299;
VIII Knox William I 40, SR 56; 165.78 143.80 147.73 1.027 g 140.77 143.30 1.018 s
Four-Forty
Parkway, I
440; I 65
I 40; I 81; I 26,
US 23; I 26,
IX Knox Unicoi 94.26 81.53 82.20 1.008 s 80.10 81.90 1.022 g
US 19W, US
23
I 26, US 19W,
US 23; I 26,
X Unicoi Knox 82.69 81.73 82.93 1.015 g 84.45 84.88 1.005 s
US 23; I 81; I
40
TTI values suggest that on routes that consist primarily of interstates, travel times are generally
higher on Weekends. On the contrary, travel times on routes that include primary and secondary
highways are higher on Weekdays. This suggests that traffic flow is dependent on day of the week
and facility type.

4.4.4 Conclusion and implications

Assessment of travel distance and travel times to the attractions suggests that the auto travel
times are proportional to the travel distance. However, attractions in Gatlinburg are an exception.
Arriving at Gatlinburg takes notably longer that other attractions. We can conclude that traffic
flow on the highways that lead to Gatlinburg could be improved to reduce travel time. In terms
of access from transit stops, attractions in Nashville are further away from transit stops than any

37
other city. Understandably, agencies must deal with various constraints such as access
management, land topography and availability, and travel demand when planning for transit
routes and stops (e.g., (Chakraborty & Mishra, 2013; Mishra et al., 2012, 2015; Sharma et al., 2020;
Sultana et al., 2018; Welch & Mishra, 2013)). Nevertheless, transit routes and stops should be
planned with the objective to allow access to attractions.
Findings from analysis of trips using INRIX Trip Analytics data suggests that states bordering
Tennessee are the biggest origin markets. Most trips to major attractions in Tennessee
originating out of state are made from bordering states. However, it could be the case that people
traveling from other states mostly use air transportation. Our findings from survey administered
on tourists revealed that people are more likely to travel via air when travel distance is large (see
Figure 4-5). Nevertheless, travel to attractions is dominated by auto and air travel accounts for a
relatively small proportion of trips. Therefore, improvement of highway-based services should
be prioritized over air transportation. Congestion on popular highway routes in the state are not
of concern at present, as suggested by the small values of TTI. However, it is noteworthy that
congestion on highway routes can peak during weekends and weekdays based on the type of
facility. Furthermore, changes in household income and pricing policies can drastically alter
mode choice and highway performance as evidenced by our findings from scenario analysis.
Planners and policy makers should be wary of this.

4.5 Current state of tourism-related transportation system in

Tennessee
Local tourism agencies were surveyed to gather a synopsis of current deficiencies in the
transportation system and services and identify measures to address them. The results from the
survey administered on local tourism agencies, mostly chambers of commerce and local tourism
departments, are presented here. These results were aggregated from 33 complete responses
received from the respondents.

4.5.1 Performance of available transportation systems and services

Almost half of the agencies that responded to the survey opined that the current state of
transportation system and services in their area/jurisdiction was satisfactory (either moderate or
good) as shown in Figure 4-22. This suggests that there is a general dissatisfaction among local
agencies with the current state of transportation system.

Figure 4-22 Current state of transportation systems in the area

4.5.2 Need for dedicated budget

All agencies agreed when asked if dedicated budget was necessary to develop tourism related
transportation infrastructure in their area (Figure 4-23). More importantly, more than half of all
respondents strongly supported the idea of allocating a dedicated budget for developing tourism
transportation in their area.

38
 Figure 4-23 Need for dedicated budget

4.5.3 Tourism transportation planning and its role in Tennessee

The agencies’ opinion of various statements associated with tourism transportation planning and
its role on the economy was collected in the survey. These statements were as follows. Summary
of responses collected for these statements are presented in Figure 4-24.
1. Regional tourism-related transportation services decisions should mainly be made by a
committee of representatives from different stakeholders
2. Involvement of local stakeholders in tourism planning enhances sustainability of the
tourism and improves the sustainability of local environment
3. Well-coordinated planning for tourism-related transportation services is critical to
managing tourist volume at destinations
4. Impacts of tourism and transportation services should be integrated to the local
sustainability plans
5. Tourism associated transportation services is a well-developed industry and employment
sector in Tennessee
6. Transportation services play a predominant role in tourism development and community's
economy
7. Tennessee has enough involvement of local communities in developing transportation
services for tourism.
8. Tennessee has a well-developed plan for solving problems related to tourism
transportation services.
9. The tourism sector affects local transportation services negatively and deteriorating
transportation infrastructures in Tennessee.
10. The tourism sector plays a predominant role in the local economy in Tennessee.

39
 Figure 4-24 Opinions on current state of tourism related transportation services

Most agencies strongly believe that impacts of tourism should be considered when developing
local sustainability plans and involvement of local stakeholders is necessary to develop tourism.
According to the agencies, the tourism sector is particularly important for the local economy. On
the contrary, most agencies disagree that tourism associated transportation system and services
is well-developed in Tennessee.

4.5.4 Issues in transportation services

To identify current issues in transportation services, a list of issues was compiled from the
literature. The survey questions collected responses by asking the agencies to specify the extent
to which they agreed these issues were prevalent in their area. The resulting responses are
shown in Figure 4-25. Of all the issues, inadequate transit service was most prevalent followed
by a lack of comfortable services and traffic congestion on major routes.

40
4.5.4.1 Prevailing issues in transportation services available to tourists

Figure 4-25 Current issues in transportation services

4.5.4.1 Solutions to address current issues in transportation services

As a follow up question to current issues in transportation services, agencies were asked to
indicate potential solutions that could address the prevalent issues. The solutions that were
popular were increased transit accessibility, improve highway tourism signage and establishment
of interactive kiosks at high visitor traffic areas (Figure 4-26).

Figure 4-26 Solutions to address current issues in transportation services

Responses entered on “Others” category where respondents could add text:

1. Service from metro to destination locations in rural areas
2. Shuttle connectivity to tourism assets and attractions
3. Bicycling sharing and docking stations.

4.5.5 Ranking of initiatives to improve transportation system and services

The agencies were provided a list of initiatives that could improve transportation systems and
services and were asked to rank them based on their preference with 1 being most preferred
and 8 being the least preferred. The objective of this was to identify the most and least preferred
initiatives from the agencies’ perspective. The responses are summarized in Figure 4-27. Installing
tourism signage, highway maintenance and establishing multimodal transportation
infrastructure were notably the most preferred initiatives (mostly ranked 1st by the agencies).

41
Results also showed that availability of parking spaces and traveler information centers were of
least concern to the agencies (mostly ranked 8th by the agencies).

Figure 4-27 Ranking of initiatives to improve transportation system and services

4.5.6 Collaboration between agencies

There may be several obstacles in interagency collaboration. To identify them, agencies were
asked to select the most prominent obstacles which were listed from the literature. The
respondents could also provide their own answer as text. The agencies recognized lack of policies
and legislative guidance, lack of involvement of private and public agencies and lack of willingness
to cooperate between stakeholders as the biggest obstacles (Figure 4-28).
4.5.6.1 Obstacles in collaboration

Figure 4-28 Obstacles in collaboration between agencies

Besides the listed obstacles, the agencies provided the following (entered as text). These are
provided here verbatim:
1. We are a very rural area with no public transportation of any kind.
2. Lack of meetings over the last several years (COVID).
3. lack of effort to bring groups together.
4. private is worried about their location and don't always take a big picture view
5. Lack of Funding is a major obstacle that has kept ideas from moving forward.
4.5.6.2 Solutions and actions to foster collaboration
A follow up to the previous question focused on identifying solutions that could foster
interagency collaboration. The question asked the agencies to select from enlisted options or
provide their answers. Agencies indicated that establishment of policies and legislative
requirement would be most helpful in fostering interagency collaboration (see Figure 4-29).

42
 Figure 4-29 Solutions to foster collaboration between agencies

Besides the listed solutions, the agencies provided the following (entered as text). These are
provided here verbatim:
1. Better communication.
2. Look at proactive measures to avoid ugly and unsafe tourism from the private sector
3. Drinking vehicles like what has happened in Nashville at an unsafe volume. It's spreading all
across Tennessee
4. Invite tourism to the TPO and RPO meetings for input purposes. Integrate greenway planning
groups as well
5. Regional cooperative groups
6. Dedicated funding local and state

4.5.7 Conclusion and implications

The survey revealed an overall dissatisfaction regarding tourism transportation systems among
local agencies. Furthermore, agencies desire allocation of dedicated budget and better
collaboration between agencies. Agencies opine that these could be remedied by state level
policies and legislative guidelines. Opinions on transportation services and their role suggests
that tourism related transportation is not well-developed in the state and local communities want
greater involvement in planning tourism and related transportation services. In terms of current
deficiencies, issues related to transit services were more prominent. The most preferred
initiatives to address current transportation issues were identification and maintenance of roads
and installation of tourism signage on highways.

4.6 Policy recommendations

Several concerns were identified from tasks undertaken throughout the research period.
Particularly, the surveys highlighted a lack of inter-agency collaboration and need to greater
involvement of state and local tourism agencies in the transportation planning process. In this
section in Table 4-2, we present a list of alternatives to these issues based on common practices
adopted and recommended by agencies across the county.

43
Table 4-2 Policy recommendations to address current concerns

Objectives Alternatives

Increased Establish formal guidelines and policies that necessitate interagency collaboration
collaboration particularly for large transportation projects.
between state
DOT, tourism Conduct regular meetings between agencies (DOTs and tourism agencies) to share
office, and the information on recent projects and activities that are related to tourism and long-
private sector range planning.

Incorporate tourism benefits and concerns in statewide transport planning, project

Involve development, and highway maintenance.
tourism
agencies in Encourage inclusion of coordination for tourism in all aspects of DOT (such as
transportation planning, project development, construction, maintenance, etc.).
project Find additional opportunities to involve local and regional tourism stakeholders in
planning multi-modal planning.

Collect and Allocate dedicated budget in transportation projects to collect tourism travel data.
utilize tourism
TDOT can coordinate with TDTD on travel data collection and its utilization for long-
data for
range planning.
transportation
planning TDOT should collect and use relevant tourism market data in project planning.

Consider new ways to prioritize projects in addition to the current practice.

Give tourism due consideration in the statewide transportation plan.

Prioritize
transportation Establish principles that guide prioritization of transportation system/services that
system needs improvement.
improvement
Involve local and regional tourism stakeholders in the decision process when
prioritizing transportation projects.

Provide additional opportunities for local agencies to improve transportation system

Increase and services at local level, for example, through additional funds and grants.
involvement
of local Increase participation of tourism businesses and organizations in transportation
agencies project planning and execution through involvement opportunities and dissemination
of updates on current and future projects.

44
Chapter 5 Conclusion
The tourism industry is a major source of revenue for Tennessee. Historical and natural
destinations in the state attract increasing number of visitors every year. There is a need for an
efficient and sustainable transportation system capable of supporting visitors traveling to tourist
destinations. The current state of transportation system is of concern considering that major
cities in the state, Memphis, Nashville, Knoxville, Cleveland, and Chattanooga, are among the
most congested cities in the US. These cities are home to some of the most popular destinations
in the state. The goal of this study was to provide Tennessee Department of Transportation with
policy recommendations to improve transportation services dedicated to tourists.
This research surveyed state DOTs and STOs across the country to understand state specific
tourism related travel demand modeling practices, tourism data sources, and data analysis
methods, tourism inclusive project selection practices, consideration of sustainable
transportation for tourism, and collaboration manner between diverse tourism stakeholders.
The survey responses were analyzed using k-means clustering technique and three clusters were
identified: states with low, medium, and high tourism impacts. The results showed that states
with greater interagency collaboration was associated with higher tourism impact. Agencies
identified lack of budget and inter-agency collaboration as the primary obstacles to developing
sustainable tourism transportation system.
Tourists who had traveled to recreational destinations in Tennessee were also surveyed to gather
information on long-distance trip characteristics such as socio-demographics and household
characteristics of tourists, mode choice, and attitude towards travel particularly considering the
COVID-19 pandemic. The survey responses were analyzed using TPB and NCA to identify
predictors of travel intentions and its necessary predictors. Results showed that travel incentives
such as flexible booking and cancellations, and better dissemination of information related to
the pandemic were necessary to manifest travel intentions and encourage people to travel more.
In the next step, a scenario-based analysis was undertaken using the national long-distance
passenger travel demand model. The analysis assumed four scenarios to forecast potential
changes in tourism travel. These scenarios were: improve transit access, reduced air fare,
reduced congestion on popular tourist routes, and increase in household income. Improvement
in any of these conditions predicted an increase in tourist travel. For example, with a 30%
reduction in air fare, personal car trips can be expected to reduce by about 1.3%, similarly, with
a 30% increase in household income, about 0.93% increase in leisure trips can be expected. Using
the long-distance model, O-Ds within the state with the most trips were also identified.
The most popular attractions in the state were identified to assess their accessibility. Travel time
and travel distance to these attractions from nearest urban areas, primary commercial airport,
interstate exit, and transit stop were obtained using Google Maps and GTFS data. Analysis
showed that it took disproportionately longer time (compared to distance) to reach Gatlinburg
from the nearest interstate exit. This suggests that the tourist routes used to travel to Gatlinburg
needs improvement. Similarly, among cities with transit facilities, travel times from the nearest
transit stop to the attractions was higher for Nashville. This finding points to inadequate transit
facilities available to tourists in Nashville. Analysis of trips made to these destinations was also
done using INRIX Trip Analytics to identify origin markets. State that produced more trips than

45
they attracted were identified for major tourist area: Memphis, Nashville, Chattanooga, Knoxville,
and Gatlinburg.
Finally, a survey was administered on local agencies involved with tourism to get an
understanding of current state of transportation system in their area, its deficiencies, and
initiatives to address them. Local and regional chambers of commerce and tourist offices were
participants in the survey. The major findings of the survey were, (i) need of dedicated budget to
develop and maintain tourism transportation system/services, and (ii) encouraging interagency
collaboration through formal standards/guidelines. Also, for currently available transportation
service, problems related to transit services (inadequate and uncomfortable service) was
identified as the most prominent issue faced by agencies at the local level. The agencies ranked
maintenance of highways and installation of tourism signages as most preferred initiatives to fix
current deficiencies in transportation system.

Recommendations
Transportation systems play a significant role in tourism development by connecting tourism-
generating regions to destinations. The distribution, capacity, efficiency, and accessibility of
transport services can affect how a destination develops, visitors' mobility, and the connectivity
of tourist experiences within destinations. However, an increased number of tourists can create
challenges in terms of the sustainability of the tourism transportation system. Thus, proper
planning and policy development are necessary to maintain the sustainability of the
transportation system and destinations. This section presents policy and guidelines for
sustainable tourism transportation services.
1. Both previous literature and surveys conducted in this study indicated a lack of effective
collaboration among the tourism stakeholders. This barrier can be removed by improving
collaboration between tourism stakeholders in project development and implementation.
TDOT, state tourism offices and tourism agencies can collaborate by exchanging available
resources, and involving in the project development, selection, and implementation process.
Moreover, the resources that need better partnership pertain to tourism data collection and
its utilization. While TDOT is already involved in roadway infrastructure development related
to tourism, such as signage, rest areas, scenic turnouts, and scenic byways, we recommend
prioritizing such development based on tourism associated benefits.
2. To promote sustainability of the tourism transportation services, sustainable modes of
transportation and services should be developed and promoted. Cycling is becoming a
popular mode of transportation for its environmental and health benefits. States can use this
potential to develop policies to promote “cycling tourism”. These policies could include
construction of ‘scenic greenways’, bicycle lanes, park and ride, and bike-sharing services.
Attractions in Memphis, Nashville, Knoxville, Cleveland, and Chattanooga, which are
congested with limited transit access, can benefit the most from such initiatives.
3. Engagement of the private sector in tourism transportation service planning and
development must be increased through collaboration. The private sector can play a key role
in tourism-related transportation planning. The principal areas for private sector
collaboration with the DOT and state tourism offices are private sector funding, marketing,
data collection, and dissemination of tourism information.

46
4. Use of data-driven practices for project selection should be encouraged. Most state DOTs and
tourism agencies do not collect tourist Origin-Destination data. They also mentioned the high
cost of third-party data. In this regard, more innovative approaches to data collection can be
identified through inter-agency collaboration.
5. A more detailed analysis of current state of transportation infrastructure in popular tourist
areas and routes is warranted particularly in Nashville, Gatlinburg, and Pigeon Forge to
identify and remedy current deficiencies. More tourists can be attracted to these destinations
if cheaper alternative to car and better public transportation is available. Considering
substantial number of tourists travel to Tennessee from the neighboring states, routes that
connect major destinations to bordering stats should be prioritized for future development
and regular maintenance.

47
References
2020 PL 94-171 Redistricting Data Summary File. (2022). The University of Tennessee Knoxville.
https://tnsdc.utk.edu/data-and-tools/2020-census/2020-pl-94-171-redistricting-data-summary-
file/

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes,
50(2). https://doi.org/10.1016/0749-5978(91)90020-T

Anderson, L., Manning, R., Pettengill, P., Reigner, N., & Valliere, W. (2011). Indicators and Standards of
Quality for Transportation in Park and Tourism Settings: Differences by Trip Purpose and Travel
Context. International Symposium on Society and Resource Management.

Ashraf, M. T., Thapa, D., Dey, K. C., Mishra, S., & Golias, M. M. (2022). Integrating Tourism Travel with
Transportation Planning and Project Development Process: A survey of State DOTs and Tourism
Departments. 101st Annual Meeting of the Transportation Research Board.

Bernardin Jr, V. L., Ferdous, N., Sadrsadat, H., Trevino, S., & Chen, C.-C. (2017). Integration of national
long-distance passenger travel demand model with Tennessee statewide model and calibration
to big data. Transportation Research Record, 2653(1), 75–81.

Bierce, E., & Kurth, D. (2014). The use of three surveys for long distance travel estimates in california.
Transportation Research Board 93rd Annual Meeting. https://trid.trb.org/view/1290078

Centiment. (2021). https://www.centiment.co/

Chakraborty, A., & Mishra, S. (2013). Land use and transit ridership connections: Implications for state-
level planning agencies. Land Use Policy, 30(1), 458–469.
https://doi.org/10.1016/j.landusepol.2012.04.017

Darling, W., Carpenter, E., Johnson-Praino, T., Brakewood, C., & Voulgaris, C. T. (2021). Comparison of
Reduced-Fare Programs for Low-Income Transit Riders. Transportation Research Record, 2675(7),
335–349.

Davis, A. W., McBride, E. C., Janowicz, K., Zhu, R., & Goulias, K. G. (2018). Tour-based path analysis of
long-distance non-commute travel behavior in California. Transportation Research Record,
2672(49), 1–11.

Dul, J. (2016). Necessary Condition Analysis (NCA): Logic and Methodology of “Necessary but Not
Sufficient” Causality. Organizational Research Methods, 19(1), 10–52.
https://doi.org/10.1177/1094428115584005

Dul, J., van der Laan, E., & Kuik, R. (2020). Necessary Condition Analysis (NCA) with R (Version 3.1.0) A
Quick Start Guide. Organizational Research Methods, 23(2), 385–395.
https://doi.org/10.1177/1094428118795272

Erhardt, G. D., Freedman, J., Stryker, A., Fujioka, H., & Anderson, R. (2007). Ohio long-distance travel
model. Transportation Research Record, 2003(1), 130–138.

Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using G*Power 3.1:

48
 Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149–1160.
https://doi.org/10.3758/BRM.41.4.1149

Frechtling, D. C., Meyer, M. D., & Pisarski, A. E. (1998). Tourism Travel and Transportation System
Development. https://onlinepubs.trb.org/Onlinepubs/nchrp/nchrp_rpt_419.pdf

Frei, A., Kuhnimhof, T. G., & Axhausen, K. W. (2010). Long distance travel in Europe today: Experiences
with a new survey. Arbeitsberichte Verkehrs-Und Raumplanung, 611.

Frick, R., & Grimm, B. (2014). Long-Distance Mobility: Current Trends and Future Perspectives.
https://www.ifmo.de/files/publications_content/2014/ifmo_2014_Long_Distance_Mobility_en.pd
f

Gonzalez-Rivera, C. (2018). DESTINATION NEW YORK.

https://nycfuture.org/pdf/CUF_Destination_New_York.pdf

Google Maps. (2021). https://www.google.com/maps

GoogleTrends. (2021). GoogleTrends. https://trends.google.com/trends/?geo=US

Guadagnoli, E., & Velicer, W. F. (1988). Relation of Sample Size to the Stability of Component Patterns.
Psychological Bulletin, 103(2), 265–275. https://doi.org/10.1037/0033-2909.103.2.265

Han, H., Al-Ansi, A., Chua, B. L., Tariq, B., Radic, A., & Park, S. H. (2020). The post-coronavirus world in
the international tourism industry: Application of the theory of planned behavior to safer
destination choices in the case of us outbound tourism. International Journal of Environmental
Research and Public Health, 17(18), 1–15. https://doi.org/10.3390/ijerph17186485

Han, Q., Zheng, B., Cristea, M., Agostini, M., Bélanger, J. J., Gützkow, B., Kreienkamp, J., & Leander, N.
P. (2021). Trust in government regarding COVID-19 and its associations with preventive health
behaviour and prosocial behaviour during the pandemic: a cross-sectional and longitudinal
study. Psychological Medicine, 1–11. https://doi.org/10.1017/S0033291721001306

Heeb, Gi. (2021, February 16). U.S. Air Travel Dropped 60% In 2020 As Covid-19 Hammered Airlines.
Forbes. https://www.forbes.com/sites/ginaheeb/2021/02/16/us-air-travel-dropped-60-in-2020-
as-covid-19-hammered-airlines/?sh=695ad9516978

Holden, A. (2016). Environment and Tourism (Third). Routledge.

INRIX. (2022). https://inrix.com/

INRIX 2021 Global Traffic Scorecard. (2021). https://inrix.com/scorecard/

Kim, M.-S., & Stepchenkova, S. (2020). Altruistic values and environmental knowledge as triggers of
pro-environmental behavior among tourists. Current Issues in Tourism, 23(13), 1575–1580.
https://doi.org/10.1080/13683500.2019.1628188

Kuhnimhof, T., Frick, R., Grimm, B., & Phleps, P. (2014). Long Distance Mobility in Central Europe: Status
Quo and Current Trends. European Transport Conference. https://trid.trb.org/view/1340422

Liu, Y., Wang, S., & Xie, B. (2019). Evaluating the effects of public transport fare policy change together

49
 with built and non-built environment features on ridership: The case in South East Queensland,
Australia. Transport Policy, 76, 78–89.

Llorca, C., Molloy, J., Ji, J., & Moeckel, R. (2018). Estimation of a long-distance travel demand model
using trip surveys, location-based big data, and trip planning services. Transportation Research
Record, 2672(47), 103–113.

Malekzadeh, A., & Chung, E. (2020). A review of transit accessibility models: Challenges in developing
transit accessibility models. International Journal of Sustainable Transportation, 14(10), 733–748.

Mather, P., & Tso, B. (2016). Classification methods for remotely sensed data. CRC press.

Mishra, S., Welch, T. F., & Jha, M. K. (2012). Performance indicators for public transit connectivity in
multi-modal transportation networks. Transportation Research Part A: Policy and Practice, 46(7),
1066–1085. https://doi.org/10.1016/j.tra.2012.04.006

Mishra, S., Welch, T. F., Torrens, P. M., Fu, C., Zhu, H., & Knaap, E. (2015). A tool for measuring and
visualizing connectivity of transit stop, route and transfer center in a multimodal transportation
network. Public Transport, 7(1), 77–99. https://doi.org/10.1007/s12469-014-0091-2

Nabors, D., Schneider, R., Leven, D., Lierberman, K., & Mitchell, C. (2008). Pedestrian Safety Guide for
Transit Agencies.
https://safety.fhwa.dot.gov/ped_bike/ped_transit/ped_transguide/ch4.cfm#:~:text=A.,stop (see
figure below)

NACTTI. (2016). NACTTI Overview Presentation. link: https://www.transportation.gov/policy-

initiatives/traveltourismcommittee/nactti-dot-overview-presentation

Nicholls, S. (2012). The 2012-2017 Michigan tourism strategic plan. Michigan State University and Pure
Michigan.

Outwater, M., Bradley, M., Ferdous, N., Trevino, S., & Lin, H. (2015). Foundational Knowledge to Support
a Long-Distance Passenger Travel Demand Modeling Framework.
https://www.fhwa.dot.gov/policyinformation/analysisframework/docs/long-
distance_model_implementation_report_final.pdf

Pappas, N. (2021). COVID19: Holiday intentions during a pandemic. Tourism Management, 84(June
2020), 104287. https://doi.org/10.1016/j.tourman.2021.104287

Petraglia, L. M., & Weisbrod, G. E. (2004). Integrating Tourism and Recreation Travel with Transportation
Planning and Project Delivery (Vol. 329). Transportation Research Board.

Pincus, M., Perry, C., Dogoe, N. K., Kinder, H., & Lindquist, E. (1999). Sustainable tourism planning and
transportation in Texas. Technical Report, Southwest Region University Transportation Center,
College.

Pooley, J. A., & O’Connor, M. (2000). Environmental Education and Attitudes. Environment and Behavior,
32(5), 711–723. https://doi.org/10.1177/0013916500325007

Richter, N. F., Schubring, S., Hauff, S., Ringle, C. M., & Sarstedt, M. (2020). When predictors of outcomes
are necessary: guidelines for the combined use of PLS-SEM and NCA. Industrial Management and

50
 Data Systems, 120(12), 2243–2267. https://doi.org/10.1108/IMDS-11-2019-0638

Rohr, C., Fox, J., Daly, A., Patruni, B., Patil, S., & Tsang, F. (2013). Modeling long-distance travel in great
britain. Transportation Research Record, 2344(1), 144–151.

Schiffer, R. G. (2012). Long-distance and rural travel transferable parameters for statewide travel
forecasting models (Vol. 735). Transportation Research Board.

SelectUSA. (2020). SelectUSA. https://www.selectusa.gov/travel-tourism-and-hospitality-industry-

united-states#:~:text=Overview,of all U.S. services exports

Sharma, I., Mishra, S., Golias, M. M., Welch, T. F., & Cherry, C. R. (2020). Equity of transit connectivity in
Tennessee cities. Journal of Transport Geography, 86, 102750.
https://doi.org/10.1016/j.jtrangeo.2020.102750

Speakman, C. (2005). Tourism and transport: Future prospects. Tourism and Hospitality Planning &
Development, 2(2), 129–135.

Stansbury, J., Spear, B., Pruvot, A., Fainsilber, O., & Alport, G. (2020). Anticipating the Travel Recovery:
Traveler sentiment survey (Edition 2). https://www.oliverwyman.com/our-
expertise/insights/2020/oct/anticipating-the-travel-recovery.html

Steinmetz, H., Davidov, E., & Schmidt, P. (2011). Three Approaches to Estimate Latent Interaction
Effects: Intention and Perceived Behavioral Control in the Theory of Planned Behavior.
Methodological Innovations Online, 6(1), 95–110. https://doi.org/10.4256/mio.2010.0030

Sultana, Z., Mishra, S., Cherry, C. R., Golias, M. M., & Tabrizizadeh Jeffers, S. (2018). Modeling frequency
of rural demand response transit trips. Transportation Research Part A: Policy and Practice, 118,
494–505. https://doi.org/10.1016/j.tra.2018.10.006

TDTD. (2018). Annual Report: The Soundtrack of America. Made in Tennessee.

https://industry.tnvacation.com/sites/industry/files/component/pod/Annual_2018_Web.pdf

Tennessee Department of Transportation. (2016). TENNESSEE TRANSIT: TODAY & TOMORROW.

https://www.tn.gov/content/dam/tn/tdot/multimodaltransportation/Tennessee_Transit_Today_
and_Tomorrow.pdf

Thapa, D., Mishra, S., Dey, K. C., Golias, M. M., & Ashraf, M. T. (2022). Recreational Travels During the
COVID-19 Pandemic: An Analysis of Antecedents and Necessary Conditions. 101st Annual Meeting
of the Transportation Research Board.

The Soundtrack of America: Made in Tennessee (2022 Vacation Guide). (2022). Tennessee Departmen of
Tourist Development. https://www.tnvacation.com/guide

The Soundtrack of America: Made in Tennessee (FY 2019 Annual Report). (2018).
https://industry.tnvacation.com/sites/industry/files/component/pod/Report_Sept_2019.pdf

Tibshirani, R., Walther, G., & Hastie, T. (2001). Estimating the number of clusters in a data set via the
gap statistic. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 63(2), 411–423.

Virginia Tourism Corportation. (2013). Virginia State Tourism Plan.

51
 https://www.dcr.virginia.gov/recreational-planning/document/vop-app-05-va-tourism-plan-
iface.pdf

Welch, T. F., & Mishra, S. (2013). A measure of equity for public transit connectivity. Journal of Transport
Geography, 33, 29–41. https://doi.org/10.1016/j.jtrangeo.2013.09.007

Yang, D., Darzi, A., Assadabadi, A., Carrion, C., Rossi, T. F., Liu, F., Mishra, S., Avner, J., Radovic, M.,
Mahapatra, S., Baber, C., & Zhang, L. (2019). Integrating Micro-Simulation Models of Short-
Distance and Long-Distance Trips for Statewide Applications. Transportation Research Board 98th
Annual Meeting. https://trid.trb.org/view/1656147

Zenker, S., Braun, E., & Gyimóthy, S. (2021). Too afraid to Travel? Development of a Pandemic (COVID-
19) Anxiety Travel Scale (PATS). Tourism Management, 84(July 2020).
https://doi.org/10.1016/j.tourman.2021.104286

52
Appendices
Appendix A: Survey of state DOTs and tourism agencies
A.1 Summary of the data and survey questions
Search engine data related to tourism travel was collected from the Google trend website for the
year 2019. The tourism travel economic impact data were collected from the US travel Association
website. These collected variables were used for the clustering process. The summary statistics
of the collected dependent and independent variables are shown in Table A-1.
Table A-1 Summary statistics of collected variables

Variable Mean SD Min Max

Total spent by domestic and international
22.99 29.01 2.50 159.30
travelers (billion)
Tourism Travel-supported job (Thousands) 177.120 215.89 17.34 1066.11
Tax Receipts from Tourism Travel (in billion) 3.31 4.32 0.29 22.20
Population Density (Per Square Mile) 404.39 1478.24 1.27 10700.08
Search Engine Data (0-100)*
Airbnb 57.80 14.21 34 100
Air tickets 62.33 13.87 38 100
Airline tickets 54.22 15.00 30 100
Airports 56.78 11.13 39 100
Booking.com 58.98 17.96 28 100
Bus ticket 56.62 14.95 25 100
Car rental 38.06 12.40 21 100
Flights 59.84 13.57 35 100
Food 81.25 8.09 68 100
Fun places 67.90 17.21 29 100
Google flights 28.53 14.69 9 100
Historic site 31.80 13.56 14 100
Hotel booking 42.90 15.87 11 100
Hotels 66.35 9.53 55 100
National parks 34.08 16.69 18 100
Night life 46.37 12.00 14 100
Recreational 36.27 17.04 19 100
Rent a car 24.86 13.50 13 100
Resort 52.76 14.47 32 100
Restaurant 66.45 13.94 42 100
Rest areas 32.11 17.69 12 100
Shuttle bus 31.16 20.43 3 100
Travel agency 59.75 12.81 34 100
Travel 62.51 12.24 45 100
TripAdvisor 42.19 14.23 23 100
* A value of 100 means highest popularity and 0 indicates that there was insufficient data for the keyword

53
A.2 Summary of the Survey

Figure A-1 Map of the survey respondents by state and agency types

A total of 33 complete and 6 partially complete (about 50%-80% complete) responses were
received from DOTs and STOs. A map of the states that participated in the survey is shown in
Figure A-1. The responding states were well distributed all over the U.S. Six states from the West
region, nine states from the Midwest region, two states from the Southwest region, six states
from the Southeast region, and four states from the Northeast region responded to the survey.
Data for the rest of the states were synthesized after the clustering process to create a complete
picture of the current practices and policies and tourism-related transportation infrastructure
and services. A summary of the survey questions that were used to compare different tourism
travel characteristics of the state clusters is presented in Table A-2.
Table A-2 Summary of response obtained from the survey

Question n % Question n %
State DOT respondents State Tourism Stakeholders
Performance Measures
Yes 5 19.23 Yes 1 5.56
No 13 50 No 14 77.78
Not Sure 8 30.77 Not Sure 3 16.67
Collaboration with State Tourism Office Collaboration with State DOTs
Always 1 4.55 Always 1 5.88
Usually 10 45.45 Usually 2 11.76
About Half of the Time 3 13.64 About Half of the Time 1 5.88
Seldom 7 31.82 Seldom 12 70.59
Never 1 4.55 Never 1 5.88
Collaboration with Neighboring State Tourism Stakeholders
Always 0 0 Always 0 0
Usually 4 18.18 Usually 0 0

54
 Question n % Question n %
State DOT respondents State Tourism Stakeholders
About Half of the Time 5 22.73 About Half of the Time 0 0
Seldom 10 45.45 Seldom 9 52.94
Never 3 13.64 Never 8 47.06
Collaboration with private sector stakeholders
Always 4 19.05 Always 0 0
Usually 9 42.86 Usually 1 6.67
About Half of the Time 3 14.29 About Half of the Time 2 13.33
Seldom 5 23.81 Seldom 12 80
Never 0 0 Never 0 0
Collection of tourism trip-related data
Yes 7 29.17 Yes 6 40
No 7 19.17 No 8 53.33
Not Sure 10 41.67 Not Sure 1 6.67
Collection of Emerging Dataset
Yes 10 43.48 Yes 6 40
No 8 34.78 No 7 46.67
Not Sure 5 21.74 Not Sure 2 13.33
Tourism Forecasting
Yes 3 16.67 Yes 5 33.33
No 10 55.56 No 10 66.67
Not Sure 5 27.78 Not Sure 0 0

A.3 k-means Clustering Results

Figure A-2 Multivariate clustering boxplot for identified state clusters

For the clustering process, four variables were used—total spending by the domestic and
international travelers, tourism travel supported jobs, tax receipts from tourism travel (in billion),

55
and the total score of the tourism-related keywords appearing in the search engine. Multivariate
clustering boxplots for the three clusters are shown in Figure A-2 and the three clusters of states
are shown in Figure A-3. Four states in Cluster 1 are California, Florida, New York, and Texas.
These states had a higher impact of tourism travel on their state’s economy. On the other hand,
Cluster 2 (medium tourism impact cluster), had above-average values for tourism travel
supported jobs, spending by domestic and international travelers, tax collected from the tourism
sector, and tourism travel characteristics. There are 15 states in this cluster with a medium impact
of tourism on the state’s economy. Lastly, cluster 3 (low tourism impact cluster) states had an
average performance in terms of tourism travel supported jobs, spending by domestic and
international travelers, and tax collected from the tourism sector.

Figure A-3 Map showing clusters of states based on tourism travel characteristics

A.4 Discussion: Cluster analysis

Transportation planners may include tourism into travel demand forecasting, planning,
prioritizing, and design processes by using tourism-related travel data. The types of tourism
travel data, their intended use, limitations of available data, and requirement of additional data
should be identified to develop a data-driven tourism-related transportation planning. The
specific types of tourism-related transportation system and service data that are frequently
collected by the agencies include tourist O-D trip data, tourist activity data (routes and modes
used, time spent, trip chaining etc.), and trade survey data. However, long-distance tourism trips
have not been an integral part of the regional travel demand model. In the survey, state DOTs
and STOs were asked about their tourism travel data collection practices, their modeling
techniques for tourism travel forecasting, limitations of tourism travel data, and types of
emerging datasets they collect for planning purposes. Aggregated result of the survey of two
clusters is shown in Figure A-4. Responses showed that 40% (6 out of 15 states) of the state DOTs
and STOs collect tourism-related data for tourism-related transportation planning,
transportation infrastructure and service developments, and associated investment decision-
making among states with medium tourism impact (Figure 6). Among states with low tourism
impact, only 28% of the states (9 out of 32) collect tourism travel-related data for planning
purposes. Also, the collection of emerging tourism travel datasets (e.g., Global Positioning System
(GPS) data, cell phone data) was more frequent (56% of 15 states) among states with medium
tourism impacts compared to low tourism impact states (29% of 32 states). In terms of tourism

56
travel forecast, 42% of the state DOTs with medium tourism impact and 28% of the state DOTs
with low tourism impact forecast tourism or tourism travel in their states. Although the practice
of tourism travel forecast is not common among both medium and low tourism impact states, a
higher percentage of states with medium tourism impact used tourism travel forecasting
methods compared to low tourism impact states. The above discussion indicates that the
inclusion of data into tourism transportation planning is still not common among low tourism
impact states.
The emergence of new data sources such as cell phone data, GPS data, and social media data
has changed how data can be used in tourism transportation planning. According to survey
respondents, the most frequently collected emerging dataset by tourism and transportation
departments was cell phone data. State DOTs and tourism agencies also collected Tourist O-D
and activity data for the planning purpose. In terms of tourism data collection issues, most of the
state DOTs expressed that the cost of acquiring data from private companies and lack of data
collection standards are the two key issues associated with tourism data collection and quality.
State tourism agencies also mentioned that cost of data collection is the main issue in acquiring
tourism data. However, they also said that limitations in data collection methods (e.g., excessive
cost) and inconsistencies in data collection (e.g., discrepancies among data sources) are the other
two most frequent issues associated with tourism data quality. In conclusion, tourism data
collection methods and data acquisition costs need to be addressed to ensure consistency in
data use among states and agencies.

Figure A-4 Cluster-wise distribution of tourism data collection and forecasting practices

A.5 Discussion: Deficiencies of tourism transportation services

In the survey, responding agencies identified the existing tourism transportation services, the
limitations of the current transportation services in supporting tourism, and how to overcome
the limitations. Most of respondents reported that their respective states had developed tourism-
focused road signage and maps in recent years. Some of the states have developed park and ride
facilities at transit stops. According to the survey responses, the most common type of tourism-
related transportation infrastructure was highway rest areas and welcome centers. Inadequate
parking facilities at popular tourist destinations are one of the limitations of tourism
transportation services (Figure A-5(a)). The figure also shows that almost 28% of the state DOTs

57
and 22% of the STOs mentioned traffic congestion on major tourism routes as a primary concern.
Apart from that, inadequate tourism signage, tourism-related information, and roadway
maintenance were also mentioned by some agencies. Higher transit access to tourist
destinations was considered critical in improving the sustainability of tourism (Figure A-5(b)).
Almost 18% of the tourism departments mentioned that developing alternate modes such as
high-speed rail, metro, and light rail was necessary to solve the current challenges in developing
tourism-related transportation services (Figure A-5(b)). Furthermore, 18% of the tourism
departments mentioned that higher air, rail, and mass transit access for strategic tourism
locations are needed for sustainable tourism transportation services (Figure A-5(b)). Initiatives
promoting active modes of transportation (e.g., bicycles and walking) are becoming popular all
over the U.S., and these modes are also important for first and last-mile connectivity to tourist
destinations. In the survey, almost 70% of the DOTs mentioned that they consider first and last-
mile transportation services in the project planning process. The survey result showed that the
most common initiatives taken by the DOTs to make these two modes more popular include the
establishment of bike lanes and bike routes, sidewalk, and improvement of the pedestrian
facility, park and ride facilities at transit stops, establishing bike racks in buses, and introduction
of bike-sharing services.

(a) Limitations of current tourism transportation services

(b) Potential solution for solving tourism related transportation problems

Figure A-5 Summary of responses on current deficiencies and its solutions

58
Appendix B: Result from analysis of travel intentions
B.1 Results from SEM analysis
Table B-1 Descriptive statistics of participants’ response to attitudinal questions

Factor
Construct Indicator Mean (SD) Source
loading
I trust the information provided by public health agencies on the pandemic. 3.41 (1.19) 0.56
I am confident that the spread of COVID-19 will be controlled sooner than later. 3.60 (1.07) 0.41
Public trust Own
I believe public health measures such as mandatory use of face coverings and scale
(α=0.78) 3.73 (1.27) 0.60
social distancing are useful in controlling the spread of COVID-19.
I believe that COVID-19 vaccines will protect me from the virus. 3.66 (1.21) 0.65
Most people who are important to me would travel for recreational purposes
3.09 (1.31) 0.68
regardless of the ongoing pandemic.
Subjective
Most people who are important to me think I should travel to tourist destinations Han et al.
norm 2.83 (1.31) 0.76
regardless of the risks from the virus. (2020)
(α=0.88)
People whose opinions I value would think it appropriate for me to travel to tourist
2.93 (1.32) 0.75
destinations despite the pandemic.
Traveling to a destination that is not seriously affected by the COVID-19 outbreak
3.84 (1.27) 0.79
for my next vacation trip is bad.
Attitude
Traveling to a destination that is not seriously affected by the COVID-19 outbreak Han et al.
(α=0.88) 3.88 (1.19) 0.81
for my next vacation trip is unpleasant. (2020)
Traveling to a destination that is not seriously affected by the COVID-19 outbreak
3.89 (1.25) 0.74
for my next vacation trip is foolish.
Whether I travel to a tourist destination that is not seriously affected by the COVID- Han et al.
3.93 (1.12) 0.51
19 outbreak is entirely up to me. (2020)

59
 Factor
Construct Indicator Mean (SD) Source
loading
Perceived I am confident that I can travel to a tourist destination that is not seriously affected
3.71 (1.17) 0.68
behavioral by the COVID-19 outbreak if I want to.
control
I have sufficient resources, time, and opportunities to visit a tourist destination
(α=0.76) 3.65 (1.17) 0.61
that is not seriously affected by the COVID-19 outbreak.

I am comfortable driving to a destination on my vehicle. 3.99 (1.21) 0.69

I am comfortable taking a flight. 3.01 (1.38) 0.71
Travel Stansbury
composure I am comfortable renting a car. 3.32 (1.28) et al., 0.64
(2020)
(α=0.88) I am comfortable using a ride share service (e.g., Uber or Lyft). 2.73 (1.35) 0.77
I am comfortable using public transport (transit bus, train, or the metro). 2.57 (1.35) 0.75
COVID-19 makes me worry a lot about my normal ways of traveling. 3.93 (1.27) 0.75
Travel
It makes me uncomfortable to think about COVID-19 while planning my vacation. 3.32 (1.28) 0.79
anxiety Zenker et
When watching the news about COVID-19, I become nervous or anxious regarding al. (2021)
(α=0.9) 3.34 (1.31) 0.78
travel.
I do not feel safe traveling due to COVID-19. 3.17 (1.35) 0.74
I am concerned about the health of my family members, friends, and relatives
2.43 (1.35) 0.54
during the pandemic
I am concerned about being quarantined away from home during my travels due
Travel 2.87 (1.48) 0.79
to COVID-19 Own
concern
I am concerned about travel restrictions on the way and at the destination after I scale
(α=0.86) 2.82 (1.41) 0.88
have started traveling
I am concerned about travel cost from sudden cancellations (e.g., cancellation of
2.91 (1.43) 0.73
tickets, lodging, etc. without refund)

60
 Factor
Construct Indicator Mean (SD) Source
loading
Compared with the average person, I know the facts about COVID-19. 3.78 (0.96) 0.75
Perceived
Han et al.
knowledge Compared with my friends, I know the facts about COVID-19. 3.76 (0.94) 0.80
(2020)
(α=0.85) Compared with people who travel frequently, I know the facts about COVID-19. 3.67 (0.99) 0.76
I would travel to a destination if the cost of travel is reduced. 3.34 (1.20) 0.81
Perceived I would travel to a destination if the cost of travel insurance is reduced. 3.17 (1.19) 0.73
Own
benefits
I would travel to a destination if the cost of dining, lodging, or services is reduced. 3.40 (1.20) scale 0.82
(α=0.89)
I would travel to a destination if it were not crowded. 3.71 (1.12) 0.61
Whenever I have a chance to travel, I will. 3.32 (1.25) 0.56
Behavioral
I will do my best to improve my ability to travel. 3.34 (1.07) Zenker et 0.76
intention
al. (2021)
(α=0.80) I will keep on gathering travel-related information in the future. 3.18 (1.05) 0.55

Note: The responses were collected on a 5-point Likert scale with 1-strongly disagree to 5-strongly agree.

61
Table B-2 Results from factor analysis

Factor
Construct Item Mean (SD) Source
loading
I trust the information provided by public health agencies on the pandemic. 3.41 (1.19) 0.56
I am confident that the spread of COVID-19 will be controlled sooner than later. 3.60 (1.07) 0.41
Public trust Own
I believe public health measures such as mandatory use of face coverings and scale
(α=0.78) 3.73 (1.27) 0.60
social distancing are useful in controlling the spread of COVID-19.
I believe that COVID-19 vaccines will protect me from the virus. 3.66 (1.21) 0.65
Most people who are important to me would travel for recreational purposes
3.09 (1.31) 0.68
regardless of the ongoing pandemic.
Subjective
Most people who are important to me think I should travel to tourist destinations Han et al.
norm 2.83 (1.31) 0.76
regardless of the risks from the virus. (2020)
(α=0.88)
People whose opinions I value would think it appropriate for me to travel to
2.93 (1.32) 0.75
tourist destinations despite the pandemic.
Traveling to a destination that is not seriously affected by the COVID-19 outbreak
3.84 (1.27) 0.79
for my next vacation trip is bad.
Attitude
Traveling to a destination that is not seriously affected by the COVID-19 outbreak Han et al.
(α=0.88) 3.88 (1.19) 0.81
for my next vacation trip is unpleasant. (2020)
Traveling to a destination that is not seriously affected by the COVID-19 outbreak
3.89 (1.25) 0.74
for my next vacation trip is foolish.
Whether I travel to a tourist destination that is not seriously affected by the
Perceived 3.93 (1.12) 0.51
COVID-19 outbreak is entirely up to me.
behavioral
control I am confident that I can travel to a tourist destination that is not seriously Han et al.
3.71 (1.17) 0.68
affected by the COVID-19 outbreak if I want to. (2020)
(α=0.76)
I have sufficient resources, time, and opportunities to visit a tourist destination
3.65 (1.17) 0.61
that is not seriously affected by the COVID-19 outbreak.

62
 Factor
Construct Item Mean (SD) Source
loading
I am comfortable driving to a destination on my vehicle. 3.99 (1.21) 0.69
I am comfortable taking a flight. 3.01 (1.38) 0.71
Travel Stansbury
composure I am comfortable renting a car. 3.32 (1.28) et al., 0.64
(2020)
(α=0.88) I am comfortable using a ride share service (e.g., Uber or Lyft). 2.73 (1.35) 0.77
I am comfortable using public transport (transit bus, train, or the metro). 2.57 (1.35) 0.75
COVID-19 makes me worry a lot about my normal ways of traveling. 3.93 (1.27) 0.75
Travel
It makes me uncomfortable to think about COVID-19 while planning my vacation. 3.32 (1.28) 0.79
anxiety Zenker et
When watching the news about COVID-19, I become nervous or anxious al. (2021)
(α=0.9) 3.34 (1.31) 0.78
regarding travel.
I do not feel safe traveling due to COVID-19. 3.17 (1.35) 0.74
I am concerned about the health of my family members, friends, and relatives
2.43 (1.35) 0.54
during the pandemic
I am concerned about being quarantined away from home during my travels due
Travel 2.87 (1.48) 0.79
to COVID-19 Own
concern
I am concerned about travel restrictions on the way and at the destination after I scale
(α=0.86) 2.82 (1.41) 0.88
have started traveling
I am concerned about travel cost from sudden cancellations (e.g., cancellation of
2.91 (1.43) 0.73
tickets, lodging, etc. without refund)
Compared with the average person, I know the facts about COVID-19. 3.78 (0.96) 0.75
Perceived
Han et al.
knowledge Compared with my friends, I know the facts about COVID-19. 3.76 (0.94) 0.80
(2020)
(α=0.85) Compared with people who travel frequently, I know the facts about COVID-19. 3.67 (0.99) 0.76
I would travel to a destination if the cost of travel is reduced. 3.34 (1.20) 0.81

63
 Factor
Construct Item Mean (SD) Source
loading
I would travel to a destination if the cost of travel insurance is reduced. 3.17 (1.19) 0.73
Perceived
Own
benefits I would travel to a destination if the cost of dining, lodging, or services is reduced. 3.40 (1.20) 0.82
scale
(α=0.89) I would travel to a destination if it were not crowded. 3.71 (1.12) 0.61
Whenever I have a chance to travel, I will. 3.32 (1.25) 0.56
Behavioral
I will do my best to improve my ability to travel. 3.34 (1.07) Zenker et 0.76
intention
al. (2021)
(α=0.80) I will keep on gathering travel-related information in the future. 3.18 (1.05) 0.55

64
Table B-3 Summary of results for the hypotheses

Hypothese Path
Relationships t-stat Decision
s coefficient

H1a Public trust → Travel attitude 0.54* 12.14 Supported

H1b Public trust → Behavioral intention 0.27* 8.56 Supported

H1c Public trust × Travel concern → Behavioral -0.07* -2.21 Supported

intention

H2a Subjective norm → Travel attitude -0.36* -10.48 Not

supported

H2b Subjective norm → Behavioral intention 0.39# 9.03 Supported

H2c Subjective norm × Travel anxiety → -0.02 -0.60 Not

Behavioral intention supported

H2d Subjective norm → Travel composure 0.45# 11.91 Supported

H3 Travel attitude → Behavioral intention -0.02 -0.57 Not

supported

H4a Perceived behavioral control → Travel 0.26# 7.08 Supported

composure

H4b Perceived behavioral control → Behavioral 0.45# 10.03 Supported

intention

H5 Travel composure → Behavioral intention 0.04 1.08 Not

supported

H6a Perceived knowledge → Perceived benefits 0.24# 7.36 Supported

H6b Perceived knowledge → Behavioral intention 0.07 1.65 Not

supported

H7 Perceived benefits → Behavioral intention 0.54# 12.17 Supported

65
B.2 Interpretation of NCA Bottlenecks and SEM results

(a) Bottlenecks of Public trust for Travel attitude

(b) Bottlenecks of Subjective norm and Perceived behavioral control for Travel composure

(c) Bottlenecks of Subjective norm, Public trust, Perceived behavioral control, and Perceived
benefits for Behavioral intention
Figure B-1 Predictor bottlenecks for the outcome to manifest

66
These bottlenecks, along with the results from SEM can results in one of three scenarios that can
be interpreted in detail as follows.
1. Predictor is significant in SEM and NCA [Scenario-1]: A change in the predictor variable will
change the outcome but a certain level of the predictor variable is necessary for the outcome
to manifest. (Subjective norm → travel composure, perceived behavioral control → travel
composure; public trust → behavioral intention, subjective norm → behavioral intention, perceived
benefits → behavioral intention, perceived behavioral control → behavioral intention)
The public trust-travel attitude bottleneck shows that 90% of intention can be derived
within the first 10% of public trust. However, beyond that, a significant amount of perceived
trust is needed to get the most favorable attitude (about 95%).
2. Predictor is not significant in SEM but significant in NCA [Scenario-2]: A certain level of the
predictor variable is necessary for the outcome to manifest but a change will not affect the
outcome. (Public trust → travel attitude; perceived knowledge → behavioral intention)
Between perceived behavioral control and subjective norm, the former manifests travel
composure more easily. However, beyond about 22% of their levels, the same increase in
travel composure can be expected for the same change in both the predictors. Until the
manifestation of 80% behavioral intention, among its predictors, perceived knowledge can
most easily increase followed by perceived benefits, subjective norm, public trust, and
perceived behavioral control. Note that the effect of perceived knowledge on behavioral
intention reflects the second scenario. Surprisingly, the increase in behavioral intention is
gradual between 5-16% of perceived behavioral control. There is a steady increase in beyond
90% of behavioral intention with a change in all its predictor variables beyond 20% of their
ranges. This suggests that for higher behavioral intention, the first 20% increase in the levels
of the predictors is crucial.
3. Predictor is significant in SEM but not in NCA [Scenario-3]: Change in the predictor variable
will change the outcome and no necessary condition exists for the predictor variable to
manifest the outcome. (Subjective norm → travel attitude; perceived knowledge → perceived
benefits)
For this scenario, we find that no minimum levels of travel anxiety and concerns are
necessary to manifest travel intention, but they are sufficient to manifest the outcome. A
similar effect of subjective norm and perceived knowledge on travel attitude and perceived
benefits respectively is also observed.

67
Appendix C: Most popular destinations
Table C-1 List of popular destinations

Region Destination Nearest UA Nearest commercial airport

West Bass Pro Shops at the Pyramid Memphis Memphis Int'l Airport
West Beale Street Historic District Memphis Memphis Int'l Airport
West Discovery Park of America Union City Memphis Int'l Airport
West Elvis Presley's Graceland Memphis Memphis Int'l Airport
West Memphis Music Hall of Fame Memphis Downtown Memphis Int'l Airport
West Memphis Rock ’n’ Soul Museum Memphis Downtown Memphis Int'l Airport
West Memphis Zoo Memphis Memphis Int'l Airport
MoSH – Memphis Museum of Science &
West Memphis Memphis Int'l Airport
History
West National Civil Rights Museum Memphis Memphis Int'l Airport
West Shelby Farms Park Memphis Memphis Int'l Airport
West Stax Museum of American Soul Music Memphis Memphis Int'l Airport
West Tennessee Safari Park Alamo Memphis Int'l Airport
Andrew Jackson's Hermitage: Home of the
Middle Nashville Nashville Int'l Airport
People's President
Middle Arrington Vineyards Arrington Nashville Int'l Airport
Middle Country Music Hall of Fame and Museum Nashville Nashville Int'l Airport
Middle Cumberland Caverns McMinnville Chattanooga Airport
Middle Downtown Nashville Nashville Nashville Int'l Airport
Middle Falls Creek Falls Spencer Chattanooga Airport
Middle Jack Daniel Visitor Center Lynchburg Nashville Int'l Airport
Middle Nashville Zoo Nashville Nashville Int'l Airport
Middle Stones River National Battlefield Murfreesboro Nashville Int'l Airport
East American Museum of Science and Energy Oak Ridge McGhee Tyson Airport
East Anakeesta Theme Park Gatlinburg McGhee Tyson Airport
East Bays Mountain Park and Planetarium Kingsport Tri-Cities Airport
East Bristol Caverns Bristol Tri-Cities Airport
East Bristol Motor Speedway Bristol Tri-Cities Airport
East Cherokee Lake Multiple counties McGhee Tyson Airport
East Dollywood Pigeon Forge McGhee Tyson Airport
East Gatlinburg SkyLift Park Gatlinburg McGhee Tyson Airport
East Gray Fossil Site and Museum Gray Tri-Cities Airport
East Historic Downtown Morristown Morristown McGhee Tyson Airport
East Hunter Museum of American Art Chattanooga Chattanooga Airport
East Jonesborough Historic District Jonesborough Tri-Cities Airport
East Manhattan Project National Park Oak Ridge McGhee Tyson Airport
East Museum Center at 5ive Points Cleveland Chattanooga Airport
East Ober Gatlinburg Aerial Tramway Gatlinburg McGhee Tyson Airport

68
East Rock City Gardens and Lookout Mountain Chattanooga Chattanooga Airport
East Ruby Falls Chattanooga Chattanooga Airport
East Sevierville Downtown Sevierville McGhee Tyson Airport
East Tennessee Aquarium Chattanooga Chattanooga Airport
East Tennessee Theatre Knoxville McGhee Tyson Airport
East The Caverns Pelham Chattanooga Airport
East The Lost Sea Adventure Sweetwater McGhee Tyson Airport
East The Old Mill Square Pigeon Forge McGhee Tyson Airport
East Titanic Museum Attraction Pigeon Forge McGhee Tyson Airport
East Tsali Notch Vineyard Madisonville McGhee Tyson Airport
East Tuckaleechee Caverns Townsend McGhee Tyson Airport
East Wetlands Water Park Jonesborough Tri-Cities Airport
East World’s Fair Park Knoxville McGhee Tyson Airport
East Zoo Knoxville Knoxville McGhee Tyson Airport

69
Appendix D: Output from INRIX Trip Analytics
D.1 Memphis

Figure D-1 Origin and destinations of trips made through I-55 E

Figure D-2 Origin and destinations of trips made through I-40 E

70
D.2 Nashville

Figure D-3 Origin and destinations of trips made through I-40 W

Figure D-4 Origin and destinations of trips made through I-40 E

71
D.3 Chattanooga

Figure D-5 Origin and destinations of trips made through I-24 E

Figure D-6 Origin and destinations of trips made through I-75 S

72
D.4 Knoxville

Figure D-7 Origin and destinations of trips made through I-40 W

Figure D-8 Origin and destinations of trips made through I-40 E

73
 Figure D-9 Origin and destinations of trips made through I-75 S

D.5 Gatlinburg

Figure D-10 Origin and destinations of trips made through US-321 S to Gatlinburg

74
Appendix E: Tourism Agencies in Tennessee
Table E-1 List of survey recipients who were invited to undertake the survey of agencies

West Tennessee
Benton County-Camden Chamber of Commerce Humboldt Chamber of Commerce
Lauderdale Chamber/Economic and Community
Brownsville/Haywood County Chamber of Commerce
Development
Carroll County Chamber of Commerce McNairy County Chamber of Commerce
City of Parsons Milan Chamber of Commerce
Covington-Tipton County Chamber of Commerce Millington Area Chamber of Commerce
Crockett County Chamber of Commerce Northwest Tennessee Tourism
Decatur County Chamber of Commerce Obion County Chamber of Commerce
Dyersburg/Dyer County Chamber of Commerce Paris/Henry County Chamber of Commerce
Greater Gibson County Area Chamber of Commerce Reelfoot Lake Tourism Council
Hardeman County Chamber of Commerce South Tipton County Chamber of Commerce
Hardin County Convention & Visitors Bureau Visit Brownsville TN
Henderson County Chamber of Commerce Visit Jackson TN
Historic Downtown Martin Weakley County Chamber of Commerce
Middle Tennessee
McMinnville-Warren County Chamber of
Bell Buckle Chamber of Commerce
Commerce
Cannon County Chamber of Commerce Metro Lynchburg/Moore Chamber of Commerce
City of Columbia Mount Pleasant Community Development Corp.
Clay County Partnership Chamber of Commerce Nashville Convention & Visitors Corp
Crossville-Cumberland County Chamber of
Robertson County Chamber of Commerce
Commerce
Dickson County Chamber of Commerce Shelbyville-Bedford County Chamber of Commerce
Fayetteville-Lincoln County Chamber of Commerce &
Smith County Chamber of Commerce
Tourism Bureau
Franklin County Chamber of Commerce Smithville-Dekalb Country Chamber of Commerce
Hartsville – Trousdale Chamber of Commerce South Central Tennessee Tourism
Hickman County Chamber of Commerce Sparta-White County Chamber of Commerce
Hohenwald/Lewis County Chamber of Commerce Stewart County Chamber of Commerce
Houston County Area Chamber of Commerce Sumner County Visitor Center
Jackson County Chamber of Commerce Tennessee’s Backroads Heritage, Inc.
Jamestown/Fentress County Chamber of Commerce Tullahoma Area Chamber of Commerce
Lawrence County Chamber of Commerce Van Buren County Chamber of Commerce
Livingston-Overton County Chamber of Commerce Visit Goodlettsville
Macon County Chamber of Commerce Visit Pulaski
Manchester Area Chamber of Commerce Wartrace Chamber of Commerce
Marshall County Chamber of Commerce Wayne County Chamber of Commerce
East Tennessee
Campbell County Chamber of Commerce & Tourism Monroe County Department of Tourism
Chattanooga Visitors Center Morgan County Tourism Alliance
Cheatham County Chamber of Commerce Morristown Area Chamber of Commerce

75
Claiborne Economic Partnership Northeast Tennessee Tourism Association
Cleveland/Bradley County Chamber of Commerce
Pigeon Forge Department of Tourism
and Tourism Development
Coker Creek Welcome Center Pikeville-Bledsoe County Chamber of Commerce
Elizabethton/Carter County Visitor Center Polk County Chamber of Commerce
Rogersville/Hawkins County Chamber of
Farragut Community Center
Commerce
Follow The Quilt Trail (Appalachian Red Council) Scott County Chamber of Commerce
Gatlinburg Convention and Visitors Bureau Sevierville Convention & Visitors Bureau
Grainger County Chamber of Commerce Sneedville-Hancock County
Greene County Partnership South Cumberland Chamber of Commerce
Sullivan County Department of Archives and
Historic Jonesborough Visitors Center
Tourism
Jellico Tourism Office Tennessee Association of RV Parks & Campgrounds
Johnson City CVB Town of Tellico Plains
Knoxville Convention and Visitors Bureau – Visit
Townsend Visitor Center
Knoxville
Marion County Chamber of Commerce Unicoi County Chamber of Commerce
Maynardville/Union County Visit Jefferson County
Meigs County-Decatur Chamber of Commerce Visit Kingsport

76