The pattern of commutes in American is affected by worker characteristics, the supply and location of jobs and housing, and the time and convenience of various modes of commuting. This report explores the changes in these conditions and the changes in commute characteristics of U.S. workers by looking at the last forty years of data from the U.S. Decennial Census. First, a national overview is presented and then the metropolitan areas are analyzed.
The work trip is often the longest distance we travel, and our work location provides a sphere of activity that anchors some of our travel, either in stops we make between home and work or in trips we make around our workplace. The commute trip is so important in understanding people's daily travel that information about the commute has been included in the U.S. decennial census. Therefore, we have detailed demographic and geographic information on US residents' travel to work over a long period of time.
The U.S. Census started including questions about commuting in 1960, so with the 2000 Census we have 40 years of decennial data. Some of the changes that impact commuting trends are:
In 1960, over half (52 percent) of the family households consisted of married couple with children. In 2000, nuclear families with children account for just over a third (35 percent) of U.S. family households-eclipsed for the first time in history by single-person households.
The most dramatic change in the workforce is the inclusion of women--61 percent of women work today compared to just 38 percent in 1960. The shift from single-earner to dual-earner families fueled the rise in household income, and household vehicle ownership, and such phenomenon as the decline in multi-occupant vehicles and the rise in trip chaining.
However, the huge increase in workers in the U.S. may be near an end as the baby boomers age into retirement years. Between 1960 and 2000, the U.S. added 63.6 million workers, 1.2 new workers for every new person. In the most recent decade, 1990 - 2000, the number of workers being added to the labor force was less than previous decades (See Exhibit B).
The U.S. is an aging society-the baby boomers will begin to reach retirement age by 2010. Baby boomers may delay retirement just as they have delayed other major life stages (marriage, children, etc.), but eventually older workers will stop working.
Immigration may be a source to fill the worker void left by retiring baby-boomers. The largest MSAs currently account for 81.4 percent of the total foreign-born population of the U.S. Policy decisions determine the amount of allowable immigration each year, but if trends continue foreign-born people will be a large factor in population and worker growth in the U.S. New immigrants to the U.S. are less dependent on auto travel than native-born people, but as they stay longer, their travel becomes more Americanized.
MSAs continue to grow in both area and population. The land area of the major metropolitan areas grew as fringe counties were adopted into the metro area, and both jobs and housing have grown outside the traditional urban centers (Exhibit C).
In 1960, there were 34 metro areas of over 1 million residents; in 1990, there were 39 areas with one million residents or more; in 2000, there were 49 large MSAs. Looking at the same metropolitan areas in 1960 and 2000 shows the growth of population and workers in suburban counties by far outpaced the growth in central counties.
On a national level, the decentralization of workers and jobs is taking place both relatively and absolutely to a much greater degree in the South and the West1-areas that are high growth with a lot of migrants and immigrants, added workers, and new housing development.
1 2000, TCRP Report 74, Transportation Research Board p.3
The 49 major metropolitan areas in the U.S. are the focus of this report and account for:
Average household size went from 3.3 in 1960 to 2.6 in 2000, a decline of over one-fifth. At the same time, vehicles per household rose from just over 1.0 to about 1.7, an increase of almost two-thirds (Exhibit D).
The census shows that in 2000 three-quarters of commuters drove alone to work(75.7 percent), followed by carpooling (12.2 percent), transit (4.7 percent), work at home(3.3 percent) and walk (2.9 percent).
In 1960, 41 million commuters were in private vehicles; by 2000, 113 million workers commuted by private vehicle, nearly three times as many (See Exhibit E). Between 1990 and 2000, drove alone continued to increase, as carpools continued to drop. By 2000, the average vehicle occupancy for the commute trip was 1.08.
The total number of workers increased in the 1990s but the number of workers using transit stayed about the same (6 million workers commute by transit). Therefore the proportion of commuters by transit, or the mode share for transit, has slightly declined.
African-Americans and Hispanics are more likely to take transit for commuting even for households where one or two vehicles are available for use. This may be due to the location of black and Hispanic households in central cities and older suburbs that have greater transit accessibility.
Work at home increased in the 1990s, and the nature of jobs usually conducted at home shifted. In the 1960s many people who worked at home were agricultural-based (farm) workers or professionals with home-based practices, in the 1990s the shift has been toward telecommuters who may work-at-home and in some other location.
The average commute increased by 2.1 minutes2 between 1990 and 2000. This is much higher increase than the 40-second increase from 1980 to 1990. By examining the ravel time distributions, we see a continued shift toward longer commutes.
2 Census reports will show an increase of 3.1 minutes between 1990 and 2000, however, changes in coding procedures between 1990 and 2000 have created confounding problems in direct comparisons. In 1990, travel time of 100 minutes or more was coded as 99 minutes, whereas in 2000 the top-code was 200 minutes. This coding change results in more accurate results in 2000. The value of 2.1 was obtained by recalculating Census 2000 data using the same topcoding as 1990.
In 2000, 14 percent of workers traveled more than 45 minutes compared to 12 percent in 1990, and 29 percent commute less than 15 minutes, compared to 31 percent in 1990. Forty percent of the commuters in large metro areas travel over 30 minutes to work, oneway, on an average day.
The pressure of time is a major factor in the travel choices people make. In 2000, more workers are driving alone, more families are living and working in the suburbs and traveling on the highway system for part of their commute, and more workers are commuting over one hour to and from their jobs on an average day.
Changes in family structure, workforce characteristics, and vehicle availability have affected mode choice throughout the 70s and 80s. Over the years as automobiles became affordable and convenient as a means of transportation, more and more people became drivers. Commuters may have shifted to POV and then drove alone to save travel time as jobs and homes became more dispersed.
The 2000 Census shows large increases in travel time in all metropolitan areas, which suggests that workers may consider other modes if trave l time can be shortened, may shift their work times (leading to peak-spreading), or may try or increase telecommuting.