Long-Distance and Rural Travel Transferable Parameters for Statewide Travel Forecasting Models (2012)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

1 The modeling of long-distance trips in statewide models differs from that of urban and regional models that focus on differentiating home-based from nonhome-based trips. Long-distance trips are more likely to be divided into categories by frequency of travel or by purpose such as recreational/tourist versus business-oriented trips. Such considerations are more likely to be indicative of long-distance variations by trip length, mode choice, and other aspects of travel. Most statewide travel demand forecasting models are built upon practices originally developed for urbanized area modeling. In the context of statewide forecasting, rural trip- making and long-distance intercity travel constitute important market segments, much more so than in urban models. Information describing these markets, and how these mar- kets vary from state to state, is somewhat sparse, and many states do not have the resources to initiate original data collection to develop a set of model parameters. Yet these same states have a pressing need to have confidence in reasonable data for rural and long-distance travel. Furthermore, for the states where local data collection has occurred, there is little basis to assess how reasonable their findings are compared with findings from other states. Statewide models in smaller and more urbanized states do not typically distinguish between urban and rural travel. However, it is generally accepted that rural area trip patterns differ from intra-urban travel, and so most statewide models should attempt to distinguish between urban and rural trip-makers. While trip rates are readily available for transfer- ability from urban and regional models, there are relatively few rural trip rates available to transfer for use in statewide travel demand models. Statewide models with trip generation rates derived from statewide surveys or the National Household Travel Survey (NHTS) Add- On samples stratified into urban and rural respondents are worth evaluating as a potential source of transferable parameters (http://nhts.ornl.gov/). Documentation related to the validation of several statewide models is available; however, no comprehensive research assessing recent national datasets had previously been performed, and there had been no analysis of the transferability of parameters among statewide models prior to NCHRP Project 08-84. For urban models, there are several sources of validation and reasonableness checking such as NCHRP Report 716: Travel Demand Forecasting: Parameters and Techniques (Cambridge Systematics, Inc. et al., 2012) and the FHWA Travel Model Valida- tion and Reasonableness Checking Manual, Second Edition (Cambridge Systematics, Inc., 2010c). These documents provide a set of excellent resources to evaluate urban models but do not provide any guidance on how nonurban (superregional, intercity, and statewide) parameters should be used, reasonable ranges of those parameters, and how those parameters should be modified for rural areas. S u m m a r y Long-Distance and Rural Travel Transferable Parameters for Statewide Travel Forecasting Models

2 Long-Distance and rural Travel Transferable Parameters for Statewide Travel Forecasting models The American Travel Survey (ATS) (http://www.transtats.bts.gov/Tables.asp?DB_ID= 505&DB_Name=American+Travel+Survey+(ATS)+1995&DB_Short_Name=ATS) was originally designed to obtain information on long-distance travelers; however, this survey was later discontinued and the more recent NHTS is not structured for a targeted sample size of long-distance trips. The 2009 NHTS Add-On programs in several states do provide usable data related to rural trip-making and, to a lesser extent, long-distance travel. The objective of this research has been to develop and document transferable parameters for long-distance and rural trip-making for statewide models. It was envisioned that this report would act as a supplement to the NCHRP “quick response” guidance on model parameters and highlight reasonable sets of parameter ranges for rural and long-distance trip-making. It will be widely used by state departments of transportation (DOTs), metropolitan planning organizations (MPOs), and consultants developing multistate and national travel forecast- ing models, statewide and intercity passenger models, and large regional models, especially those covering areas of low-density rural development patterns and undeveloped lands. Differences in Rural and Long-Distance Trip-Making In the context of statewide forecasting, rural trip-making and long-distance intercity travel constitute important market segments. Information describing these markets and how they vary from state to state has been sparse, and many states do not have the resources to initiate original data collection to develop a set of model parameters. Yet these same states have a pressing need for confidence in reasonable transportation planning results for rural and long-distance travel. Furthermore, for the states where local data are available, there has been little basis to assess how comparable their assumptions are with those from other states. This topic is addressed in this study by identifying differences in rural and urban travel, in various states, from existing surveys. This includes preliminary analyses of 1995 ATS, 2001 NHTS, 2009 NHTS, and select statewide, superregional, and tourist survey data to (1) see how differences in rural and long-distance trip-making occur in different geographic regions and (2) identify any explanatory variables that could be used to adjust average values and reflect conditions in a particular state. The most recent NHTS contains over 20 separate add-on partners, some representing full states and some MPO planning areas (which may include rural areas within the MPO boundary). It was important in conducting analysis that rural and long-distance data on transferable parameters be compared against urban short-distance data and typical model parameters. For example, according to the 2009 NHTS, short trips account for the vast majority of per- sonal trips in the United States—three-quarters of vehicle trips are less than 10 miles in length. However, these trips account for less than one-third (28.9 percent) of all vehicle miles traveled (VMT). Trips of over 100 miles account for less than 1 percent of all vehicle trips, but 15.5 percent of all household-based vehicle miles. With the potential impact on VMT, travel demand forecasts depend on knowing more about the current amount and nature of long-distance and rural travel in the United States. Statewide Model Statistics This Guidebook explores the characteristics of statewide models further to identify sources that could be used in comparing, developing, and recommending trip production and attrac- tion rates, friction factors, mode choice coefficients, and peak-to-daily/time-of-day factors,

Summary 3 among other model parameters, for estimating rural and long-distance travel. The final report for NCHRP Statewide Model Validation Study (Cambridge Systematics, Inc., 2010d) included a series of tables describing model parameters and benchmark statistics from statewide models, including information on long-distance and rural trip purposes, where these were separated from typical urban model purposes. Some of the information in this Guidebook was derived either from recent work on the NCHRP Model Validation Report or prior work on national model research for FHWA. Establishment of trip purposes used in statewide models is important because this largely determines the stratifications used in subsequent model statistics (i.e., these are reported by trip purpose). Some trip purposes in statewide models are duplicative, using different names but meaning the same thing. This has been fleshed out through discussions with state DOT contacts and their consultants. Some models differentiate short-distance from long-distance trip purposes while others do not. Therefore, although this task focuses primarily on long- distance trip purposes, it also includes home-based and nonhome-based trip purposes for relevant models that do not include separate long-distance purposes. Model statistics compiled by trip purpose included aggregate trip rates, percent trips by purpose, average trip length/duration (in time and distance), vehicle occupancy rates, and mode splits. General Guidance on Transferability of Model Parameters This Guidebook provides general guidance on when and when not to transfer model parameters. General analysis by the research team has shown that population density is a potential indicator of model transferability. This is particularly the case with mode choice for long-distance travel, because private passenger vehicles predominate in long-distance travel in smaller-sized urbanized areas and rural areas while long-distance travel is more common on alternate modes in large metropolitan areas. Clearly, there is a relationship between population density and available transportation modes that also explains the mode choice issue. Density, in the sense of urban versus rural travel, shows up consistently in most analysis documented in this study with, for example, urban areas having higher trip rates but rural areas having higher average trip lengths. With respect to analysis of median income impacts on trip-making, it stands to reason that lower income households would make fewer long-distance trips than higher income households. Likewise, household decisions on transportation modes for long-distance travel should include an income component. Analysis completed for this report deepens under- standing of the relationship between income and rural trip-making. Key employment types and industries can impact rural trip-making. A good example of this is tourism and lodging, which has a large need for low-income workers who cannot afford to live in proximity to resort developments. Such areas are also magnets for long- distance travel because visitors to resorts usually reside outside of the region. The source of the model parameter is a key decision point in parameter transferability because there is a wide variety of sources considered in establishing such settings, includ- ing state DOT surveys (both household and intercept), surveys from adjacent or similar states, national surveys, MPO surveys, NCHRP Report 716 and other model guidance docu- ments, as well as other statewide models. Furthermore, smaller states (e.g., Rhode Island) might have more in common with urban and regional models than statewide models, with a smaller percentage of long-distance trip activity and dominated by urbanized land.

4 Long-Distance and rural Travel Transferable Parameters for Statewide Travel Forecasting models Clearly, long-distance model parameters should be derived from surveys with a statisti- cally valid sample of such trip-makers. Rural model parameters require a survey with both urban and rural resident components in order to ensure that the resulting rates are in fact the result of differences in residential and/or work location and not just due to error in survey execution or design. Although reported statistics from statewide models and documentation of general guidance are useful to provide context, such comparisons are no substitute for analysis of travel survey data. The limitations of the data sources must also be considered, especially as these relate to geographic limitations or trip definition. The minimum amount of data needed for the geography intended (national, regional, state, or metro area) must be assessed for each of the parameters. It is important for readers of this document to understand the limitations of the datasets used during this study when transferring parameters provided in this report. Some potentially transferable parameters important to properly estimating long-distance and rural travel patterns and comparative benchmark statistics in statewide models are described below. Transferable parameters recommended for estimation include the following: • Daily (weekday and weekend) rural trip rates per household by household characteristics (e.g., number of workers by industry) and by trip purpose; • Monthly or annual long-distance trip rates per household by household characteristics (e.g., median income) and by type of trip (trip purpose); • Friction factors, gamma functions, or utilities for rural travel by trip purpose; • Friction factors, gamma functions, or utilities for long-distance travel by trip purpose; • Auto occupancy rates for rural vehicle trips by trip purpose; and • Party size for long-distance trips by trip purpose. In addition to the transferable parameters recommended above, and the dynamics noted earlier in this Summary, reasonableness values are documented in this research for the following: • Percent of rural trips by purpose; • Percent of long-distance trips by trip purpose; • Average (mean) person trip length of rural trips by mode and purpose; • Average person trip length of long-distance trips by mode and trip purpose; and • Percent of long-distance and rural trips by mode (private vehicle, rail/bus, air, other) and travel distance. Consideration of Other Trip Characteristics Beyond demographic and mobility characteristics are considerations as to what should constitute a statewide model trip. Even this varies among different statewide models, with a few that essentially do not include intra-urban trips (e.g., Louisiana). Trips could be defined by person, household, or even vehicle in some cases. Sometimes, it might make sense to include intermediate stops as trip ends; however, this would seemingly go against the con- cept of long-distance trips. In fact, what travelers typically think of as a “(round) trip” is what transportation planners consider a “tour.” A few statewide models (e.g., Ohio, Oregon, and New Hampshire) use the concept of tours instead of trips. For rural travel analysis, average weekday conditions would likely be preferable. Simi- lar to regional models, while it might be best to exclude travel on weekends and holidays, such limitations would result in sample size problems. NHTS staff indicates that approxi- mately 25–30 percent of surveys were conducted on weekend travel; however, weekend travel

Summary 5 includes Friday after 6:00 p.m. (teleconference with Adella Santos, FHWA; Vidya Mysore and Frank Tabatabaee, Florida DOT; and Rob Schiffer, Cambridge Systematics, Inc. on August 10, 2011), a timeframe that is similar to other weekday evening peak periods in many regions. In states with a singular, well-defined peak season, consideration could be given to surveys that constitute peak season average weekday traffic instead of annual average daily traffic (AADT), although such a timeframe of analysis would not be recommended for a study on national transferability such as this. Conversely, since long-distance travel is not an everyday occurrence in most households, monthly or annual statistics must at least be considered in survey analyses. Also, it is essential to include weekends and holidays in any survey analysis of long-distance travel because these time periods reflect where the greatest amount of such travel takes place. Consideration was given to developing time-of-day factors both for rural and long-distance trips during this study; however, with the infrequency of long-distance trips, use of trip rates by time of day might be overkill. In addition to temporal considerations, there are other aspects to be considered in defining a trip for the purposes of research and analysis. The first of these is consideration of person trip versus vehicle trip analysis. Since the majority of statewide models deal with person trips and starting with vehicle trips almost precludes a mode choice process, the recommendation is to conduct survey analysis by person trip rather than vehicle trip. Long-distance trip-making was considered at two to three different thresholds to determine how parameters differ at each threshold. Another consideration was how to deal with intermediate stops and whether these should constitute a trip end or not. Clearly, long-distance trips require stops for gas, food, and/ or lodging. In the context of a regional model, these intermediate stops for shopping, etc., would each represent a unique nonhome-based trip. In the context of most statewide, multi- state, or national modeling, however, these intermediate stops are not of tremendous impor- tance in defining and simulating a trip. On the other hand, it is probably worth considering an intermediate stop at the end of the day for lodging as the end of a daily trip, assuming the analysis is daily rather than monthly or annually. The location of intermediate stops, relative to congestion on Interstate highways or crossroads, could result in greater interest about intermediate travel patterns. The number and duration of stops was also addressed in this research. The topic of intermediate stops also leads directly to consideration of tours versus trips. The previous lodging example might be better addressed as a stop during a tour, rather than the endpoint of the trip; however, the majority of statewide models are still trip-based. Those statewide models that are tour-based were developed using statewide travel surveys and, as a result, will not likely have as much use for transferable parameters. However, the preparation of tour-based parameters is beyond the scope of this project. Process for Using Data Sources to Develop Parameters A key analytical step in this research was to compare trip generation statistics for households in “rural” areas, using various rural definitions to assess if there are differences in trip-making. Such analyses also needed to account for urban trip characteristics to identify differences. Ana- lytical comparisons necessitate a typology of rural activity, such as defining rural households nearer to urban centers versus those farther away. Another unique characteristic of some rural areas, yet more difficult to quantify, is proximity to major recreational areas.

6 Long-Distance and rural Travel Transferable Parameters for Statewide Travel Forecasting models Demographic profiles are also helpful, defining household characteristics such as size, life cycle, income, and/or number of workers by worker status and occupation. An interesting topic, should such data be available, would be to include comparisons of Internet availability and use this information to impute if rural households are more or less likely to shop online, based on a lack of options to shop locally. The propensity of rural residents to link trips is another unique factor as those with long daily commutes are likely to do their shopping and other personal business prior to leaving the urban area at the end of the work day. The research team identified opportunities to leverage some of the analysis already con- ducted for urban transferable parameters (NCHRP Report 716) and much of the thinking on new typologies, especially sociodemographic, were helpful for NCHRP Project 08-84 efforts. The Version 2 NHTS 2009 has a number of enhancements that were helpful for analyti- cal purposes, including estimates based on the 2008 American Community Survey (ACS) and land-use descriptors for the household and workplace locations from Claritas/Neilson. Selected characteristics of urbanized areas from the annual “Highway Statistics” publication of the FHWA Office of Highway Policy Information (OHPI) can also potentially assist in defining characteristics that separate rural from urban settings. Long-Distance Travel Parameters and Benchmarks One key to implementing the analytical plan and developing transferable parameters was to obtain access to all datasets from the American Travel Survey (ATS) and iden- tify trip purposes, average trip lengths, vehicle occupancies, and other statistics typified by long-distance travelers. The 1995 ATS datasets are dated; however, these data are the only long-distance data that provide statistically sound estimates of long-distance travel in and between the states. Although the 2001 National Household Travel Survey (NHTS) had a long-distance com- ponent, this survey did not have sufficient samples to calculate estimates of long-distance travel for most states (New York and Wisconsin were exceptions to this, because of the large add-on in the former and stratified sampling of the latter, although neither add-on was included in the official 2001 NHTS long-distance file). The approach to using NHTS 2001 data was based on discussions with FHWA NHTS support staff, both past and present, as well as members of the research team with extensive experience using different versions of the NHTS. All of these discussions pointed to concerns over the use of NHTS 2001 for long- distance trips and at least some of these concerns are documented elsewhere in this study. All of the NHTS 2001 long-distance data were made available for use by the consultant team as well, including state add-on samples. These two long-distance datasets can be used together, yet separately, since the 2001 ques- tionnaire relied heavily on the 1995 ATS as a template. Definitional categories for mode and purpose are comparable. The study team also obtained readily available state DOT survey data and documentation from statewide household travel surveys for Michigan and Ohio. Additionally, the study team coordinated with Canadian officials to identify available long-distance travel parameters readily available from their recent household travel surveys. Finally, recent travel surveys using Global Positioning Systems (GPS) were mined for param- eters on long-distance travel as well as rural parameters. Transferable rural travel parameters largely focused on the 2009 NHTS and its State Add-On surveys. Analysis of variance (ANOVA) and other statistical tests were run on 2009 NHTS data in an attempt to identify which available attributes best explain differences in

Summary 7 rural trip-making and whether certain parameters should be stratified for different condi- tions such as urban clusters and proximity to urbanized area boundaries. Existing statewide models also played a significant role in this analytical plan, in terms of quantifying reasonableness ranges against which to compare resulting ATS/NHTS survey-based model parameters. Also, documented model parameters were identified for potential transferability to other statewide models, based on the characteristics of the state where the data were collected versus the state to which a parameter might be proposed for transferability. Interregional, or intercity, travel components are included in some statewide models to capture both intrastate and interstate trips. The core model design feature is the recognition that interregional travel is very different from urban area travel, where different set(s) of explanatory variables are involved or different sensitivi- ties to levels of service. A set of typical long-distance and rural trip purposes was established from this analysis so that model parameters could be stratified by such categories and reasonableness bench- marks could be established for percent trips by purpose. Mean trip length statistics, both in miles and minutes, also were estimated from the survey databases for use as benchmarks in future statewide model validation efforts; however, the survey analysis for this study did not include the calculation of state-by-state trip lengths. As discussed previously, statewide models and travel surveys have used a range of thresh- olds to define long-distance trip-making. Most sources cited in this study used either 50, 75, or 100 miles as the minimum threshold for trips to be considered “long-distance.” In an effort to maximize the number of long-distance trip samples, this Guidebook looks at model parameters at three different long-distance trip thresholds: 50–100 miles, 100– 300 miles, and more than 300 miles. Separating 50–100 mile trips from 100–300 mile trips allows for differentiation of long-distance trips by the two most common thresholds, beginning and ending at 100 miles. The rationale for using 300 miles as another cutoff point is that preliminary data analysis indicated a mode shift from personal auto to air travel at this distance. Rural Travel Parameters and Benchmarks Identification of rural travel parameters took a different focus than long-distance travel parameters. First, rural trip-making data are well represented in the recent 2009 NHTS. Therefore, the study team was able to focus primarily on this one survey database, unlike the multiple and considerably older survey databases used to identify long-distance travel parameters. Second, the points of reference are quite different for rural trips. Long-distance travel characteristics were generally summarized by different trip length categories, whereas rural travel parameters required establishing typologies for classification and comparison against comparable statistics on travel in urbanized areas. Finally, the tempo- ral issues for rural travel are not as complex as those for long-distance trips. For example, the database does not deal with international travel or multiple stops and the greater share of travel is on the weekdays, with a much smaller share of weekend travel than with long- distance trips. The first step in the assessment of rural travel parameters was the identification of rural typologies and an exploration of how these different typologies can be used to describe the trip-making of rural households. This also includes the need to define what is and is not rural travel, and how typical rural travel behavior differs from that in more urban settings. These

8 Long-Distance and rural Travel Transferable Parameters for Statewide Travel Forecasting models efforts started with a focus on attributes contained within the NHTS 2009 “DOT version” of the database, including the Claritas attributes described earlier. The following attributes from the 2009 NHTS DOT version were used to identify poten- tial rural typologies: • URBAN—Identifies whether or not the home address is located in an urban area, typi- cally defined as a concentrated area with a population of 50,000 or greater. • URBRUR—Identifies whether or not the home address is located in a rural area. • URBANSIZE—Population size of the urban area in which the home address is located. • HBHUR—Urban/Rural Indicator, appended to the NHTS by Claritas (http://nhts.ornl. gov/2009/pub/UsersGuideClaritas.pdf). This classification reflects the population density of a grid square into which the household’s block falls. • HBRESDN—The number of housing units per square mile by block group. • HBPOPDN—The population per square mile by block group. Additionally, the rural typologies recommended as part of NCHRP Project 25-36 also were considered in this effort. The four typologies recommended by NCHRP Project 25-36: Impacts of Land Use Strategies on Travel Behavior in Small Communities and Rural Areas (http://apps.trb.org/cmsfeed/TRBNetProjectDisplay.asp?ProjectID=2987) were as follows, along with the study definitions of each, as quantified by “commuting zones” developed by the USDA’s Economic Research Service: • Population Density—Computed as number of people divided by unit area of developed or developable land. • Road Density—Calculated as road length in miles per square mile of developed or devel- opable land. • Land Use Mixture—A proxy of land-use mixture measuring how residents, jobs, and other activities are distributed in relation to each other. • Variation in Population Density—Variation in population density distinguished where most residents are located in a relatively small set of concentrated areas at relatively high densities from locations where residents are spread more evenly. This project did not pursue full consideration of commuting zones, which are defined in NCHRP Project 25-36 as “multicounty regions that convey the typical pattern of commuting trips in a spatially defined labor market: a much higher proportion of commuting trips have origins and destinations that are both inside the zone than those trips for which one end is outside” (Department of City and Regional Planning Center for Urban and Regional Studies University of North Carolina at Chapel Hill, 2011). In place of data on commuting zones, the analysis presented here uses readily available data to simulate some of these typologies. Population Density already was an attribute included in the 2009 NHTS dataset so it was easily addressed. Road Density was calculated using the 2005 National Highway Planning Network and geographic information systems (GIS) tools, based on a simple formula of Road Length/Census Tract Area. The resulting Road Density was a continuous variable, so a regression analysis was conducted and then the variable was recoded as a categorical variable. There was no practical way to simulate land-use mixture or the variation in population density using the data readily available for this project. One additional typology analyzed was “urban proximity” because the NCHRP Project 08-84 research team thought that the proximity to urban areas could impact the number and purpose of trips. Latitude/longitude address information was not stored for each household in the 2009 NHTS DOT database, which is necessary for accurate depiction in GIS. The database did have Census Tract and Block Group information, and this information was

Summary 9 appended to an NHTS Census Tract/Block Group shapefile. Once the 2009 NHTS DOT database was joined to the NHTS CT/BG shapefile by a Census Tract/Block Group ID number, the households were spatially referenced to the Block Group. In cases where a Block Group was in proximity to multiple urban areas, distance to the closest urban area was applied. Unfortunately, Proximity to Urban Area did not show any clear trip rate trend, so the remaining analysis focused on the other measures. Comparisons and Conclusions The subject area of this study was wide ranging and although there are a multitude of ways to analyze the topic of rural and long-distance travel, there were limitations to the resources available for this study. Study findings were largely focused on the 1995 ATS for long-distance trips and 2009 NHTS for rural trip-making parameters. This section presents a few compari- sons among the different surveys and travel parameters analyzed during this study. Originally, it was intended to look at the impacts on long-distance trip rates of proximity to areas with substantial tourist activity. Unfortunately, the ATS and NHTS databases do not include information on proximity of residence to “tourist areas.” Manual geocoding of known tourist sites was considered to analyze trip rates based on proximity to tourist areas; how- ever, there were concerns about arbitrarily coming up with a list of tourist sites and possibly excluding some regionally important tourist sites. National parks are an obvious attraction and easily mapped as are the locations of well-known nonurbanized tourist areas such as Branson, Gatlinburg, the Outer Banks, etc. However, should every amusement park in the United States be included in such an analysis? Also, the “production” of long-distance trips would not likely be influenced so much by proximity to tourist areas, as would be trip attractions. This topic might be worthy of another research effort to provide a more objective assessment of differing types and sizes of rural tourist destinations. Rural accessibility/proximity to employment was also considered; however, the NHTS 2009 database had limited data on work location. Instead, proximity to urbanized areas was tracked in its relationship to rural trip production. Trip rates for long-distance and rural trips were provided from several different sources. Table S.1 presents overall long-distance person trip rates per household from the 1995 ATS, 2001 NHTS, and recent GPS household surveys. Annual rates from the ATS and NHTS were divided by 365 days and rounded to two decimal places to derive a daily rate for comparison against a recent GPS survey database. As indicated, all survey databases result in daily person long-distance trip rates of 0.03–0.04 per household. Likewise, total daily person rural trip rates were reported from several sources, includ- ing 2009 NHTS, Michigan DOT, and the GPS household survey database. As depicted in Table S.2, person trip rates per rural household appear to be in a relatively similar range for different stratifications of 2009 NHTS, while different subareas and years from the Michigan and Ohio surveys tend to show lower household trip rates by comparison. Rural trip rates Survey Data Source Daily Person Trips per Householda 1995 ATS 0.03 2001 NHTS 0.03 Recent GPS Household Surveys 0.04 (average of four surveys) a Annual trip rates were divided by 365 for 1995 ATS and 2001 NHTS, rounded to hundredths. Table S.1. Comparative long-distance household trip rates.

10 Long-Distance and rural Travel Transferable Parameters for Statewide Travel Forecasting models from the GPS household survey database fall within a range similar to the NHTS, Michigan, and Ohio household person trip rates. The impact of the recent economic recession on 2009 NHTS trip rates is unknown at this time and beyond the scope of this research effort. A brief summary of findings and key conclusions based on survey analyses is presented below, with long-distance trips discussed first, followed by rural trips. • Long-distance trip rates are generally consistent when compared among several data sources and years. The percentage of long-distance trips by purpose/type appears consis- tent between the 1995 ATS and 2001 NHTS long-distance component: – Business—28.38 percent for NHTS 2001 versus 22.25 percent for ATS; – Pleasure—54.84 percent for NHTS 2001 versus 58.97 percent for ATS; and – Personal Business—16.78 percent for NHTS 2001 versus 18.78 percent for ATS. • Long-distance trips are generally longest for business purposes (954 miles) and shortest for personal business (704 miles), with pleasure trip lengths in the middle of the others (828 miles). • Auto occupancy rates are considerably higher for long-distance trips (3.10) than urban or rural travel (1.54), lowest for long-distance business trips (2.11), and higher for other long-distance types (3.33–3.46). • Private automobile is the dominant transportation mode for long-distance travel (82 per- cent); however, trip length and purpose/type figure prominently in shifting to air travel. • Rural trip rates vary somewhat among different data sources; household trip rates from Michigan and Ohio surveys are generally lower than those from the 2009 NHTS, as depicted earlier in Table S.2. • Rural trip rates (9.69) appear lower than suburban area trip rates (10.34), but otherwise are not that different from urban trip rates (9.36–9.50), using statistics based on one of several stratifications found in Appendix E. • The percentage of rural work trips (12 percent) appears to be less than that experienced in most urban settings (typically 15–20 percent). • Rural trip travel times (19–24 minutes, nonwork versus work) are generally shorter than urbanized areas with 1 million plus population and subway or rail (20–32 minutes, non- work versus work). • Rural auto occupancy rates (1.54) are generally higher than small- and medium-sized urbanized areas (1.49–1.52) but equal to, or lower than, the largest metropolitan areas (1.54–1.63). It is strongly recommended that the rates provided in this study from the 1995 ATS for long-distance travel and 2009 NHTS for rural travel be considered for use where local trip rates are not available. Other trip rates in this report, including secondary source parameters (Michigan, Ohio, Canadian surveys, GPS surveys) and NHTS 2001 statistics, are provided for comparative purposes only. Daily Person Trips per Household 2009 NHTS 9.78–10.06 (dependent on stratification) Michigan Travel Counts Surveys 7.64–9.41 (dependent on area and year) Ohio Statewide Household Travel Survey 7.78 (no substratifications) GPS Surveys 8.24–13.56 Data Source Table S.2. Range of comparative rural household trip rates.