Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration
Office of Planning, Environment, & Realty (HEP)


The decennial Census is a vital source of information on the commuting characteristics of the nation. For the last 40 years the Census has consistently collected information on:

These data allows us to track trends and identify changes in commuting behavior, to link demographic characteristics of households and workers to mode of travel to work,vehicle availability, and other related characteristics of U.S. commuters. We also obtain geographic flows of workers from place of residence to place of work.

The analysis in this report addresses the trends in the nation, the states, and in the 49 largest Metropolitan Statistical Areas (MSAs) which are those with a population of over one million residents. MSA- level analysis includes the Census years 1980, 1990 and 2000 (the geographic changes to metropolitan areas complicates the trends prior to 1980). Between 1980 and 1990 six metro areas were added to the largest group of those over one million in population. Between 1990 and 2000 ten areas were added to the largest group.

Sources and Limitations of the Data

All of the demographic and travel data presented in this report are from the U.S. decennial census, unless otherwise indicated. Even though the census collected these data, changes in methods, geography and coding in the decades between 1960 and 2000 may inhibit direct comparison of the data.

Changes in Geography

The United States Office of Management and Budget (OMB) defines metropolitan areas (MAs) according to published standards that are applied to Census Bureau data. The general concept of an MA is that of a core area containing a large population nucleus, together with adjacent communities having a high degree of economic and social integration with that core. MAs in this report are based on application of 1990 standards (which appeared in the Federal Register on March 30, 1990) to 1990 decennial census data and to subsequent Census Bureau population estimates and special census data. This report uses the June 30, 1999 definition of MAs (new definitions were published by OMB on June 3, 2003, but are not used in this report). A metropolitan area is called a Consolidated Metropolitan Statistical Area (CMSA) if it meets requirements of an MSA,has a population of 1 million or more, if the component areas are recognized as primary metropolitan statistical areas (PMSA), and if local opinion favors the designation. For example, the Washington, D.C. CMSA incorporates the Washington, D.C. PMSA, Baltimore, MD PMSA, and Hagerstown, MD PMSA.

Metro Area Definitions for all data (except 1990 Median income, and 1990 Median age)for the Census years 1980, 1990, and 2000 will use the June 1999 definitions of MSAs;therefore comparisons will be valid and straightforward. 1990 Median age and 1990 median income values are based on Census Bureau's published values using June 30,1993 definition of MSAs.

The 1990 data for MSAs were prepared using county level data from 1990 Summary File 3 files, and then aggregating the data for 1999 geographic definitions : Central Counties are the same as those defined in the 1990 Journey to Work Trends Report. That report included a designated central county for 39 metropolitan areas. For the ten MSAs that were not included in the 1990 report, ONE County was selected based on the primary downtown area.

For the New England area counties are not a basic geographic component. When possible we use NECMAs, or county-based metro areas, and the central county is designated as central county when the NECMA has a city designated as central city in it. Making the geography as comparable as possible is done on a case-by-case basis for the New England area.

For more information on the June 30, 1999 definition for MSA geography, please refer to the Census Bureau website at:

Mode to Work

Public transportation (transit) includes bus or trolley bus, streetcar or trolley car, subway or elevated, railroad, ferryboat, and taxicab.

Travel Time to Work

In the 1990 Census (including the CTPP data), the maximum travel time assigned to any worker was 99 minutes. Workers who reported travel times of 100 minutes or more were coded to 99 minutes in 1990. The maximum travel time was increased to 200 minutes for Census 2000, thus the 2000 data are more accurate because they include the actual value for these long trips. The impact of this coding change is that increases in travel time between 1990 and 2000 are somewhat over-stated. At the national level, the Census Bureau estimates that about 29% (0.9 minutes) of the 3.1-minute increase in average travel time is attributable to the coding change.

The state-wide difference between the average travel time for workers for 2000 for top codes of 99 minutes and 200 minutes is anywhere from 0.7 to 1.7 minutes (the State of Louisiana). Nationally, average travel time is 24.6 minutes using a top code of 99 minutes. That means the national average of reported travel time increased by 2.2 minutes from 1990, rather than 3.1 minutes.

The story of increasing commute time in the last decade remains the same, regardless of the issue of coding changes. From 1980 to 1990, 9 states and the District of Columbia showed a decrease in average travel time and of the remaining 41 states only 12 showed an increase in average travel time over one minute. Using the same top code in 2000 (99 minutes), every state showed an increase in travel time of more than one minute.

Number of jobs, Employed persons, and Workers at work

In examining decennial-census based counts of workers, it is important to understand definitional differences between workers and employed population and the differences between total employment (jobs) and workers-at-work. A general rule-of-thumb should be that total employment should be 7 to 9 percent higher than the Census 2000 count of workers- at- work. Two percent of the difference can be attributed to weekly absenteeism (see Item 2a), and six percent of the difference can be attributed to workers with multiple jobs (see Item 2c). These are general estimates based on national figures,and the exact measures for each MSA may be different.

  1. Employed persons versus Workers-at-work

    "Workers", as used in Journey-to-work and CTPP, refers to all those persons 16 years or older who were at-work in the reference week (including people in the Armed Forces).

    "Employed" is defined as all persons 16 years or older who were:

    1. At work (except in the Armed Forces).
    2. With a job but not at work for the whole week (due to illness, personal businees,vacation etc.)

    The Census Bureau considers the terms "employed" and "civilian employed" as exactlythe same. People who volunteered to work (without pay), and people who worked for the armed forces are excluded from "Employed".

    Exhibit 7.1 shows all persons 16 years of age and older, workers, total workers, civilian employed population, and people working in the armed forces for the nation from the decennial Census.

    Exhibit 7.1 Employed population versus Workers: 2000

    CategoryUnited States
    Total Population: 16 years or older217,168,077
    Total population in the labor force138,820,935
    Total Workers128,279,228
    Civilian Employed129,721,512
    Armed Forces1,152,137
    Civilian Employed + Armed Forces130,873,649
  2. Reconciling Total Employment (jobs) and Workers-at-work

    The decennial census based data for workers are derived from the long form question,"At what location did this person work LAST WEEK?"

    If the person worked at more than one location they are instructed to print where they worked most last week. Thus, these data are tagged to a particular reference week.People are not being asked their usual workplace location. Also, the Census asks for ONLY ONE job. People with multiple jobs can write in information about only one job on the Census form.

    There are three main adjustments that are needed to make TOTAL EMPLOYMENT (JOBS) data comparable to census workers-at-work data:

    1. Weekly absenteeism adjustments

      The Census reports only workers (full-time or part-time) who worked in any time during the week prior to the survey. An adjustment must be made to reflect workers who may not work every day or who may not go to work on an occasional day due to illness, vacation, personal business or other reasons. The FHWA publication "Transportation Planner's Handbook on Conversion Factors for the Use of Census Data" notes that studies by local agencies suggest that the typical WEEKDAY absenteeism factor is in the range of 15-20 percent.

      One way to calculate absenteeism for you area is to compare the values for "Civilian Employed" + "Armed Forces" with "Total Workers."

      Absenteeism factor = (Civilian Employed + Armed Forces) - Total Workers]*100 / Total Workers

      Using the values from Table 1, the national average for WEEKLY absenteeism is about 2 percent. This procedure can be used to calculate weekly absenteeism factors for all geographies (eg: state, county, place, tract, or block group).

    2. Seasonal fluctuations in employment adjustments

      Both the labor force, and employment opportunities fluctuate with different seasons. The decennial census does not measure any "typical" week in the year-the reference week may be anytime between March-April 2000.

    3. Multiple jobholding adjustments

      In May 2001, 7.8 million persons worked at multiple jobs in the United States, a figure representing 5.7 percent of all workers1. The percent of workers holding multiple jobs varies based on geographic location, cost of living etc.

Updated: 10/20/2015
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000