Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 137
Appendix E
Current Data Programs
for Monitoring Passenger Travel
and Freight Movement
T
his appendix describes the major sources of travel data in the United
States today. It covers programs addressing passenger travel, those
addressing freight movement, and those addressing both. Issues
and gaps associated with each program are highlighted.
Data Programs for Monitoring Passenger Travel
A comprehensive picture of passenger travel requires measurements of
both local and long-distance travel. Local travel is frequent and often
repetitive, dominated by journeys to work, shopping, schools, and services.
For most people, long-distance travel is less frequent, is dominated
by tourism and business trips, and involves a different set of mode
choices than local travel. A comprehensive picture of local and long-
distance travel across all modes at the national scale has yet to be developed,
though some initial work to this end is under way at the Federal Highway
Administration (FHWA). Building blocks for a national picture of local
travel include the National Household Travel Survey (NHTS), the Census
Transportation Planning Package, and the National Transit Database,
supplemented by a half-century of surveys by local agencies and metro-
politan planning organizations (MPOs). Building blocks for national
measures of long-distance travel include the American Travel Survey
137
OCR for page 138
138 How We Travel: A Sustainable National Program for Travel Data
(ATS), the survey of international air travelers conducted by the Department
of Commerce’s Office of Travel and Tourism Industries (OTTI), tourism
surveys conducted by the private sector, and air carrier traffic statistics
(discussed in the section on data programs for monitoring both passenger
travel and freight movement).
National Household Travel Survey
The NHTS is the only source of national data on personal travel by all
modes in the United States. Data on travel characteristics—trip frequency,
length, and time; travel mode (including nonmotorized modes); and
purpose—are linked with household and personal data (e.g., household
composition, income, age, work characteristics, general location type)
and vehicle ownership and use data to provide a snapshot of personal
travel (Contrino 2009). Limited data are available for states and some
large metropolitan areas, but not below this geographic level. The 2009
NHTS focused on short trips (within 50 miles). Although respondents
recorded their trips of all distances for one day, long-distance trips
were so infrequent that these data could not be used for any substantive
analysis.
The NHTS started in 1969 as the Nationwide Personal Transportation
Survey (NPTS), a home-interview survey conducted in roughly 5-year
cycles. The NPTS was merged with the ATS (described below) and renamed
the NHTS in 2001 to capture both local and long-distance travel in one
survey. The long-distance portion of the NHTS was not successful and
has since been dropped.
Historically, the NHTS has been funded primarily by FHWA, with
modest contributions by other administrations at the U.S. Department of
Transportation (U.S. DOT). States and MPOs can purchase larger local
samples through add-ons to the national sample. With the creation of the
Bureau of Transportation Statistics (BTS) in 1991, that agency became a
cosponsor of the 1995 survey; the cost was split equally in 2001 in the
effort to replace the ATS for measuring long-distance travel. The most
recent survey (2009) was delayed when BTS announced it could no
longer support the effort. FHWA reassumed full responsibility for its
funding and administration and, largely with the support of the depart-
ments of transportation (DOTs) of 14 states and six MPOs that provided
the bulk of the funding—$21 million of the $24 million cost—the survey
OCR for page 139
Current Data Programs for Monitoring Passenger Travel 139
moved forward.1 Currently the program is staffed with only 1 full-time
equivalent (FTE) from FHWA and 2.5 FTE on-site contractors.
In addition to issues of funding stability, the NHTS faces the increasing
challenge of low response rates. Data collection conducted primarily through
landline telephone surveys yielded only a 23 percent response rate to the
initial recruitment for the 2009 survey, a sharp decline from the 56 percent
response rate for the 2001 NHTS.2,3 U.S. DOT staff acknowledge that
the growing share of cellular telephone–only households will require
abandoning complete reliance on landline telephone communication in
future surveys. In addition, greater use of web surveys, Global Positioning
System (GPS) data recorders, and other approaches may be needed. In fact,
a small pilot test of surveying cellular telephone users was conducted as
part of the 2009 survey (Contrino 2010). Subsequent analysis showed that
cellular telephone–only respondents have different travel patterns from
those of other respondents, an issue that must be revisited as the next
survey plan is developed. As a step in this direction, FHWA, together with
the Office of the Secretary of Transportation, is funding a $1.6 million
study to explore a wide range of methods for conducting the next NHTS.
BTS is part of the study team but has not contributed funding.4 Response
rate issues are not limited to the NHTS and are not new, having been
identified in a letter report from the Transportation Research Board
(TRB) to BTS in 2002 and in a TRB Special Report the following year
(TRB 2003).5
Another critical issue is the timeliness of the data. Although many data
products are provided within 6 months to 1 year after survey completion,
the length of time between surveys and the one-snapshot, cross-sectional
approach are problematic. When the 2009 survey was conducted, for
1. The 14 states were California, Florida, Georgia, Indiana, Iowa, New York, North Carolina, South Carolina,
South Dakota, Tennessee, Texas, Vermont, Virginia, and Wisconsin. The six MPOs were Chittenden
County MPO (Vermont), Linn County Regional Planning Commission (Cedar Rapids, Iowa), Maricopa
Association of Governments and Pima Association of Governments (Arizona), Piedmont Regional
Transportation (North Carolina), and Omaha–Council Bluffs Metro Area Planning Agency (Nebraska)
(Contrino 2010). The state and MPO contributions were handled as a pooled-fund project of FHWA, an
arrangement that enables agencies to pool resources for a common purpose. In the case of the NHTS,
the matching fund requirement was waived (Contrino 2009).
2. These are the response rates submitted to the Office of Management and Budget (OMB). The retention
rate for those initially recruited was much higher—80 percent—even higher than the 2001 rate of
70 percent, reflecting a greater effort to obtain responses and the use of incentives.
3. T. Tang, FHWA, personal communication, June 11, 2010.
4. T. Tang, FHWA, personal communication, June 11, 2010.
5. The letter report appears in Special Report 277 as Appendix A.
OCR for page 140
140 How We Travel: A Sustainable National Program for Travel Data
example, the nation was in a deep recession, so that travel, particularly
discretionary travel, was suppressed.
Census Transportation Planning Products
The CTPP provides nationwide passenger work trip data, anchored on
home and work locations. Through 2000, these data were drawn from
the long form of the decennial census of population and housing; more
recently, they have been drawn from the continuous American Community
Survey. Initially collected in the 1960 census to support the definition of
metropolitan areas by the Office of Management and Budget, Journey
to Work data provided through the CTPP support regional and local
transportation planning with geographically detailed information on
where people live and work, how they get to work, and when they
depart for work (Pisarski 2006). These data are available at a fine-grained
geographic level—typically by traffic analysis zones used by planning
agencies or census block groups—to provide inputs for regional travel
demand models for larger MPOs and to serve as primary source data for
smaller agencies with limited modeling capabilities (Pisarski 2006).
Although work trips now account for only slightly more than 20 percent
of all household vehicle trips (Hu and Reuscher 2004, 16) and just 15 percent
of all person trips (Pisarski 2006, 3), these data are vital for understanding
peak demand loads on the transportation system and overall levels of and
options for addressing congestion.
The Census Bureau collects Journey to Work data through its regular
programs, and the transportation community funds the special tabulations
to support transportation planning through the CTPP (Pisarski 2006;
Christopher 2009). FHWA supported initial development of the CTPP,
and worked with the American Association of State Highway and Trans-
portation Officials (AASHTO) and the National Association of Regional
Councils to encourage states to use their federal State Planning and
Research (SP&R) funds and MPOs to use their federal planning funds
through a pooled-fund project to support the CTPP.6 In 2000, a budget
of $3 million supported dedicated staff within AASHTO for continued
development and deployment of the CTPP. The budget has grown to
$5.9 million for 2007 through 2011.
6. SP&R funds and planning funds are authorized from the Highway Trust Fund for specific purposes
enumerated in Sections 104(f ), 134, 135, and 505 of Title 23 and Sections 5303, 5304, and 5305 of Title
49, U.S. Code.
OCR for page 141
Current Data Programs for Monitoring Passenger Travel 141
The effectiveness of the CTPP is now dependent on the American
Community Survey.7 That survey ensures much more timely data relative
to the decennial census and smoothes out data collection costs over time,
but continuous data collection has its own problems. Not the least of these
are smaller (timely) sample sizes, which have not only increased the
variability of the data but also entailed disclosure restrictions, particularly
for the modal and small-area data routinely used by transportation modelers.
Establishing methods for properly interpreting continuously collected
data for trend analysis and accumulating sufficient data for reliable small-
area analyses are works in progress. Resolution of these issues will have
important implications for other surveys, such as the NHTS, for which
moving to a continuous data collection approach is an option.
National Transit Database
Starting in 1978, Congress required all recipients of federal urbanized
area formula grants for transit to participate in a uniform reporting system,
then known as Section 15 for the section of the legislation establishing the
requirement. Today, the NTD, for which the Federal Transit Administration
(FTA) is responsible, is the primary national database of statistics on the
finances, operations, and service characteristics of more than 700 transit
providers in urbanized areas across the United States and more than
1,300 transit providers in nonurbanized areas (FTA 2009). The NTD is
funded by an annual $3.5 million designation from FTA’s grant programs,
and all grantees from the Urbanized Area Formula Program (Section 5307)
and Other-Than-Urbanized Area Formula Program (Section 5311) are
required to report (Giorgis 2009).8 Not surprisingly, compliance is high,
particularly for urban transit systems, because the data are used, among
other purposes, for grant apportionments for transit properties located in
urbanized areas. The data are also used to support the transit section of
the biennial Condition and Performance Report required by Congress and
to measure transit agency performance and serve other benchmarking
purposes.
The data on passenger travel by transit are limited. Passenger trips, or
boardings, are unlinked trips aggregated for each urban and rural transit
7. The decennial census and the American Community Survey were performed in parallel in 2000 to test
comparability and ensure continuity. However, the long form was dropped in the 2010 census.
8. The burden on local transit properties reported to OMB some 3 years ago was 229,634 hours at an esti-
mated cost of about $3.4 million.
OCR for page 142
142 How We Travel: A Sustainable National Program for Travel Data
property. If, for example, a passenger travels by rail and then transfers to
a bus, each boarding is counted as a separate unlinked trip.9 Providers in
urbanized areas also report passenger-miles, derived mainly by sampling,
with larger systems sampling annually and smaller systems every 3 years
(Giorgis 2009). Transit properties in urban areas report passenger data by
mode (e.g., heavy, light, and commuter rail; buses of various types) and by
type of service (direct and contracted out) annually and monthly for unlinked
passenger trips only. (The American Public Transportation Association
[APTA] also gathers monthly data on unlinked passenger trips from its
membership.)10 States report annually on behalf of all rural transit prop-
erties by mode only (Giorgis 2009). The NTD has no data on the charac-
teristics of transit riders or the purpose or time of their trips. It is necessary
to rely on the NHTS for some of these data, but the two data sets are not
linked or linkable. The NTD has no data on passenger trip costs, travel
times, crowding, or service levels and schedule adherence.
Local Travel Surveys
Since the metropolitan transportation studies of the 1950s and 1960s,
regional and local agencies have conducted surveys of local travel to
support transportation planning. Originally funded by programs of FHWA
and Section 701 grants from the Department of Housing and Urban Devel-
opment (HUD), household travel surveys and other data collection activities
of MPOs are eligible for funding from both federal highway and federal
transit programs. Surveys are conducted to support travel demand models
used for regional transportation plans, corridor studies, and the develop-
ment of major projects. While these surveys provide a wealth of local
information, the lack of standardized methods and data definitions inhibits
comparisons among areas or aggregation into a national picture of local
travel. Local travel surveys became less common as costs increased, and
9. According to BTS, which uses monthly unlinked passenger trip data gathered by the American Public
Transportation Association (APTA) as source data for the Transportation Services Index (described
in Chapter 1), all ridership data reported relate to trips, not to people. The use of passes, transfers, joint
tickets, and cash by people transferring from one vehicle to another, one transit mode to another, and
one public transit agency to another makes it difficult to count people. Boardings (unlinked passenger
trips) can be counted more accurately. At the largest public transit agencies, even boardings may be
estimated for portions of the ridership.
10. APTA is a nonprofit association of transit systems and commuter rail operators, transit associations,
state DOTs, and other organizations. APTA collects these data on a voluntary basis from its member-
ship, which includes virtually all of the larger and many medium-sized transit properties as well as
some small transit properties that are not included in the NTD database.
OCR for page 143
Current Data Programs for Monitoring Passenger Travel 143
MPOs became dependent on the CTPP for small-area data and the NHTS
for model calibration factors.
American Travel Survey
The ATS was conducted in 1995 to capture travel by all modes for trips
to destinations more distant than 50 miles. Telephone interviews of
67,000 households conducted four times during the year provided data
on passenger flows among states and major metropolitan areas. The
$18 million survey was conducted by the Census Bureau and funded entirely
by BTS. The ATS was the successor to the smaller National Travel Survey,
conducted by the Census Bureau in the 1970s to measure travel among states.
As described earlier, the ATS was combined with the NPTS to create
the 2001 NHTS in the hopes of reducing costs and establishing a com-
prehensive, internally consistent picture of local and long-distance travel.
However, the NHTS failed to produce reliable passenger flow data among
states and major metropolitan areas. While sample sizes and response
rates were higher for the ATS than for the NHTS, the ATS shared many of
the same challenges of respondent burden and cost.11
Recent interest in high-speed intercity rail investments underscores
the continuing importance of data on intercity passenger flows by origin
and destination. The ATS data are more than 15 years old but are the
only national source of publicly available information on passenger
travel by surface transportation modes to support analyses of intercity
transportation.
Survey of International Air Travelers
OTTI has conducted a survey of international air travelers since the early
1980s. The survey is an important source of data on expenditures of
foreign visitors to the United States and corresponding expenditures
of U.S. residents while traveling abroad, and it has been used as a measure
of foreign travel by Americans and domestic travel by foreigners. In the
past, these data were a primary input to balance-of-payments information
for the U.S. national accounts, but they have largely been supplanted by
credit card company data on inbound visitor spending.12
11. See, for example, the discussion in TRB Special Report 277 (TRB 2003).
12. Credit card companies doing business in the United States are now required by regulation to transmit
data quarterly to the Department of Commerce on the expenditures made by foreigners visiting the
United States.
OCR for page 144
144 How We Travel: A Sustainable National Program for Travel Data
Data are obtained from survey instruments administered to outbound
travelers either on board flights or in the departure gate area at 30 U.S. gate-
way airports (24 of which are major international gateways) for a sample
of aircraft flying between the United States and foreign destinations.13 To
adjust for over- and undersampling, survey observations are weighted to
census data from Immigration and Customs forms that regulate inbound
and outbound visitors.14 The surveys seek to obtain representative data on
all international air travel.15 Given the complex combinations of carriers,
origin–destination pairs, airports, and flights, however, this coverage is
not always feasible; thus, the focus is on the top 20 origin countries.
Survey data on travel within the United States are most important from
a transportation perspective. Survey data are collected on travel to states
and major U.S. destinations by nationality of visitor, use of transportation
facilities, mode of transport within the United States, group size, and length
of stay. The 2008 survey cost about $1.7 million, and the data were released
for 20 states and Guam in April 2009.16 The manual form I-94 for inbound
travelers to the United States by sea and air is being phased out for those
countries participating in the Visa Waiver Program, which will affect more
than two-thirds of overseas travelers (OTTI 2010).17 The same informa-
tion will be collected from automated sources related to the passenger’s
itinerary, which should improve the accuracy and timeliness of reporting.
Data from the Private Sector: Tourism Surveys
Tourism is a major economic activity for a number of states and large
companies. The result has been the development of private sources of
data to guide market decisions. Timely data are of the essence for the
tourism industry, which is focused on “this season,” while public-sector
surveys typically measure long-distance travel activities over a minimum of
13. The program targets two separate populations: (a) non-U.S. residents who have traveled to the United
States and who are returning home, and (b) U.S. residents departing the United States on the originating
leg of their flight. Foreign visitors are being asked to account for their activities and expenditures
retrospectively, while U.S. outbound travelers are asked to estimate their activities and expenditures
prospectively.
14. The I-92 form is required of all domestic and foreign air carriers to report total passengers by flight,
and the I-94 form is required of foreign visitors to provide information on their prospective visit to the
United States.
15. Mexican air travel is included, but through an arrangement with Statistics Canada, that agency provides
its survey to the United States, thus avoiding duplication.
16. Data provided by OTTI staff and reported by D. Frechtling, George Washington University, June 29, 2010.
17. The manual I-94W form will continue to be required at land borders and for non–Visa Waiver Program
countries.
OCR for page 145
Current Data Programs for Monitoring Passenger Travel 145
1 year, thus lagging behind industry needs. One source of current tourism-
oriented travel data is the D. K. Shifflet & Associates (DKSA) Survey System.
Founded in 1982, DKSA is a private firm serving an array of clients,
typically destination attractions, cities, theme parks, and travel associations,
but also public entities, such as state offices of economic development and
tourism promotion. DKSA retains data sets dating from 1995, derived
from a family of monthly surveys, including mail surveys and Internet-based
panel interviews, which yield data on more than 75,000 U.S. resident
traveling households and their travel each year, measured over a 3-month
recall period.18 The historical data permit clients to track long-term trends
pertaining to their interests. Coupled with modeling capabilities, the
company provides estimates of traveler volume by such metrics as trips;
number of travelers; length of stay; purpose of stay and travel activities;
visitor spending; and other attributes, such as demographic data, that
serve the market research and service planning needs of clients.19,20
Transportation-related data include mode of transportation, trip destina-
tion, and traveler transportation expenditures.
Because of the expense of this ongoing activity, great care is taken to
ensure that proprietary information is not divulged by clients or others.
Data are typically licensed to clients for use, but DKSA retains ownership.
The proprietary nature of the data can be a particular concern with regard
to services to public entities, which may have problems with observing
market protection agreements because of Freedom of Information Act
requests. This is a good example of the need to establish a realistic sense
of appropriate and effective boundaries between private data providers
and data users at the federal, state, and local levels.
Other National Surveys Containing Personal Travel Data
Several other surveys are conducted in which transportation data are
gathered for the purpose of comprehensiveness, not to meet a specific
transportation need. Three such surveys are reviewed here: the Consumer
Expenditure Survey (CEX) and the American Time Use Survey (ATUS),
18. DKSA uses a panel company to recruit nationally representative panels of households that have
agreed in advance to participate in periodic surveys. Extensive information about the household is
gathered at the time of recruitment, and panel response is typically two to three times higher than that
obtained by contacting households randomly.
19. Using DKSA’s visitor volume and spending database as input, IHS Global Insight is able to generate
estimated revenues as well as direct, indirect, and induced spending.
20. J. Caldwell, DKSA, personal communication, June 30, 2010.
OCR for page 146
146 How We Travel: A Sustainable National Program for Travel Data
which measure how people spend their money and their time, respectively,
and the American Housing Survey (AHS), which contains questions on
work travel and availability of transportation services. The Bureau of Labor
Statistics sponsors the annual CEX and ATUS, and HUD sponsors the AHS.
The Census Bureau conducts all three. These and other similar surveys
provide opportunities for expanded transportation applications.
Consumer Expenditure Survey
Data for the CEX are collected through a direct interview and a detailed
diary kept by selected households to facilitate recall of those expenditures
easily overlooked. The number of observations is sufficient for national
summaries, but detailed area statistics are available for only 18 metropolitan
areas, down from 24 to 28 areas in past years. The surveys include a
comprehensive set of descriptive variables, such as persons, workers, and
vehicles in the household; type of housing; income and age; and gender
and racial characteristics. Expenditures on about 10 transportation items
are collected, but more detail is available on a selective basis to researchers
where the sample size permits.21 The survey does not address individual
trips but rather records aggregate expenditures for the reporting period,
which are then accumulated to an annual total.22 One of the strengths
of the survey is that it differentiates purchases in the home community
(restaurant food and transit) from spending away from home or during
out-of-town travel. A weakness from a transportation perspective, but not
according to the survey’s intent, is that the expenditures tallied include
only those paid for by the consumer and not those reimbursed by others.
This means that business travel, or even event travel for which a school or
church reimburses users, is not included. This represents a substantial
gap in the understanding of transportation spending.
American Time Use Survey
The first survey of time usage was conducted in 2003 and reported in
late 2004. Respondents are selected from among households that have
21. For example, the item “public transportation” is reported as a single value in standard reporting, but
unlike its use in transportation parlance, this item includes all modes of transportation for which one
might purchase a ticket, including airlines, buses, rail lines, cruise lines, taxis, and of course urban
mass transportation.
22. It is possible, however, to calculate actual trip costs from the data. Thus, the average expenditure per
household may be small (e.g., the average amount spent on cruise trips per year), but when the total
number of households making such trips is recorded (e.g., say, 2 percent of all households), the average
per trip expenditure can be calculated (e.g., $40/.02 = $2,000 per cruise).
OCR for page 147
Current Data Programs for Monitoring Passenger Travel 147
nearly completed their participation in the Current Population Survey.
One person in the selected household is interviewed regarding his or her
previous day’s activities. As in the CEX, a diary is employed to support
recall. Seventeen main categories of time use with multiple subelements
are reported. The variables that describe the respondents parallel those of
the CEX, so that distinctions made on the basis of age, gender, race, and
income are possible. More broadly, the survey parallels the CEX in that
the ways people spend time and money can be quite similar, particularly
for out-of-home activities. Recognizing this link, the Bureau of Labor
Statistics joins the two surveys in its website presentation. The ways in
which people spend time are important to transportation analysts and
modelers not just because they constitute travel time information such as
the time spent in traveling to recreation, but also because transportation
modeling is increasingly focused on activity-based models that tie travel
to the actual activities in which people engage.23 The use of this survey is
in its infancy, and it could become increasingly significant to transportation
analysts.
American Housing Survey
The AHS has been conducted since 1973 (then called the Annual Housing
Survey). Its main focus is on the condition and characteristics of the
U.S. housing stock, with associated demographic statistics for the related
households. The housing unit is the main sampling reference, and the
surveyors return to that unit repeatedly, interviewing whoever resides
there. New housing units are added to represent new construction.
The survey consists of a representative national sample and a series of
metropolitan samples that are rotated among more than 40 of the larger
metropolitan areas. Both are conducted in every odd-numbered year
rather than, as previously, in alternating years. The survey covers a set of
work travel questions paralleling those of the ACS, including the number
of vehicles owned, mode of travel to work, time of departure, and travel
time. In addition, it includes questions on distance to work and time spent
working at home. From a metropolitan planning viewpoint, the AHS is
probably the only survey that asks questions about the quality of housing
and the immediate neighborhood, such as noise, smoke odors, and reasons
for moving away from a neighborhood, as well as the availability of
nearby services, such as transit and shopping. It is also the sole source of
23. The smallest geographic unit for which the data are available, however, is the states.
OCR for page 150
150 How We Travel: A Sustainable National Program for Travel Data
by sample size and disclosure issues, which are more severe than in pas-
senger surveys given the heterogeneous and uneven nature of business
transactions. The sample size was cut to 50,000 establishments in 2002
but restored to 100,000 in 2007, which allowed better representation of
international gateways. CFS and FAF regions are states and selected large
metropolitan areas within states.29 County-level detail, requested by many
users, cannot be derived without the collection of local data to supple-
ment the CFS and FAF.30
Concern about supply chains is twofold. The CFS and FAF track flows
among regions rather than among business establishments and do not
capture many of the variables of mode and route choice that can be affected
by public policy. The CFS also depends on shippers’ knowing where and
how shipments were sent, yet many supply chains often involve third-party
logistics firms that manage the freight shipments (Schofer 2006; TRB 2003).
Thus shippers often do not know by what mode a shipment is made or
through what intermediate facility, which results in inaccurate or incom-
plete responses to the CFS.
Transborder Freight Data Program and Foreign Trade Statistics
The North American Transborder Freight Database was developed in
response to the signing of the North American Free Trade Agreement
(NAFTA) by the United States, Canada, and Mexico in late 1992 to help
monitor trade flows among these countries. BTS contracted with the
Census Bureau to provide previously unpublished data from the census
foreign trade statistics (BTS undated, a). Data on freight flows are reported by
commodity type, mode of transportation (all surface modes, air, and water),
and U.S. port of entry or exit for imports from and exports to Canada and
Mexico (BTS 2010a).31 Commodity detail by port is not available because
of disclosure limitations (BTS undated, a). State of origin and state of
destination for exports and imports, respectively, are also reported for
29. Boundaries of metropolitan areas are either Metropolitan Statistical Areas (MSAs) or Consolidated
Statistical Areas (CSAs) as defined by OMB. When a CSA or MSA spans a state line and both state
portions are large enough to support CFS tabulations, each state portion becomes a separate CFS
region. When only one state portion is large enough to support CFS tabulations, only the larger portion
becomes a separate CFS region; the smaller portion is included in the state totals. The FAF uses the
same geographic boundaries as the CFS.
30. Methods of local freight data collection are being developed through the National Cooperative Freight
Research Program.
31. BTS also provides incoming border-crossing data for vehicles and passengers, containers, and pedestrians
for land ports on the U.S. border with Canada and Mexico (BTS 2010a).
OCR for page 151
Current Data Programs for Monitoring Passenger Travel 151
surface transportation modes, but these data may not reflect the physical
point of origin.32 No data are available at the metropolitan-area level.
In addition to monitoring North American trade flows, the North Amer-
ican Transborder Freight Database is used by FHWA, which reimburses
BTS for the purchase of the data from the Census Bureau,33 to implement
the Coordinated Border Infrastructure Program and to provide one of the
data sources for the FAF.34 Finally, the database is used by states, MPOs
and local governments, and the private sector for trade corridor studies,
transportation infrastructure planning, and marketing and logistics planning
(BTS 2010a). Monthly and annual data are available online on the BTS
website, where users can access the data through an interactive search-
able interface or download the data in raw table formats (BTS undated, a).
The transborder data are part of the larger collection of foreign trade
statistics and suffer the same limitations associated with using trade as a
surrogate for transportation. Data on inland destinations of imports and
origins of exports are frequently inaccurate, data on domestic modes used
between international gateways and inland locations are generally lack-
ing, and in-transit flows (moves between foreign countries through the
United States) can be identified only by joint efforts of all three NAFTA
partners. Some of these limitations were addressed in a survey of domestic
transportation of U.S. foreign trade in 1970 and 1975; more recent efforts to
deal with these limitations depend heavily on models (CNSTAT 2005).
Finally, by definition the data are restricted to North American trade flows.
Rail Carload Waybill Sample
The railroad industry reports data on rail freight shipments to the Surface
Transportation Board (STB), which replaced the Interstate Commerce
Commission in 1995 (the industry was partially deregulated in 1980). All
railroads terminating 4,500 or more revenue carloads per year for 3 years
in a row are required to file a stratified sample with the STB, averaging
32. In general, import data are more accurate than export data because U.S. Customs uses the former data for
enforcement purposes (BTS undated, a). Although the original intent of collecting data on state of origin
was to capture the state where the goods were grown, manufactured, or produced, in practice the state of
origin may represent the mailing address of the U.S. exporter or an intermediary or (mainly for agricul-
tural and bulk shipments) the consolidation point or port of exit, rather than the physical state of origin.
33. In 2010, it cost $52,575 to purchase the transborder freight data from the Census Bureau, which was
reimbursed by FHWA. BTS provides programming support consisting of 0.4 BTS staff plus 1 FTE
contractor, and the Census Bureau provides support consisting of 0.2 FTE.
34. BTS prepares custom tables of the transborder and border crossing data for FHWA to use in calculating
the apportionment of funds to states under the Coordinated Border Infrastructure Program (BTS 2010a).
OCR for page 152
152 How We Travel: A Sustainable National Program for Travel Data
about 3 percent of their waybills. Virtually all of the data are filed elec-
tronically. A contractor (RAILINC) collects, compiles, and edits a stratified
sample of carload waybills in a confidential version. STB and the Federal
Railroad Administration (FRA) share equally the annual cost of collecting
and processing the data, which totaled $322,000 for the 2010 Carload
Waybill Sample.35 Access to the data is restricted because the sample con-
tains competitive shipping and pricing information.
The contractor also creates a public-use file—the Public Use Waybill
Sample—available for download on the STB website at no cost.36 The data
include origin and destination points, types of commodities shipped
(aggregated to the five-digit Standard Transportation Commodity Code
[STCC] level),37 number of cars, tons, revenue, length of haul, participating
railroads, and interchange locations (CTRE undated). Origin and destina-
tion points are reported by Economic Areas (as defined by the Bureau of
Economic Analysis [BEA]), and junction points are reported by state or
province, rather than by freight station or city name, to avoid disclosure of
data from individual rail carriers.38 As a result of these restrictions, only
about 45 to 50 percent of the total waybill records in the public-use file
contain full geographic data (STB undated).
Waterborne Commerce Statistics
The Rivers and Harbors Appropriations Act of 1922 granted the U.S.
Army Corps of Engineers (USACE) the legal authority to collect, process,
distribute, and archive data on all foreign and domestic waterborne com-
merce on U.S. waters (BTS 2010a).39 For domestic commerce, all vessel
operators of record must report to USACE waterborne traffic at the ports
and harbors and on the waterways and canals of the United States and
its territories (NDC undated). For movements with cargo, the point of
loading and unloading of each commodity must be delineated. To protect
confidentiality, commodities are grouped into general categories, and if
three or more vessel operating companies do not carry a particular com-
modity from a region of origin to a region of destination, that commodity
35. J. Palley, FRA, personal communication, May 27, 2010.
36. P. Aguiar, STB, personal communication, May 28, 2010.
37. An STCC code is a seven-digit numeric code representing 38 commodity groupings, developed on the
basis of commodity descriptions used by freight rail and motor carriers.
38. The origin and destination BEA areas for a commodity shipment are included only if there are at least
three freight stations and at least two more freight stations than railroads in the BEA.
39. Amended and codified in 33 U.S.C. 555.
OCR for page 153
Current Data Programs for Monitoring Passenger Travel 153
is reclassified as “unknown and not elsewhere classified.” The Waterborne
Commerce Statistics Center of USACE compiles and publishes state-to-
state and region-to-region origin–destination movements by tons and
commodity by code in a series of annual publications entitled Waterborne
Commerce of the United States.40 The fiscal year 2010 appropriations for
these activities totaled $4,488,660, and a staff of 28 FTEs supports data
collection and processing.41,42
Foreign waterborne commerce includes imports, exports, and in-transit
traffic among the United States, Puerto Rico, and the Virgin Islands and
any foreign country. Since calendar year 2000, foreign waterborne statistics
have been derived primarily from data purchased from the Port Import
Export Reporting Service (PIERS). Data from U.S. Customs and the Census
Bureau also are used. Prior to 2000, data on foreign waterborne commerce
were supplied solely by the Census Bureau.
USACE uses the Waterborne Commerce Statistics to analyze the feasi-
bility of new water transportation projects and activities, to set priorities
for new investment and rehabilitation, and to manage the operation and
maintenance of existing projects. The data are also used by other federal
agencies, for example, as input for the U.S. national accounts and for
emergency management and homeland defense. Summary statistics
that do not disclose movements of individual companies are released to
the public.
Private Sources of Freight Data
Shippers and carriers maintain significant amounts of data on freight
movements for their own uses and sometimes share that data with con-
sultants and trade associations. Two private sources with long histories
include TRANSEARCH and PIERS.
TRANSEARCH
This database was initially developed by Reebie Associates to meet freight
industry needs for market data, and it has since been acquired by IHS
Global Insight. After passage of the Intermodal Surface Transportation
40. Parts 1–4 summarize data on the movements of vessels (trips) and commodities (in short tons) at ports
and harbors and on waterways and canals in the United States and its territories (NDC undated). Part
5 provides summary national statistics on foreign and domestic waterborne commerce on U.S. and
territorial waters by tonnage and ton-miles of commodities (USACE 2009).
41. Data provided by S. Hassett, USACE, July 20, 2010.
42. The staffing numbers do not include the director, two administrators, and one supervisor.
OCR for page 154
154 How We Travel: A Sustainable National Program for Travel Data
Efficiency Act of 1991, the product was adapted under a competitively
awarded federal contract to develop and then privately maintain annual
freight flow data organized by county. To this end, the database uses
public, commercial, and proprietary data sources, among them a variety
of federal data sources including the CFS and an ongoing, long-term
private shipment sample from many of the nation’s largest motor and rail
carriers.43 As a result, TRANSEARCH supplies information on U.S. freight
flows by county origin and destination, four-digit commodity, and mode and
submode of transportation, and offers such additional geographic units as
zip codes, Economic Areas, states, and the nation (IHS Global Insight 2010).
The data are based in part on IHS Global Insight’s economic data, issued
annually for the previous year, and can be paired with short- and long-term
freight forecasts at consistent levels of data detail. The database is sold to
a wide range of customers, including railroads, trucking companies, and
port authorities for market and network assessment; states and MPOs for
freight planning; and financial groups for public infrastructure investment
analyses. One drawback of the database is that its private shipment sample
depends on voluntary participation and thus is not a random sample,
although the vendor attempts to attract a diversity of carrier types.
A second drawback is the proprietary nature of the database, and hence
its lack of transparency. Users can obtain a reasonably complete account
of the construction of the database, and its elements are subject to a degree
of market testing in that industry clients can and do provide feedback to
the vendor. Nevertheless, users must accept on faith the validity of the
results, particularly at the county level.
Port Import Export Reporting Service
Launched more than 30 years ago by the Journal of Commerce, now a
division of UBM Global Trade, PIERS collects data on imports and exports.44
Import information is gathered from vessel manifests and from U.S.
Customs Automated Manifest Systems data from all U.S. ports. The PIERS
quality control staff verifies the data monthly by comparing them against
a list of all vessels arriving at U.S. ports, provided by U.S. Customs, to
43. The incentives for carriers to provide the data include assured confidentiality in the treatment of the
collected data, free analyzed data in return, and no attempt to collect sensitive data (e.g., pricing)
(Ciannavei 2010).
44. PIERS is the oldest data set on private waterborne trade data and most established in the U.S. federal
government market. Other companies, however, such as Zepol and Datamyne, also sell such data and
are competitors to PIERS.
OCR for page 155
Current Data Programs for Monitoring Passenger Travel 155
identify any discrepancies (PIERS 2010). Export information is gathered
by dedicated PIERS staff from bills of lading at all U.S. ports, again verified
by PIERS quality control staff monthly against a list of all vessels exiting
U.S. ports supplied by U.S. Customs (PIERS 2010).45 UBM Global Trade
sells these trade data, enhanced with detail on commodity type and value
and available monthly, primarily to private companies (e.g., large container
companies) for determining market share and analyzing the competition.
Data Programs for Monitoring Both Passenger Travel
and Freight Movement
Some data programs measure both passenger travel and freight movement,
in cases in which people and goods are generally carried in the same
conveyances or on common infrastructure. These programs include the
air carrier traffic statistics, the Highway Performance Monitoring System
(HPMS), and the Vehicle Inventory and Use Survey (VIUS). Such data are
also collected by several private sources.
Air Carrier Traffic Statistics
Data on air passenger travel are some of the best in the transportation
industry. The data are both detailed and timely (released monthly). The
Office of Airline Information in BTS makes available a public version of
what is known as the Origin & Destination (O&D) Survey—a database of a
continuous 10 percent sample of airline tickets sold by certificated air car-
riers in scheduled domestic passenger service. The data for each passenger
include point of origin, destination, airline, class of service, and fare
(BTS undated, b). BTS spends $300,000 annually for data collection by a
contractor, which is supported by 0.5 FTE at BTS.46
BTS also collects airline traffic data from all U.S. carriers, which include
the number of passengers and the weight of cargo (mail and freight) by
nonstop flight segment and by market or in-flight segment (BTS 2010a).47
45. There are 48 of these ports throughout the United States, including Alaska, Hawaii, and Puerto Rico.
46. S. Smith, Research and Innovative Technology Administration, personal communication, April 29, 2010.
47. The data on air cargo from carriers are limited, partly because of the structure of the air cargo industry
where the carriers are often wholesalers not knowing much more than the weight of what they are
carrying, at least domestically. As a result, these data are of limited use to analysts and decision makers
because they reveal so little of the “what and why” of the use of air cargo and almost nothing about
where the air cargo leg fits into the supply chain of the shipment being flown.
OCR for page 156
156 How We Travel: A Sustainable National Program for Travel Data
Domestic data, that is, for traffic between airports located within the
boundaries of the United States and its territories, are unrestricted; inter-
national data, that is, between the United States and a foreign point, are
restricted for 6 months after the report date (BTS undated). These data as
well as other nontravel airline statistics gathered by BTS are widely used
by customers within U.S. DOT as well as Congress, the Department of
Homeland Security, state and local governments, the air transportation
industry, researchers, academia, and the public (BTS 2010a).
Highway Performance Monitoring System
Development of the HPMS was initiated in 1978 to meet congressional
requirements to report on the nation’s highway needs, among other
purposes, and is widely used by FHWA and other federal agencies, states,
MPOs, local governments, and other customers. National-level data on
highway inventory, condition, performance, and operating characteristics
are collected annually for all public roads, with the greatest detail being on
major highways.48 Data on vehicle miles traveled (VMT) are based on short-
term as well as continuous traffic counters. State DOTs collect and report
the data to FHWA in a bottom-up approach. FHWA provides guidelines on
data collection in the HPMS Field Manual (FHWA 2010) and provides
technical support and software to facilitate reporting and assist with data
quality checks. Nevertheless, the quality of the data continues to reflect state-
to-state variability. SP&R funds are available to fund data collection by the
states, which also use their own funds. FHWA provides about $400,000
annually for system development and support, performed by about five
FTEs.49,50
FHWA uses HPMS data in its biennial Condition and Performance
Report to Congress, in the calculation of apportionment formulas for
federal highway funds (a strong incentive for states to provide the data),
in support of analytic models,51 and for other reports such as the annual
48. Data are collected on the distribution of travel by six vehicle classes for all public roads except for
non-federal-aid local roads and rural minor collectors. Areawide summary information is provided
for urbanized, small urban, and rural areas and for air quality nonattainment and maintenance areas
(FHWA 2008). Major data items, such as average annual daily traffic, are counted in full, not sampled
(R. Gillmann, FHWA, personal communication, April 16, 2010).
49. No dollar amount for the cost of state data collection was available. In a burden estimate reported to
OMB, however, FHWA estimated that the 52 responses would take approximately 93,600 hours.
50. R. Gillmann, FHWA, personal communication, April 16, 2010.
51. The HPMS is the data source for the Highway Economic Requirements System model, which in turn
produces the information for the biennial Condition and Performance Report to Congress. HPMS data are
also used by the FAF to calibrate base-year assignments and forecast future freight flows (FHWA 2008).
OCR for page 157
Current Data Programs for Monitoring Passenger Travel 157
Highway Statistics and the monthly Traffic Volume Trends (Gillmann 2009).52
The HPMS is also used for safety reporting by National Highway Traffic
Safety Administration (NHTSA) staff, who combine their safety databases
with HPMS data on VMT to report fatality and injury rates by road class.
Finally, states and MPOs use HPMS data to assess highway investment
needs and conduct air quality conformity analyses. A recent reassessment
of the HPMS (FHWA 2008) resulted in recommendations to, among
other things, reduce variability in state-to-state reporting; provide for
geographic locating, analysis, comparison, and reporting of data; and
expand data collection on VMT to ramps and interchanges. FHWA hopes
to have these changes implemented by 2011.
VMT estimation is probably the most common use of HPMS (FHWA
2008). VMT is calculated and used at the national, state, and local levels.
One of the original intents of the HPMS was to develop a consistent basis
for VMT estimation nationally. Nevertheless, VMT data are supplied by
individual states, often using their own data collection procedures.
Vehicle Inventory and Use Survey
The VIUS was launched by the Census Bureau in 1963 as the Truck
Inventory and Use Survey and was conducted every 5 years from 1967
through 2002. The 2002 VIUS included typical configurations, weights, fuel
usage and economy, miles traveled, economic activity served, commodities
carried, and other characteristics for a sample of 130,000 trucks, vans,
minivans, and sport utility vehicles drawn from state registration files.
The sample supported summary tables by state where the vehicles were
based or registered. Data on where the vehicles operated were limited to
percent miles out of state or in Canada and Mexico.
VIUS information has been used in a variety of national and regional
studies. VIUS data on payloads by commodity and vehicle type have been
central to converting FAF and CFS tonnages into vehicle movements.
VIUS data also are major inputs to models of fuel use, carbon footprint, and
the air quality consequences of vehicle activity. They are essential com-
ponents of truck size and weight studies and highway cost allocation
studies. Finally, VIUS provides data on passenger travel for personal and
business purposes by pickups, vans, minivans, and sport utility vehicles.
52. Traffic Volume Trends is based on hourly traffic count data reported by the states on the basis of data
collected at approximately 4,000 continuous traffic counting locations nationwide; those data are
used to estimate the percent change in traffic for the current month compared with the same month
in the previous year. Estimates are readjusted annually to match the VMT reported from the HPMS.
OCR for page 158
158 How We Travel: A Sustainable National Program for Travel Data
Plans to expand the VIUS to automobiles and buses, providing a com-
plete picture of vehicle use for passenger travel and freight movement by
all vehicles except those owned by government, died with the survey as
the result of a governmentwide budget rescission to help defray the costs
of the war in Iraq and Hurricane Katrina. The cost of restoring the VIUS
is about $10–14 million depending on the scope and whether national- or
state-level precision is sought (BTS 2010b).
Private Data Sources
Popular concern about highway congestion has inspired the development
of several private sources of travel data. Congestion monitoring started
with the HPMS and the biennial Condition and Performance Report and
was refined through the annual Urban Mobility Reports of the Texas
Transportation Institute. Private firms began capturing speed data directly
through traffic reporting services sponsored by local media outlets and
through carriers’ reports on their geographic locations to communications
vendors for dispatching and fleet management services; they now capture
these data as well through individuals reporting their locations via smart
phones. A recent example is the system developed by INRIX, described in
more detail in Chapter 3, which provides congestion information to drivers
in return for being able to monitor their speed through their smart phones.
While traffic monitoring services provide little beyond speed data, future
services may be able to include data on traffic volumes, traveler character-
istics, and perhaps even goods carried.
References
Abbreviations
BTS Bureau of Transportation Statistics
CNSTAT Committee on National Statistics
CTRE Center for Transportation Research and Education
FHWA Federal Highway Administration
FTA Federal Transit Administration
NDC Navigation Data Center
OTTI Office of Travel and Tourism Industries
PIERS Port Import Export Reporting Service
STB Surface Transportation Board
TRB Transportation Research Board
U.S. Army Corps of Engineers
USACE
OCR for page 159
Current Data Programs for Monitoring Passenger Travel 159
BTS. 2010a. Significant Accomplishments, Fiscal Year 2009. Research and Innovative
Technology Administration, Washington, D.C.
BTS. 2010b. Options and Costs for Restoring the Vehicle Inventory and Use Survey.
Briefing presented at the Meeting of the Advisory Committee on Transportation
Statistics, Washington, D.C., Oct. 8.
BTS. 2008. Commodity Flow Survey. Survey Overview and Methodology and
Frequently Asked Questions. http://www.brs.gov/publications/commodity_flow_
survey/. Accessed March 11, 2010.
BTS. Undated, a. North American Transborder Freight Data: Overview and
Frequently Asked Questions. http://www.bts.gov/cgi-bin/breadcrumbs/Print
Version_redesign.cgi?date=21165415. Accessed April 21, 2010.
BTS. Undated, b. Office of Airline Information, Sources of Aviation Data. http://www.
bts.gov/programs/airline_information/sources/. Accessed April 16, 2010.
Christopher, E. 2009. Census Data for Transportation Planning. FHWA, Matteson, Ill.
Briefing presented to the Committee on Strategies for Improved Passenger and
Freight Travel Data, Washington, D.C., Dec. 10.
Ciannavei, P. 2010. IHS Global Insight, Inc., West Hartford, Conn. Briefing and
handout presented to the Committee on Strategies for Improved Passenger
and Freight Travel Data, Washington, D.C., Feb. 18.
CNSTAT. 2005. Measuring International Trade on U.S. Highways (J. L. Horowitz
and T. Plewes, eds.), National Academies Press, Washington, D.C.
Contrino, H. 2010. The National Household Travel Survey. FHWA, Washington, D.C.
Presentation to a stakeholder meeting at the American Automobile Association,
Washington, D.C., Feb. 25.
Contrino, H. 2009. The National Household Travel Survey. FHWA, Washington, D.C.
Briefing presented to the Committee on Strategies for Improved Passenger and
Freight Travel Data, Washington, D.C., Dec. 10.
CTRE. undated. Freight and Trade Datasets for Transportation Planning, Rail Waybill
Data. http://www.ctre.iastate.duc/research/bts_wb/cd-rom/freight/rail.htm.
Accessed March 16, 2010.
Donnelly, R. 2010. Best Practices for Incorporating Commodity Flow Survey and
Related Data into the MPO and Statewide Planning Processes. Parsons Brinck-
erhoff, Inc., Albuquerque, New Mexico, requested by the AASHTO Standing
Committee on Planning, August.
FHWA. 2010. Highway Performance System Field Manual. Office of Highway Policy
Information, U.S. Department of Transportation, Washington, D.C., September.
FHWA. 2008. HPMS Reassessment 2010+, Final Report. Office of Highway Policy
Information, U.S. Department of Transportation, Washington, D.C., September.
FTA. 2009. National Transit Database. http://www.ntdprogram.gov/ntdprogram.
Accessed April 15, 2010.
Gillman, R. 2009. HPMS 2010+ Reassessment Overview. FHWA, Washington, D.C.
Briefing presented to the Committee on Strategies for Improved Passenger and
Freight Travel Data, Washington, D.C., Dec. 10.
Giorgis, J. 2009. The National Transit Database: Passenger Travel Data. Federal
Transit Administration, Washington, D.C. Briefing presented to the Committee
OCR for page 160
160 How We Travel: A Sustainable National Program for Travel Data
on Strategies for Improved Passenger and Freight Travel Data, Washington,
D.C., Dec. 10.
Hu, P. S., and T. R. Reuscher. 2004. Summary of Travel Trends: 2001 National House-
hold Travel Survey. FHWA, U.S. Department of Transportation, Washington,
D.C., December.
IHS Global Insight. 2010. Transearch® and Related Applications. http://www.ihs
globalinsight.com/ProductsServices/ProductDetail2322.htm. Accessed April 6,
2010.
NDC. Undated. U.S. Waterway Data, Waterborne Commerce of the United States.
http://www.iwr.usace.army.,mil/ndc/data/datawcus.htm. Accessed April 19, 2010.
OTTI. 2010. DHS Streamlines Visa Waiver Travel Process, Arrival/Departure Form
(I-94W) Changeover to Electronic Record. TI News, Washington, D.C., June 23.
PIERS. 2010. About PIERS. UBM Global Trade. http://www.piers.com/index.crm?
page=about. Accessed Sept. 24, 2010.
Pisarski, A. E. 2009. Testimony on Research and Development to Support the
Department of Transportation’s Strategic Goals. Submitted to the Subcommit-
tee on Technology and Innovation of the House Committee on Science and
Technology, U.S. House of Representatives, Washington, D.C., Nov. 19.
Pisarski, A. E. 2006. NCHRP Report 550 and TCRP Report 110: Commuting in
America III: The Third National Report on Commuting Patterns and Trends.
Transportation Research Board of the National Academies, Washington, D.C.
Schofer, J. L. 2006. Freight Database for the Future: Workshop Summary:
Observations on Improving the 2007 CFS. In Transportation Research Circular
E-C088: Commodity Flow Survey Conference. Transportation Research Board of
the National Academies, Washington, D.C. http://onlinepubs.trb.org/onlinepubs/
circulars/ec088.pdf.
STB. Undated. Industry Data—Economic Data: Waybill. http://www.stb.dot.gov/
stb/industry/econ_waybill.html. Accessed March 16, 2010.
TRB. 2003. Special Report 277: Measuring Personal Travel and Goods Movement.
Transportation Research Board of the National Academies, Washington, D.C.
http://onlinepubs.trb.org/onlinepubs/sr/sr277.pdf.
USACE. 2009. Waterborne Commerce of the United States, Calendar Year 2008,
Part 5: National Summaries, Introduction. Waterborne Commerce Statistics
Center, New Orleans, La., Dec. 31. http://www.iwr.usace.army.mil/ndc/wcsc/
pdf/wcusnatl08.pdf. Accessed April 19, 2010.
U.S. Census Bureau. 2009. Commodity Flow Survey. Program Overview. http:///
www.census.gov/svsd/www/cfsdat/cfsoverview.htm. Accessed March 11, 2010.
