Commuting In Amercia: Part 2


Published on

2006 Pisarski Report on Commuting in America

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Commuting In Amercia: Part 2

  1. 1. PART 2 COMMUTERS IN THE NINETIES Population and Worker Growth 3SOME SURPRISESIt is reasonable to expect very few surprises in popu- Why does this shift in growth matter for com-lation growth trends. The factors to be considered muting? Had the increase come about as the resultare relatively clear cut and demographers understand of increased birth rates it would have had a very lim-them well. In most cases, the changes in those fac- ited impact on commuting, at least for another 18tors—birth rates, death rates, population migra- to 20 years. Had the increase come from unexpectedtion—shift at something like a glacial pace. What increases in longevity it would have had almost noa disconcerting surprise then when Census 2000 effect. Because the increase was the result of immi-results revealed about 6 million people more than gration, it indicates almost an instant increase inexpected; instead of 275 million, the count was over commuters. Most of the immigrant arrivals are of281 million. working age and quickly join the labor force. The A very simple but roughly reliable approach to foreign-born population arriving in the 1990s wasunderstanding the nation’s population growth and its particularly concentrated in the 25-45 age group.7projections into the future that served well for the last Only 29% of the native population was in thishalf of the past century was that about 25 million in group but 44% of immigrants were in that range.population were added each decade from 1950-1990, Thus, a shift in population due to immigration canand about 25 million per decade were expected, as have an immediate impact on the number of work- FIGURE 2-1 Population Change, 1950-2050 30 Over 80% Population change (millions) 25 of immigrants 20 arriving in the 15 5 years prior to the census were 10 in the 16-65 age 5 group, the main 0 working age 2040-2050 2030-2040 1970-1980 1990-2000 2010-2020 1950-1960 1980-1990 1960-1970 2020-2030 2000-2010 group, with very few older than 65. Actual Projectedshown in Figure 2-1, to be added out to 2050. Thus, ers and commuting. In this case, the size of the agethe expectation was for 100 years of very stable, pre- group from 16-65 years old, the main working agedictable growth.6 Instead of about 25 million in the group, reached a level in 2000 that had not beenperiod from 1990-2000, however, the census showed expected until increase of almost 33 million. The growth rate ofover 13% for the decade was well beyond the rate of 6 Census Bureau, U.S. population trends as of September 13, 1999.less than 10% that had been expected. The simple 7 Throughout this report, numbers in a range go to, but do notanswer to what happened is immigration. include, the ending number in the range. C O M M U T I N G I N A M E RI C A III | 15
  2. 2. centage terms, was expected (Commuting in America TABLE 2-1 Worker and Population Increase, 1950-2000 II earmarked 1990 as the turning point that signaled Total Workers Worker Increase Worker Increase Population the end of the worker boom). However, many are Year (Millions) (Millions) (%) Increase (%) hard-pressed to understand the sharper-than-expected 1950 58.9 N/A N/A N/A declines, particularly given the larger-than-expected increases in immigrants who largely were of working 1960 65.8 6.9 11.7 18.5 age. This situation, critical to commuting, will be 1970 78.6 12.8 19.5 13.3 discussed later in this chapter. 1980 96.7 18.1 23.0 11.4 Another point of note in the table is that the 1990 115.1 18.4 19.2 9.7 30-year decline in the rate of population growth as the baby boom waned took a sharp reversal in 2000 2000 128.3 13.2 11.5 13.2 and returned to the growth rates of the 1970s. Since Overall Change 69.4 117.8 86.0 2000, worker growth has been quite limited given the events of 9/11 and the recession that followed. Between 2000 and 2003, only 1.4 million work- TABLE 2-2 Workers Age 65 and Older ers were added, about a 1.1% increase, according to the American Community Survey (ACS). In Age Group 1990 2000 Change 2004, Bureau of Labor Statistics (BLS) employment 65- 75 2,947,744 3,305,563 357,819 statistics showed worker growth surged, adding 1.7 75+ 549,718 942,575 392,857 million from December 2003 to December 2004, but were still an anemic 2.4% growth since 2000. 65+ 3,497,462 4,248,138 750,676 The worker/population ratio of 62.4 was the same at the end of 2004 as at the start of the year. A key concern for commuting is the rate of Perhaps the best example of the effects of immi- growth within the work-age segment of the popula- gration is the pre–Census 2000 projections for the tion (ages 16-65). BLS has traditionally used 16 number of 18-year-olds, a useful indicator of new or over as their base measure of the labor force age labor force arrivals. According to projections, this group, but in the coming years that approach may age group, which had dropped below 4 million after prove misleading, as the baby boom ages and the 1990, would not reach 4 million again until 2008; share of population over 65 surges. The key point, in fact, 4 million was reached in 2000. and one to monitor carefully in the future, is that in 2000 only 3.3% of workers were age 65 and older, not much greater than the 3% registered for 1990. PARALLEL LABOR FORCE TRENDS The population at work among those 65 and older The boom in workers can be measured by the period rose by roughly 750,000 from 3.5 million in 1990 to from 1970-1990 when the rates of increase in workers 4.25 million in 2000, about half of the growth com- exceeded population increase, as seen in Table 2-1. ing from those age 75 and older as shown in Table Note also that in the overall period from 1950-2000, 2-2. The number of workers age 65 and older rose by workers in the population more than doubled. All of over 21% in the period while the population in that this said, the sharp decline in the number and the rate group only rose about 12%. As that group’s share of of growth in workers in the 1990s comes as another the population increases sharply after 2010, a key demographic surprise. Some decline, certainly in per- question for commuting will be the extent to which FIGURE 2-2 Workers by Age Group 50 Worker Nonworker 40 Workers (millions) 30 20 10 0 16-25 25-35 35-45 45-55 55-65 65-75 75+ Age group16 | COMMUTING IN AMERICA III
  3. 3. TABLE 2-3 Population Growth Rates by Age Group, 1990-2000 Age Group 1990 (Millions) 2000 (Millions) Change (%) 1990 Distribution (%) 2000 Distribution (%) < 16 56.9 64.3 13.0 22.9 22.8 16 - 65 160.6 182.2 13.4 64.6 64.7 65+ 31.2 35.0 12.1 12.5 12.4 All 248.7 281.4 13.1 100.0 100.0 Note: Data include both household and general quarters population.8 Figure 2-3 does a good job FIGURE 2-3 Population and Labor Force Trends, of explaining the effects of the 1960-2000 baby boom on commuting. It shows how divergent the rate 35 of growth of the 16-65 age 30 group has been over the last 25 half of the twentieth century, Growth (percent) The number only moving in synch with 20 of workers age the total population since 15 1990. But note that a new 65 and older 10 factor is significant: the civil- rose by more ian labor force did not grow 5 in tandem with the 16-65 age than 21% in the 0 group as it has since 1970. period while the 1960 1970 1980 1990 2000 If these data are expanded population in that to examine the patterns by gender, as in Figure 2-4, it group only rose Total population Age 16-65 Civilian labor force is clear that both male and about 12%. As female labor force growth that group’s share rates have declined somewhatpersons in that age group continue to work. and that female growth rates show a sharper decline. of the population Another point is the percentage of these age This figure also displays the prominent role played increases sharplygroups that are workers. Figure 2-2 shows how this by women joining the labor force in extraordinary after 2010, astatistic plays out within the age groups. About 55% numbers over the period.of those in the 16-25 age group are workers, rising to Another perspective on the data is shown in Fig- key question for75% in the main working years, until the 55-65 age ure 2-5, which starkly depicts some of the dramatic commuting willgroup when it again drops to 55%. From age 65-75, shifts of the era. Perhaps most fascinating is the be the extent toonly about 18% or 19% of the population works, decade from 1970-1980, in which total population,falling to 6% for those 75 and older. These values the 16-65 age group, and the civilian labor force all which persons invary distinctly by gender. The peak for men is 84% grew by almost exactly the same amount. The period that age groupin the 35-45 age group, whereas the peak for women from 1980-1990 shows a sharp drop in the increase continue to only 70% and it occurs in the 35 through 54 age of those 16-65 as the tail end of the baby boomergroups. In the 65-75 age group, it is 24% men versus group arrived. Finally, the 1990-2000 group shows14% women and 10% men versus 4% women for the small growth in labor force relative to the growththose 75 and older. So, despite the fact that there are in population. In that sense, the period appearsmany more women than men 65 and older, a greater reminiscent of 1950-1960, when most of the babynumber of men in the over-65 age group work, boomers were first born.roughly 2.5 million men to 1.7 million women. As can be seen from Table 2-3, the rates ofchange in numbers were about the same for all ofthe main age segments of interest here, and the dis-tribution of the population between the age groupsremained effectively identical for 2000.8 In tables throughout this report, numbers may not add due torounding. C OMMUT I NG I N A MER I CA III | 17
  4. 4. FIGURE 2-4 Male–Female Labor Force Trends, Looking Beyond the Numbers— 1960-2000 The Group Quarters Population 50 Most of the U.S. population is organized into house- holds, and further classified into family and nonfamily 45 households. There is a segment of the population, however, that is not household based. These individuals are gener- 40 ally referred to as being in group quarters. The Census 35 Bureau recognizes two general categories of people in group quarters: 1.) the institutionalized population, which Growth (percent) 30 includes people under formally authorized, supervised care or custody in institutions (such as correctional institutions, 25 nursing homes, and juvenile institutions) at the time of 20 enumeration and 2.) the noninstitutionalized population, which includes all people who live in group quarters other 15 than institutions (such as college dormitories, military quar- ters, and group homes). From a commuting point of view, 10 only that segment of the group quarters population not in 5 institutionalized settings is of interest and many of these individuals—who may be college students, members of 0 the military, farm camp workers, or members of religious 1960 1970 1980 1990 2000 orders—often work on the same site where they live, so their work travel has limited impact on others. The total group quarters population in 2000 numbered Population age 16-65 Total civilian labor force about 7.8 million, just below 3% of the population, with a Women in civilian labor force Men in civilian labor force higher share for men than women. Of this group, it is the noninstitutional population of about 3.7 million (2 million men and 1.7 million women) that has the potential to be FIGURE 2-5 Population and Labor Force Increase, commuters. Of these, about 2 million are college students, 1950-2000 about 600,000 are in the military, and the remainder are in other group arrangements. Generally, this report addresses the travel behavior 2000 of the 273.6 million members of the household popula- tion. The household population includes both families 1990 and nonfamilies. An example of a nonfamily household is several unrelated people sharing an apartment where there are common kitchen and bath facilities. This would be 1980 considered a nonfamily household and not a group quarters arrangement. 1970 1960 BABY BOOM WORKERS APPROACHING RETIREMENT 0 5 10 15 20 25 30 35 Even with the population growth surge from Increase (millions) immigrants, the strong impact of the baby boom generation still remains very clear. Figure 2-6 shows Total population Age 16-65 Civilian labor force the sharp shifts in net population change by 5-year age group. The leading edge of the baby boom is very clear at ages 55-60 and the trailing edge at ages 35-40. Something of a surprise is that the 65-70 age group actually registers a decline in population, lag- ging previous cohorts, as a result of the lack of births during the Depression Era.18 | COMMUTIN G IN AM E RICA III
  5. 5. FIGURE 2-6 Net Population Change, 1990-2000 7 6 5Population change (millions) 4 3 2 1 0 -1 -2 -3 <5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 50-55 55-60 60-65 65-70 70-75 75+ Age group Up to the present, the labor force effects of TABLE 2-4 Nonworker and Worker Population by Age Group these changes have been mild but will start to shift Age Group Nonworkers Workers Total Population Workers (%) sharply between 2005 and 2010. The share of those < 16 64,113,087 0 64,113,087 0 of working age has remained stable at just below 16-25 14,188,649 17,810,367 31,999,016 14.03 65% (64% for women and 65% for men) for the last decade. By 2010, however, the first of the baby 25-35 9,581,498 28,890,182 38,471,680 22.76 boomers will reach 65, and there will be a sharp 35-45 10,478,404 34,557,990 45,036,394 27.22 rise in the 65-70 age group. According to interim 45-55 8,966,440 28,262,886 37,229,326 22.27 Census Bureau projections prepared in the summer 55-65 10,657,419 13,167,192 23,824,611 10.37 of 2004, the working age share drops sharply after 2010 as the over-65 group rises from 13% to 16% 65-75 14,549,867 3,305,563 17,855,430 2.60 in 2020 and to 20% by 2030. 75+ 14,165,277 942,575 15,107,852 0.74 Table 2-4 shows the share of population in Total 146,700,641 126,936,755 273,637,396 100.00 the worker population by age group for 2000. These patterns will be key for monitoring future worker populations. Small shifts in the percentages FIGURE 2-7 Male–Female Labor Force Increase, can make for great swings in the labor force. Of 1960-2000 particular interest will be the 10.37% rate among workers ages 55-65 and the 2.60% rate for those age 65-75. 16 Labor force increase (millions) 14 12 10 MALE–FEMALE LABOR FORCE TRENDS 8 Figure 2-7 shows the absolutely dominant role 6 women have played in labor force growth over the 4 last 50 years; 60% of labor force increase in the 2 period can be attributed to women. As a result, 0 the female share of the labor force rose from 28% 1960 1970 1980 1990 2000 in 1950 to almost 47% of all workers. In the years since 2000 through 2003, it has stabilized at just Male Female above 46% according to both the annual ACS and the BLS employment statistics. C O M M U T I N G I N A M E RI C A III | 19
  6. 6. RACE AND ETHNICITY IN WORKER TRENDS Among the other groups with lower shares of It is clear that the 16-65 population group, consti- working-age population are Alaskan Natives (61.1%) tuting about 65% of the total population, is critical and the newly reported group of those belonging to for understanding commuting. It is very significant the category of two or more races (57.4%), both due that the share for this group varies rather consider- to very large shares of population under age 16. ably by race and ethnicity.9 The main patterns are Another important social group to consider is shown in Table 2-5. Note that the Asian popula- the new immigrant population—of whatever race or tion represents the largest working-age group in ethnicity. Among those who arrived in the United percentage terms, primarily because of a relatively States within the 5 years just prior to Census 2000, much smaller older population. The African-Ameri- 80.5% were in the 16-65 age group (81.5% for can population has only a slightly smaller share of men), indicating a very strong orientation to work- working-age population but their younger and older ing age, and less than 3% were age 65 and older. populations are sharply skewed from the average. Recent BLS data from the Current Population The Hispanic population has an even larger young Survey (CPS)10 provides valuable input regarding population and an even smaller older population in the role of the foreign born in the labor force. As percentage terms. Given their importance in the new of 2003, BLS counted 21.1 million people, 67.4% worker mix, the percentages of the Hispanic popula- of the foreign born, as in the labor force, somewhat tion are shown in Figure 2-8. greater than the 66.1% of the native-born popula- tion. A major demographic distinction between 9 To preserve the accuracy of the original data, terms used to denote the foreign-born population and the native-born race or ethnicity throughout this report appear as they did in population is the role of men in the labor force as their original data source. Census 2000 used the following racial shown in Table 2-6. While the participation rates categories: American Indian or Alaska Native; Asian; Black or African-American; Native Hawaiian or Other Pacific Islander; White; in general for both groups are roughly the same, and Some Other Race. Ethnicity choices for Census 2000 included: the differences between men and women within Hispanic or Latino and Not Hispanic or Latino; for the sake of the groups speak to a significant cultural distinc- brevity, ethnicities are referred to here as Hispanic or non-Hispanic. tion. Foreign-born men have a participation rate of Additional information on the composition of these categories can be found in “Summary File 4, 2000 Census of Population and Housing: over 80% contrasted to 72% for the native born; Technical Documentation,” Census Bureau, March 2005. among women it is almost the opposite situation with foreign-born women at 54%, 6 percentage points below native-born women. As a result, TABLE 2-5 Population by Major Age Group foreign-born men are 16% of the total national Race/Ethnicity Age Group (%) labor force whereas foreign-born women are slightly more than 12%. Part of the distinction < 16 16-65 65+ may be in that the foreign-born labor force is a White, non-Hispanic 20.6 64.9 14.5 significantly younger group as noted earlier from Black, non-Hispanic 29.2 62.7 8.2 the census data. BLS considers the prime work Asian, non-Hispanic 21.6 70.7 7.7 years to be the ages of 25-55. This age grouping accounts for almost 77% of the foreign born but Hispanic 31.9 63.3 4.8 only 69% of the native born. As a result, the for- All 23.4 64.5 12.0 eign born constitute close to 16% of the labor force age group that is in their prime work years. Another important facet of the group is its educa- FIGURE 2-8 Hispanic Share of Population tional makeup. Among the foreign born, nearly 30% 20 over age 25 had not completed high school, contrasted 17% to only 7% of the native born; however, the college Percent of total population 15 graduation rates were very similar, 31% and 32% 12% 13% respectively. This indicates a strong bi-modal character- 11% istic regarding education among the foreign born. 10 A quick way to summarize the linkage between 5% population, households, and workers is shown in 5 Table 2-7, which identifies the number of workers per household, a key component of the relationship 0 among these three elements. The central item of <16 16+ 16-65 65+ All Age group 10 Bureau of Labor Statistics, Current Population Survey. “Labor Force Characteristics of Foreign-Born Workers in 2003,” December 1, 2004, U.S. Department of Labor, Washington, D.C.20 | COMMUTING IN AMERICA III
  7. 7. TABLE 2-6 Role of the Foreign Born in the Labor Force, 2003 Foreign Born Native Born Gender Population Civilian Labor Force Participation Rate Population Civilian Labor Force Participation Rate (Thousands) (Thousands) (%) (Thousands) (Thousands) (%) Male 15,669 12,634 80.6 90,766 65,603 72.3 Female 15,662 8,482 54.2 99,072 59,790 60.4 All 31,331 21,117 67.4 189,837 125,393 66.1 Source: BLS, Current Population Survey, 2003. TABLE 2-7 Households and Population TABLE 2-8 Distribution of Hours by Workers in Household Worked per Week, 1999 Roughly 70% (Millions) Hours/Week Workers Percent of the workers Workers/ 1 through 8 1,814,996 1.47 Households Population Workers in America live Household 9 through 24 10,687,829 8.65 0 27.8 50.7 0 in households 25 through 32 9,370,184 7.59 1 38.9 91.0 38.9 33 through 40 63,872,119 51.71 with at least one 2 31.6 99.5 63.9 3+ 7.2 32.5 24.0 41 through 48 12,824,312 10.38 other worker. This Unrounded Total 105.4 273.6 126.7 49 through 56 15,132,230 12.25 affects their op- 57 through 64 6,174,702 5.00 tions and choicesgreat importance in this table is that roughly 70% 65 through 72 2,237,437 1.81 in commutingof the workers in America live in households with 73 through 80 951,102 0.77 behavior in manyat least one or more other workers. This affects 81 through 99 461,649 0.37 ways.their options and choices in commuting behavior Total 126,936,755* 100.00in many ways. Note that 24 million workers live *Includes 3,410,195 workers recorded as not in the universe, did not workin households of three or more workers. This is in 1999, or under age 16.particularly significant in the interaction withimmigrant status. Although those workers who hadbeen living outside the United States 5 years before “regular” 33- through 40-hour week with roughlyCensus 2000 constituted only 2.8% of workers, 20% working less and 30% more than that as shownthey constituted almost 5% of workers living in in Table 2-8. A further distribution by worker genderhouseholds of three or more workers. shows that men tend to work more than average and When we think of work, there is a tendency to women less, as shown in Figure 2-9. These resultsassume that the 40-hour week is standard. Only a are supported by the recent BLS American Timebit more than half the worker population works the Use Survey (ATUS), which shows men working an FIGURE 2-9 Workers by Hours Worked and Gender, 1999 35 30 Male Female 25 Workers (millions) 20 15 10 5 0 41 through 55 <15 15 through 20 35 through 40 21 through 34 55+ Hours worked/week Source: CTPP. C O M M U T I N G I N A M E RI C A III | 21
  8. 8. than the CPS. Given the nature of the surveys, and TABLE 2-9 Average Hours Worked the fact that the CPS, which is the national source of per Day, 2003 closely watched monthly unemployment statistics, is Job Status All Male Female designed specifically to obtain work-related data with All jobs 7.59 8.01 7.06 more questions and greater probing by interviewers to get complete and accurate information, this is not Full time 8.09 8.33 7.72 surprising. What was surprising was the degree of dif- Part time 5.40 5.74 5.19 ference between the two surveys observed this time, Source: BLS, American Time Use Survey, 2003. greater than any time since the 1950 census. The percentage differences in each decade had declined over time, but in 2000 the difference was three times average of about 8 hours a day, about an hour per the previous census and was exceeded only by the day more than women at 7.06 hours, averaging out 1950 difference. The long-term pattern is shown in In each decade to 7.6 hours per day for the entire workforce. Part of Table 2-10. the gap between the explanation for the difference is that women tend The findings of one internal review11 of the the census and to work more part-time hours than men. Table 2-9 differences, based on the April 2000 CPS results, provides a summary of the ATUS findings. presented the following observations: other sources ofemployment statis- 1. The 2000 decennial census estimate of the num-tics, primarily the ABOUT THE SURPRISES IN ber of employed people, 129.7 million, was about 7.2 million, approximately 5%, lower than theBLS, had declined WORKER GROWTH April 2000 CPS estimate of 136.9 million. Earlier discussion identified that the decennial census over time, but in 2. The 2000 decennial census estimate of the num- had shown a surprising lack of growth in workers in 2000 the differ- ber of unemployed people, 7.9 million, was about the period from1990-2000, both in percentage and 2.7 million, or over 50%, higher than the CPS ence was three absolute terms. A decline in the percentage growth estimate of 5.2 million. rate was expected, but not to the levels actuallytimes the previous 3. The “civilian labor force” is the sum of the found in the census. Preliminary examination of thepopulation census employed and unemployed values, and therefore 2000 decennial statistics by the Census Bureau has the disparities balanced somewhat with the differ- and was the big- shown that the disparity between the census results ence of the decennial census at 137.7 million, at and other surveys also conducted by the Bureau that gest gap since about 3.1%, or 4.5 million below the CPS value report labor-force-related statistics (most notably, the 1950. of 142.2 million. BLS Monthly Employment Statistics) were substan- 4. The decennial census also showed disparities in all tial. The main household survey that supports the of the usual rate measures that accompany these BLS reporting system is called the Current Popula- statistics. tion Survey (CPS), which provides 60,000 observa- tions per month. In addition, an establishment-based Because the interest here is only in the worker survey of 160,000 businesses and government agen- side of the equation, for the purposes of commuting cies also is conducted to get information from the analyses these disparities do not balance out. The employer side. Each of these surveys reports slightly number of workers observed by the decennial census different amounts of employment. Over the decades, who worked the previous week was 128.3 million. the decennial census typically reported fewer workers This varies slightly from the number of civilian employed of 129.7 million observed by the census. TABLE 2-10 Civilian Labor Force Comparison The small difference is attributable to those people with jobs but who were not at work in the survey Decennial Census Absolute week for various reasons (e.g., sickness, vacation, job Year CPS (Millions) Difference (%) (Millions) Difference (Millions) stoppage, weather, etc.). Since BLS does not produce 1950 58.2 61.5 3.3 5.67 an estimate of those employed but not at work, an 1960 67.5 69.1 1.6 2.37 adjustment for this difference would place CPS 1970 80.1 82.0 1.9 2.37 employed and at-work estimates at about 135.4 mil- 1980 104.4 105.6 1.2 1.15 lion, which is roughly 7 million (5.5%) more than 1990 123.5 124.8 1.3 1.05 reported by the decennial census. 2000 137.7 142.2 4.5 3.27 One problem that led to sharp disparities were Source: BLS, Current Population Survey, December 2004. the anomalies in the group quarters statistics in the Census Bureau, “Comparing Employment, Income, and Poverty: 11 Census 2000 and the Current Population Survey,” September 2003, U.S. Department of Commerce, Washington, D.C. 22 | COMMUTING IN AMERICA III
  9. 9. misunderstanding because TABLE 2-11 Comparison of E/P Ratios the CPS-based BLS report E/P Ratio for a year is an average of 12 monthly reports, each of which Variable Census 2000 April 2000 CPS Difference reports one week in the month. All 61.2 64.6 3.4 The census ostensibly reports In the case of the Age 25+ 62.2 65.7 3.5 employment for the week African-American Age 65+ 13.1 12.5 –.6 previous to the reporting date and Hispanic occurring on April 1, but in Male 68.1 71.8 3.7 fact census forms are collected populations, the Female 54.9 57.9 3.0 in April through July and so gaps between White 62.4 65.1 2.7 the employment statistics repre- the decennial sent a composite of that period. Black 55.5 61.4 5.9 census and the In that degree, they cannot Hispanic 56.4 66.1 9.7 fully be comparable with, for BLS, as measured example, the April employment by employment/decennial census as a result of very odd reporting by reports from BLS used here for quarters members, notably college students. Another notable difference in the surveys is that population ratioThis resulted in vast overstatements of the college in comparison to the census, the CPS observes a statistics, wereunemployed on the order of one-half million. One continued higher rate of labor force growth (13.9% greater than foreffect of this was dramatic unemployment rates in contrasted to 11.5% for the census) and a lower ratecollege towns. When the group quarters population of population growth (12.3% contrasted to 13.2% the population inis excluded from the analysis, the values of the two for the census). general.surveys are closer together. For example, just as aresult of dropping out group quarters, the decen-nial unemployment rate drops from 5.8% to 5.2%, ADJUSTMENTS TO THE 2000bringing it closer to the CPS. Because the focus of DECENNIAL CENSUS Although nationalthis work is on the household-based population, this It may be desirable to consider adjustments to the levels are stable,shift is a positive development. 2000 decennial census numbers in some cases in The disparities do seem to be relatively stable by the volatility in metropolitan areas or states where disparities seemdemographic groups in that there is some unifor- great based on local data; at the national level it does individual statesmity in the differences. However, differences for not seem appropriate, given the need to maintain can be great. Inmen, women, and age were more uniform with the comparability within and between national sources.notable difference that those over 65 were closer no state is the In cases where it seems significant, note will be madethan other age groups. In the case of the African- of these potential differences throughout this text. Current Popula-American and Hispanic populations, there tended to It must be remembered that the main population tion Survey of thebe greater gaps between the decennial census and the benchmark control of the CPS is the decennial cen-CPS as measured by employment/population (E/P) BLS lower than sus. Since 2000, the CPS has been adjusted severalratio statistics and shown in Table 2-11. The E/P times to bring it into alignment with the decennial the census. Inis a ratio of civilian employed population to total benchmark and with population estimates produced 32 states, the twononinstitutional population over age 16. Another each year since. This does not affect the E/P ratioimportant facet of the differences is that although estimates were observed in the CPS but rather the population tothe national levels are as shown, the volatility in which it is applied. Since the end of the 2001 reces- not statisticallyindividual states can be great. In no state is the CPS sion, the trend has been for the CPS to show a rise significant in theirlower than the census. In 32 states, the two estimates in employment as a function of population increases,were not statistically significant in their difference. difference. although the E/P ratio itself has declined. An impor-Some states, such as California (1.3 million fewer tant facet of these considerations is that the new ACSemployed), Florida (525,000 fewer), and Texas process is now intended to replace the decennial cen-(716,000 fewer), however, saw substantial scale sus as an annual survey of population characteristics.differences in total employment. The significant This will mean that both the CPS and ACS shouldundercount of Hispanics may have affected these be in closer alignment because they are both funda-states particularly. mentally benchmarked to the same sources. There are many facets to the process of compari-son between the surveys. Attempting to comparethe data for common years can lead to a degree of C O M M U T I N G I N A M E RI C A III | 23
  10. 10. 4 Population and Household Trends GEOGRAPHIC DISTRIBUTION OF GROWTH Regional Growth Chapter 3 noted that the 2000 census delivered An important facet of national population growth, something of a surprise in terms of the total popu- critical to commuting, has been the regional popula- lation of the nation, counting at least 6 million tion shift that has been occurring at least since the end more people than expected. This chapter elaborates, of World War II. Figure 2-10 shows that together the describing the geographic distribution of the popula- Northeast and Midwest (as pictured on Figure 1-8) tion observed in the census and its demographic gained only a little more than one-quarter of the last characteristics. The many facets of population and half century’s population growth, with the West gain- household change as they bear on commuting and ing one-third and the South the remainder. Continu- its attributes now and in the future are discussed. ing this trend in the last decennial period, the South Regional and West each increased their share of the national FIGURE 2-10 Regional Shares of population by 1 percentage point as the Northeast population National Growth, and the Midwest each gave up 1 percentage point—ashares continue 1950-2000 pattern that has repeated roughly each decade since their strong 1960. About 45% of growth from 1990-2000 went to 11% the South and another 32% to the West. When com-pattern of sharp pared to the 1980s, this actually represents a slowing growth in the 33% of the share of growth going to those regions. In the 15%South and West first half of the 1980s, the South and West obtained 94% of the nation’s growth, dropped to about 83% in and stability or the later half of the decade, and arrived at about thedecline in share present rate early in the 1990s. The long-term decadein the Northeast trend is shown in Figure 2-11. Note that by Census 2000, the West surpassed the Northeast and closed and Midwest on the Midwest. By 2002, the West surpassed the up through 41% Midwest. The South continues to expand its lead and available data reached 100 million by the end of 2000. Northeast Midwest South West Population growth since 2000 has basically as of 2004. continued this trend. From July 2003 to July 2004 all 10 of the fastest growing states were in the FIGURE 2-11 Regional Population Growth, 1950-2000 South or West and those 10 accounted for half the national growth. The share of population change 120 since 2000 has roughly held constant for the Northeast and the South but the Midwest dropped Population growth (millions) 100 from a 14% growth share in the 1990s to less than 80 11% in the first years of the new decade and all of that lost share was gained by the West rising from 60 32% of the growth in the 1990s to almost 35% in 40 the new decade. These trends clearly suggest that the Northeast and Midwest continue to lose share 20 while the South and West gain. 0 The national population was estimated by the 1950 1960 1970 1980 1990 2000 Census Population Division to be slightly more than 293.5 million in July 2004, an addition of approxi- mately 11.5 million people since the decennial cen- Northeast Midwest South West sus. This represents a growth rate of almost exactly24 | COMMUTIN G IN AM E RICA III
  11. 11. the decade. Figure 2-12 shows the estimated total FIGURE 2-12 Regional Population, 2004 amount and the share of population by region as of July 2004. Long-term projections out to 2030 place 67,409,440 55,401,511 88% of the nation’s population growth in the South 23% 19% and West.12 State Growth Current In the decennial period from 1990-2000, all states population has except Washington, D.C. (which, for the sake of com- pleteness, is reported as a state by the Census Bureau) been growing gained in population. At the extremes, California at about 1% per added more than 4 million and North Dakota gained year, putting the 105,114,601 65,729,852 less than 4,000. This was a very positive change from 36% 22% United States at past decennial trends where there had been significant Northeast Midwest South West loses among several states. More recently, the growth over an estimated rates for the period from 2000-2004 were still sharply 296 million in Note: ACS data for 2004; does not include qroup quarters population. skewed by state, although only North Dakota (and again Washington, D.C.) showed actual declines in June 2005, an1% per year. The estimated rate for the most recent population. Of the states showing lower-than-average addition of aboutyear available, from July 2003 to July 2004, was just national growth, five were below 1%, and eight were 15 million sinceslightly under 1%. The Census Bureau population between 1% and 2%. The 20 states showing greaterclock showed an estimated 296 million in June 2005 growth than the national average, accounting for over the 2000 census,with the prospect of crossing 300 million some time 76% of the nation’s growth, are presented in Table and with the pros-in 2006. At such rates, the population increase for 2-12. With the exception of New Hampshire, all of pect of crossingthis decade will be less than 30 million, considerably these states are in the West or South. The three stateslower than that observed in the last decade, unless 300 million somethere is another demographic surprise at the end of 12 Census Bureau Projections Release, April 21, 2005. time in 2006. This rate will not TABLE 2-12 Growth by States Exceeding National Growth Rate produce as many State July 1, 2000 Population July 1, 2004 Population Growth Growth Rate (%) additions to popu- Nevada 2,018,104 2,334,771 316,667 15.69 lation as did the Arizona 5,165,765 5,743,834 578,069 11.19 Florida 16,047,807 17,397,161 1,349,354 8.41 past decade. Texas 20,949,316 22,490,022 1,540,706 7.35 Georgia 8,230,094 8,829,383 599,289 7.28 Idaho 1,299,610 1,393,262 93,652 7.21 Utah 2,243,129 2,389,039 145,910 6.50 Colorado 4,326,872 4,601,403 274,531 6.34 North Carolina 8,077,662 8,541,221 463,559 5.74 Delaware 786,397 830,364 43,967 5.59 California 33,999,879 35,893,799 1,893,920 5.57 Virginia 7,104,852 7,459,827 354,975 5.00 Washington 5,911,043 6,203,788 292,745 4.95 Oregon 3,430,706 3,594,586 163,880 4.78 New Hampshire 1,240,472 1,299,500 59,028 4.76 Maryland 5,311,531 5,558,058 246,527 4.64 New Mexico 1,821,544 1,903,289 81,745 4.49 Alaska 627,576 655,435 27,859 4.44 South Carolina 4,023,129 4,198,068 174,939 4.35 Hawaii 1,212,343 1,262,840 50,497 4.17 Total (20 states) 133,827,831 142,579,650 8,751,819 6.54 United States 282,177,754 293,655,404 11,477,650 4.07 C OMMUT I NG I N A MER I CA III | 25