Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand (2014)

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

9 Starting just after World War II, the number of miles driven annually on America’s roads steadily increased. The rising numbers were related to societal shifts, such as women joining the workforce, families moving to the suburbs, and the greater affordability of more cars for more people. As indicated in Figure 3-1, vehicle miles traveled (VMT) per capita tracked growth in gross domestic product (GDP) per capita from the 1930s to the 1960s. After that, VMT per capita grew at a higher rate than GDP per capita until around 2007. Then, Americans started driving less. The question is, why? Certainly, high gasoline prices or the cost of driving could have been factors. In 2007, gasoline prices reached the highest level since the 1981 oil shock (Money CNN 2007). According to a 2013 auto club AAA report, on average, the cost of driving 15,000 miles a year rose 1.2 cents per mile in 2012, with increases in fuel, tires, financing, license, registration, and taxes (AAA 2013). Sivak (2013a) analyzed FHWA and Census data from 1984 to 2011 to ask and answer the question, “Has motorization in the U.S. peaked?” His research, which examined the total VMT of light-duty vehicles [cars, pickup trucks, sport utility vehicles (SUVs), and vans] from 1984 to 2011, found that total VMT peaked in 2006 at 2.8 trillion miles. In 2011 (the latest year for which data are available), the number was 2.6 trillion miles (a 5 percent reduction from 2006). Sivak noted decreases in VMT per licensed driver, per household, and per registered light-duty vehicle as well. Part of the answer lies in the economic recession, which has had a very slow recovery time. But since these rates peaked before the onset of the recession (prior to 2008), Sivak concluded that they reflect other soci- etal changes that influence the need for vehicles. It is clear that VMT per capita is declining, but it is not evident what will happen or why it will happen in the future with a full economic recovery. Given numerous changes occurring in population, demographics, and travel patterns, this situation illustrates the fact that the future is difficult to predict and is shaped by many interacting factors (Curtis and Perkins 2006; Polzin 2006; Guequierre 2003; Polat 2012). 3.1 Eight Socio-Demographic Trends The following sections discuss eight key trends (past and future) associated with the socio- demographic variables that contribute to the impact on U.S. travel demand and various aspects of travel behavior (e.g., VMT, car ownership, mode choice, and trip rates). Those variables C H A P T E R 3 Key Trends, Drivers, and Projected Impact on Travel Behavior Chapter 3 Takeaways 8 Socio-demographic Trends Associated with Travel Demand • Trend 1: The next 100 million • Trend 2: The graying of America • Trend 3: The browning of America • Trend 4: The changing American workforce • Trend 5: The blurring of city and suburb • Trend 6: Slow growth in households • Trend 7: The Generation C • Trend 8: The salience of environmental concerns “We have to face the growing reality that today young people don’t seem to be as interested in cars as previous generations.” Jim Lentz, Toyota President Author

10 The Effects of Socio-Demographics on Future Travel Demand include age, household structure, acculturation, race/ethnicity, household income, workforce participation, and residential location. In addition, in selecting these trends, the study team focused on several external factors that are intertwined with socio-demographics, such as changes in vehicle and information technologies and cultural shifts in attitudes toward sustain- ability or environmental consciousness. It is important to note that demographic trends will vary substantially by region of the coun- try. A review of national socio-demographic trends is provided here. Trend 1: The Next 100 Million The United States is growing more slowly. • Drivers: Population growing but aging, declining fertility rates among white women, extended life span, and less immigration. • Impact on Travel Demand: Overall increase in total VMT due to population growth; VMT per capita appears to be declining. The 2000s marked the lowest decennial rate of population growth since the Depression (see Figure 3-2). Between 2000 and 2010, the U.S. population grew by 27.3 million (about 10 percent), Sources: VMT data from FHWA 2012a; GDP data from BEA 2013. 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 1930 1950 1970 1990 2010 R ea l G DP p er c ap ita (C on sta nt 20 05 U S $) VM T pe r c ap ita VMT GDP Figure 3-1. Trends in VMT and GDP—1930 to 2010. Source: Brookings Institution 2010. 7.3 14.5 18.5 13.3 11.5 9.8 13.2 9.7 0 2 4 6 8 10 12 14 16 18 20 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s Pe rc en t Figure 3-2. Population growth by decade, 1930s–2000s.

Key Trends, Drivers, and Projected Impact on Travel Behavior 11 which was smaller than the growth from the decade before (1990 to 2000), both in absolute terms and as a percentage of initial population. The decrease from the 13 percent growth in the 1990s reflects slower U.S. economic growth, reduced immigration, declining fertility rates among white women, and aging Baby Boomers. Whether this decade-long trend in slowing population growth represents a long-term change will be influenced by the extent to which the relevant societal changes become permanent. Net change in the U.S. population results from adding births, subtracting deaths, adding people who migrated to the United States, and subtracting people who left. Based on these factors, the U.S. Census Bureau projects that the U.S. population will grow over the next four decades (2010–2050) from 310 million to just over 400 million by 2051 (Census 2013). Based on trends from the last four decades (1970–2010), it is estimated that the majority of the popu- lation growth (to 2050) will be due to immigrants and their descendants. During these past decades, the U.S. population increased by 52 percent, from about 203 million to 310 million (Hobbs and Stoops 2002). Much of this growth—32.7 million people—occurred during the 1990s. This was the largest numerical increase of any decade in U.S. history, although the growth varied geographically, with large population increases in some areas of the country and little growth or decline in others. The 1990s population growth was largely immigration- driven. The previous record increase was in the 1950s, a gain fueled primarily by the post- World War II Baby Boom. Total population growth is also due to a decline in U.S. mortality. One measure of mortal- ity levels, independent of population age structure, is life expectancy at birth. Life expectancy has increased at a steady rate, with occasional one-year declines, over the past 35 years. There is a lack of consensus as to whether life expectancy will continue to increase, or is approach- ing some biologically fixed limit (Sonnega 2006). Increasing life expectancy at birth (and thus decreasing mortality rates) clearly leads to increases in the total population, all else being equal. Also, it is important to note that America’s population is growing at a faster rate than the rest of the world’s developed nations. An article in The Economist postulated that by 2040, and pos- sibly earlier, America will overtake Europe in population. According to past trends, any popula- tion growth should stimulate an increase in total U.S. VMT (Economist 2002). Trend 2: The Graying of America America is becoming “grayer.” The population age 65 and older will significantly increase as the Baby Boom generation enters this demographic group. • Drivers: Population aging, extended life spans, “boom-and-bust” birth rate patterns. • Impact on Travel Demand: Decreased per capita VMT, decreased work trips, increased vehicle age, decreased auto ownership, increased carpooling, decreased transit use. Population aging is evident in the increasing share of the population in the older age catego- ries as the Baby Boom generation becomes older (see Figure 3-3). The large Baby Boom cohort is just now reaching age 65. The oldest members of the generation, born between 1946 and 1964, entered seniorhood at the end of the 2000s, as the youngest members crossed fully into middle age (Passel and Cohn 2008). As the Boomers age, the percentage of Americans who are age 65+ is estimated to nearly double from one of every 8 Americans in 2000 to more than one of every 5 in 2050. Their growing numbers, whose households are smaller on average than adults under 65, has tended to add to a decreasing overall size of households over the past several decades (discussed later in this chapter).

12 The Effects of Socio-Demographics on Future Travel Demand When talking about the aging of the U.S. population, it makes sense to discuss this in terms of the succession of accepted generations, demarcated by historical circumstances. The United States has experienced a generational boom-and-bust birthrate pattern, as described below: • Depression Babies—Born 1930–1945 and age 65–80 in 2010. Their numbers are small due to low birth rates in a poor economy. • Baby Boomers—Born 1946–1965 and age 45–64 in 2010. Their numbers are large, and their entry into any age group had a major impact on that group’s growth. • Generation X/Baby Bust—Born 1966–1980 and age 30–44 in 2010. Their numbers are small, being an “echo” of the Depression era generation. • Generation Y/Millennials—Born 1981–1995 and age 15–29 in 2010. As the children of the Baby Boom generation, their numbers are relatively large. This generation is much more racially and ethnically diverse and will become more so, as immigrants gradually increase the number of young adults in the United States. During the last several decades, Baby Boomers, most of whom are non-Hispanic white, have dominated the U.S. population. But as Baby Boomers reach old age, their dominance is being replaced by another younger cohort (Millennials) that is much more likely to be Hispanic, Asian, or multiracial (Jacobsen et al. 2011). The rapid increase in diversity among younger cohorts may be creating a new kind of generation gap. Although historically the generation gap has been defined by different cultural tastes in music, fashion, or technology, this new demographic divide may have broader implications for transportation-related planning and policy making, as is discussed later in this chapter. Age plays a significant role in patterns of vehicle use. Traditionally, it was well established that VMT levels change with age and were at their highest for middle-age adults who are in peak levels of both work-related and household-serving travel. For the Depression era generation, there is also decreasing capability to drive. For Baby Boomers, there is less work travel as they enter retirement. At the same time, Sivak and Schoettle (2011) found substantial increases from 1983 to 2008 in the percentage of older people with a driver’s license. For people age 65–69, the portion increased from 79 percent in 1983 to 92 percent in 2008, and for people age 70+, it increased from 55 percent in 1983 to 80 percent in 2008. While younger people appear to be delaying licensing as discussed later in this chapter, older people are retaining their licenses and continuing to drive (Stokes 2012). Still, per person, older people also tend to drive less, so VMT per capita will be less. Source: Brownell et al. 2013. 26 43 31 26 40 34 24 37 39 20 25 30 35 40 45 50 Less than 18 18 to 44 45 and older Pe rc en t 1990 2000 2010 Figure 3-3. Percentage of U.S. population by age group, 1990–2010.

Key Trends, Drivers, and Projected Impact on Travel Behavior 13 It is also clear in recent cross-sectional data sets at the regional level that the percentage of trips people make by nonauto modes (transit, walk, and bike) tends to decrease with age. This can be explained to some extent by the physical demands of biking, walking, and getting to and from bus stops becoming more difficult with age. However, the findings could also be explained by cohort effects, where older generations have gotten “out of the habit” of walking and biking, which they may have done more when they were younger. As an example, in a 2010 survey by KRC Research and Zipcar, participants were asked to what extent they agreed with the state- ment: “In the past year, I have consciously made an effort to reduce how much I drive, and instead take public transportation, bike/walk or carpool when possible (KRC 2010, Davis et al. 2012).” Only 24 percent of people age 55+ agreed with this statement, compared with 31 percent of those age 45–54, 41 percent of those age 35–44, and 45 percent of those age 18–34. Trend 3: The Browning of America America is becoming “browner.” The white population has grown more slowly than every other race group in the second half of the 20th century. • Drivers: Structural changes in population distribution by race/ethnicity, relatively high fertility rates among Hispanic women, continuing immigration in younger age groups. • Impact on Travel Demand: Increase in VMT per capita, increase in auto age, greater public transit use. White non-Hispanics accounted for a majority of the U.S. population in 2010, but their share has declined over time as the shares of other groups, particularly Hispanics and Asian/ Pacific Islander populations, have grown significantly faster (see Figure 3-4). Immigration policy reforms in 1965, and more significantly in 1986, resulted in a wave of immigration between 1965 and 2000. As an example, the Hispanic population more than tripled from 1970 to 2000, from Source: Census Bureau 2011 Population Estimates (Passel and Cohn 2008). 6317 12 5 2011 47 29 13 9 2050 85 3.5 11 0.6 1960 White Hispanic Black Asian Projected Figure 3-4. Population by race and ethnicity, actual and projected (% of total).

14 The Effects of Socio-Demographics on Future Travel Demand about 9 million to 35 million. Between 2000 and 2010, the total U.S. population grew by 27 million, the U.S. Hispanic population grew by 15.2 million—56 percent of the total in national growth (Arce 2011). In 2010, Hispanics numbered nearly 50 million and accounted for 16 percent of the U.S. population. The U.S. population is bound to become more diverse over the next 30–50 years due to the demographics of America’s children. Between 2000 and 2010, the number of children in the United States grew by less than 1.9 million. A closer look shows that the number of Hispanic children grew by 4.8 million, and the number of non-white, non-Hispanic children grew by just 1.4 million. Meanwhile, the number of non-Hispanic white children shrank by 4.3 million— creating the increase of less than 1.9 million children overall. Another way to look at this is that today 80 percent of Americans over the age of 65 are white non-Hispanic, and only 7 percent are Hispanic. In contrast, 55 percent of children under age 18 are white non-Hispanic, and 23 percent are Hispanic. Although the Census Bureau expects Hispanic birth rates to drop in the coming years due to acculturation, their rates are currently higher than those of non-Hispanics of any race (2.7 com- pared with 1.83 for non-Hispanic blacks and 1.90 for non-Hispanics of other races) (Pendall et al. 2012). So even as Hispanic birth rates fall, Hispanics will constitute larger percentages of the national population, resulting in more total births, even if the rate is lower. Based on such influences, the Census Bureau projects that the share of children who are Hispanic will rise from 23 percent today to 35 percent in 2050, which will have major implications for the His- panic proportion of the total population (Arce 2011). In 2030, an estimated one of four (24 percent) Americans will be Hispanic; in 2050, three of ten (29 percent) will be Hispanic. Births, rather than immigration, will play a growing role in Hispanic and Asian population growth. As a result, in the future, a much smaller proportion of both groups will be foreign-born than now. Current travel demand forecasting models rarely explicitly use input variables related to race, ethnicity, or acculturation level. Their effects are picked up indirectly though other variables, such as household income and geography. As a result, the influences of immigration and accul- turation on travel behavior over time have received very little attention in the field of travel demand forecasting. Yet, because they could be very important considerations for the future, they are accounted for in the scenario analysis tool Impacts 2050. The 2009 National Household Travel Survey (NHTS) data show that drivers and workers per household are significantly higher for U.S.-born Hispanics than for foreign-born Hispanics and non-Hispanics, indicating a greater propensity for trip making (FHWA 2012b). U.S.-born Hispanics also have more vehicles per household and own more newer vehicles compared with foreign-born Hispanics. Foreign-born Hispanics show a higher propensity for carpooling, walk- ing, and transit due to a number of factors, including a lack of driver’s license and the costs asso- ciated with owning a vehicle and driving (Liu and Painter 2012). As Hispanics become a larger portion of the total U.S. population and if the trends among Hispanic households continue, then the United States could experience increasing public transit use and aging of the vehicle fleet. Trend 4: The Changing American Workforce America’s workforce is growing older, more female, and more diverse. • Drivers: Boom-and-bust birth rate patterns, population aging, female work participation patterns, female longevity, structural changes in racial/ethnic distribution of labor force, immigration. • Impact on Travel Demand: Decreased VMT per capita, increased work-related VMT, lower growth in work-related VMT, increased carpooling.

Key Trends, Drivers, and Projected Impact on Travel Behavior 15 The overall size of the U.S. labor force has been increasing over time due to population growth and increased female participation in the labor force. Between 1980 and 2010, the proportion of women in the labor force increased from 52 percent to 59 percent of all women over age 16, while the proportion of men decreased from 77 percent to 71 percent of all men over age 16 (Hobbs and Stoops 2002). The labor force is projected to increase by 10.5 million in the next decade, reaching 164 mil- lion in 2020 (Toossi 2012). This represents a slower rate of growth than previous decades, pri- marily the result of a slower rate of population growth since 2000 and a decrease in the labor force participation rate stemming from the 2007–2009 recession and its aftermath. The Bureau of Labor Statistics (BLS) bases its labor force projections on the civilian noninstitutional popula- tion (age 16 and older). The population will continue to exhibit structural changes that will have significant impacts on the U.S. workforce (see Figure 3-5). For example, according to the BLS, the share of 16–24-year-olds in the workforce is declining—from 17 percent in 1992 to 16 percent in 2012 to a projected 14 percent in 2022. Even more significant declines are observed among 25–54-year-olds, who represent the prime age group for workers. This group’s share of the civilian noninstitutional population dropped from 56 percent in 2002 to 51 percent in 2012, and is projected to drop further to 48 percent in 2022. This “baby bust” generation reflects the drop in birth rates that took place from 1965 to 1975. By contrast, the 55 and older age group increased its relative share in the civilian noninstitutional population from 26 percent in 1992 to 28 percent in 2012, and is expected to grow to 38 percent in 2028. Interestingly, the share of women in this group was 8.7 million more than men in 2010, and is expected to be 7.9 mil- lion more in 2022. So as the civilian noninstitutional population shifts to higher age groups, it becomes more female. These structural changes will have significant impacts on labor force participation rates, which have declined over the past decade. The primary driver of this overall trend is the aging workforce. As the Baby Boom generation moves from middle age, with high participation rates, to the older age groups, with significantly lower participation rates, the overall labor force par- ticipation rate will decline. The trend accelerated during the recession, suggesting that many more people dropped out of the workforce than otherwise would have if the economy were in better shape. This rate peaked at 67 percent from 1997 to 2002, and then declined from 2002 to 2012, dropping to 64 percent. As the economy improves, however, the workforce will continue to age (Toossi 2013). All things being equal, the United States will experience slower economic growth, which should result in decreased VMT per capita. Source: BLS, Monthly Labor Review, December 2013 (Toossi 2013). 0 20,000 40,000 60,000 80,000 100,000 120,000 16 to 24 25 to 54 55 and older Th ou sa nd s 1992 2002 2012 2022 Figure 3-5. Civilian labor force by age.

16 The Effects of Socio-Demographics on Future Travel Demand As noted, the U.S. Hispanic population is growing at the fastest rate of all racial and ethnic groups. The Hispanic share of the civilian noninstitutional population is projected to increase from 12.4 percent in 2002 to nearly 19 percent in 2022 (Toossi 2013). Hispanics in the aggregate and Hispanic men had the highest labor force participation rates in 2010. Also, Hispanics have a younger population than other racial and ethnic groups and a greater proportion in the prime working-age groups. These trends will continue and will increase the trend of even more racial and ethnic diversity in the workforce in the next four decades. Labor force participation is important because commute trips are a major contributing fac- tor to peak period congestion. In 2010, 86 percent of workers drove to work (76 percent alone, 10 percent in a carpool) (McKenzie and Rapino 2011). The percentage of Hispanic and Asian workers who drove alone did not exceed 70 percent. While 5 percent of all workers used public transport, the rate of public transit use among foreign-born workers was nearly twice that of native-born workers (11 percent versus 4 percent). However, as Hispanic workers acculturate, their use of public transit is likely to decrease, while their auto use increases. Trend 5: The Blurring of City and Suburb The differentiation between cities and suburbs is fading. • Drivers: Population growth, housing starts, population aging, age structure, household structure. • Impact on Travel Demand: Decreased VMT per capita, increased nonmotorized trips, increased transit trips. U.S. population density, defined as the number of people per square mile of land area, increased from 50.7 in 1960 to 87.4 in 2010 (see Figure 3-6). Over the same period, central cities have become less dense, and the density of suburbs has changed very little (Census 2012, Hobbs and Stoops 2002). However, at the turn of the 21st century, urban population growth acceler- ated. Census data indicate that many city centers grew faster than their suburbs between 2010 and 2012 for the first time in decades (Census 2012). Viewed as a whole, though, U.S. suburbs have continued to grow faster than city centers in every decade since the 1920s. Nevertheless, the division between city and suburb is blurring. There is no longer always a clear line between an economic center where people work and suburbs where people live. Both can be home to employers and residences. It is also misleading to think of all suburbs in the Source: Census 2012 and Hobbs and Stoops 2002 50.7 57.4 64 70.3 79.6 84.7 1960 1970 1980 1990 2000 2010 Figure 3-6. Number of people per square mile of U.S. land area.

Key Trends, Drivers, and Projected Impact on Travel Behavior 17 same way; in the past decade, inner-ring suburbs have experienced population changes more similar to those in center cities than to outer-ring suburbs. Data from the 2010 Census also show that many closer-in suburbs linked to a city with public transit or well-developed roadways are benefiting from strong city growth, while exurbs near the metropolitan edge are not prospering quite as well. And many trends vary with the region of the country and the economic prosperity of the metropolitan region. One factor that is contributing to the blurring of cities and suburbs is that U.S. internal migra- tion rates have been declining for several decades (Census 2009). These rates reached record lows in 2009, with 2 percent of Americans moving from one state to another, and 4 percent moving from one county to another (Cooke 2011). The most recent spike in migration rates occurred in the late 1990s (along with strong economic and population growth). Then migration rates began a strong downward turn that was accelerated by the economic recession of 2007–2008 and has continued long after, indicating that economics does not completely explain the trend. Behind the total numbers, different trends are happening in suburban and urban growth. As the Baby Boomers age, many of them are choosing to remain in suburban areas. As a result, the suburbs are both growing and aging quicker than central cities. An AARP analysis of 2010 Census data showed that 9 of 10 older adults nationally were living in the same communities where they raised their children (Faber et al. 2011). In 2000, 34 percent of suburban residents were over 45; by 2010, 40 percent were. In contrast, in central cities, the population over 45 increased from 31 percent to 35 percent. More and more suburban households are made up of singles, empty nesters, or retirees—they just happen to be older than urban residents fitting these household structure types. However, this trend is not necessarily consistent across the country; some metropolitan suburbs have successfully attracted younger residents, while others have shed them (Frey 2011). Among central cities in the 100 largest metropolitan regions, two-thirds gained population from 2000 to 2008, continuing a trend that began in the 1990s. Some of this was attributed to immigration, because the largest cities remain magnets for newcomers, and some to the fact that as housing prices began to decline in 2006, center city residents who might have moved to the suburbs instead remained in cities. In a recent study by the Urban Land Institute, Millennials were more likely than older Ameri- cans to prefer living in a big city, and showed the strongest preference for communities with mixed uses and different types of housing (BRS 2013). Likewise, 2010 Census data indicate that 20–34-year-olds who are delaying marriage much longer claim a disproportionate share of new city residents since 2008. Urban markets have greater job densities that are appealing to younger adults facing poor job prospects due to the recession. In Washington, D.C., for example, 28 per- cent of recently arriving migrants (2008–2010) were age 25–34, compared with the 15 percent they comprised of the region’s population (Sturtevant 2013). In cities, especially their centers, car ownership and use are declining. According to 2010 Census data, the share of metropolitan residents without a car has grown since the mid-1990s. Currently, 13 percent of people in cities of more than 3 million people have no car, compared with 6 percent of people living in rural areas. There are various reasons for this. Public mass- transit systems are, in the main, faster than they used to be, with increased capacity in many cities. More recently, private alternatives to car ownership or car-sharing services, such as Zipcar and Car2Go, have growing membership. In addition, an emerging autonomous vehicle market may also influence car ownership and use. Such population and demographic changes are gen- erational events that will take decades to fully shake out. It is difficult to know exactly what form it will ultimately take, but re-energized cities and more compact suburbs may lead to a less car-dependent way of life, with more walking and biking and use of public transit.

18 The Effects of Socio-Demographics on Future Travel Demand Trend 6: Slow Growth in Households The rate of new household formation has plunged since 2006, creating more single households and also more multigenerational and larger households. • Drivers: Poor labor market, aging population, lifestyle choices of Millennials. • Impact on Travel Demand: Decreased per capita VMT, decreased auto ownership among young people, increased carpooling, increased public transit use. Between 2006 and 2010, an average of 850,000 households were formed per year, compared with an average of 1.68 million per year over the previous five years (see Figure 3-7). In fact, household formation during 2006–2011 appears to have been far lower than in any five-year period over the past 40 years (Paciorek 2013). New households can be formed when children move out of their parents’ homes, when couples separate, or when unrelated individuals choose to live singly after previously sharing a residence. The number of households can decline if two households combine, either through marriage or by sharing a residence to reduce housing costs. In the current environment, household formation rates may well be depressed both because fewer young people are living on their own, because established households are combining to lower costs, or because of the loss of homes through foreclosure. While the total U.S. population increased greatly during the 20th century, the percentage increase in the number of households was even greater, reflecting a trend of higher proportions of people living in smaller households (Hobbs and Stoops 2002). The average household size has been decreasing from 4.6 people per household in 1900 to 3.3 in 1960 to a low of 2.59 in 2000. The long trend of falling household sizes had five main drivers: lower fertility or fewer children, aging Baby Boomers, longer life spans, women entering the labor force, and rising incomes. Then between 2000 and 2010, the average U.S. household size rebounded to its 1990 level of 2.63. This rebound has been tied to young adults responding to the “Great Recession” by moving back in with their families and delaying moving from home for the first time (Kochhar and Cohn 2011, Taylor et al. 2010). American Community Survey data indicate that in 2006, prior to the Great Recession, 15 percent of young adults age 25–34 lived in the same household with one or both parents (Payne 2012). In 2010, the share among this age group living with parents was 18 percent, reflecting an increase of approximately 1.27 million young adults living with their parents. Residing in the parental home is often an adaptive strategy during times of economic distress (Furstenberg 2010). Over time, single young adults are consistently more likely to live in their parent’s home. Just more than three-fourths did so in the 1940s, and about half in 2010. What Source: Federal Reserve Board of Cleveland, Economic Commentary, No. 2012-12, 2012 (Dunne 2012). 0 500 1000 1500 2000 2500 3000 3500 4000 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 H ou se ho ld s, T ho u sa n ds 15 year average Figure 3-7. Net new households, 1996–2010.

Key Trends, Drivers, and Projected Impact on Travel Behavior 19 is different about today is that the age at first marriage has increased. The vast majority (90 per- cent) of 18–24-year-olds have never been married. So there is both an economic element to this trend, but also an element of the lifestyle choice of the Millennial generation. This living in another’s home is not just a lifestyle choice of the young. According to the Pew Research Center, in 2008 some 49 million Americans, or 16 percent of the total U.S. population, lived in a family household that contained at least two adult generations or a grandparent and at least one other generation (Taylor et al. 2010). That is a significant increase from 1980, when the figure was about 28 million, or 12 percent of the population. There are a number of possible factors for this increase, from population aging to financial reasons to the growth of the Hispanic and Asian populations. The composition of households may affect transportation demand through both the number of people and their ages and relationships. In particular, households with children have higher VMT than households without children. According to the 2009 NHTS, households with chil- dren averaged 30,400 VMT per year, while households without children averaged only 14,400 VMT per year (FHWA 2012b). But 2010 Census data indicate that households with children under 18 years have grown at the slowest rate over the period from 1960 to 2010, and increased by only 0.5 percent between 2000 and 2010. Also, young adults (Millennials and Gen X) are delaying marriage, which also has a depressing effect on creating households with children. Over this same decade (2000–2010), households without children under 18 years increased by 15 percent, from 37.2 million to 42.8 million. Single-person households also increased by 15 percent, from 27.2 million to 31.2 million. Multiple-person nonfamily households increased by 23 percent, from 6.5 million to 8 million. It is also likely that the low propensity to acquire a car could be related to young adults living with their parents. The probability of buying a new vehicle peaked with people 55–64 years of age (Sivak 2013b). In 2011, one vehicle was purchased for every 14.6 drivers in that age group. By com- parison, the rate was one vehicle for every 222 drivers age 18–24 and for every 35 drivers age 25–34. Trend 7: The Generation C Mobile broadband will become increasingly more important and ubiquitous, creating a new Generation C. • Drivers: Technology evolution, lifestyle choices, age structure. • Impact on Travel Demand: Reduced VMT per capita for some trip purposes, decreased car ownership. The growing influence of digital and mobile devices in the way people live, work, and social- ize has spawned a new generation. Generation C is not necessarily a demographic group, as it is a lifestyle segment. It represents people of any generation who are connected, communicating, content-centric, computerized, and community-oriented (Friedrich et al. 2010). Generation C is “always clicking.” However, some researchers, such as A.C. Nielsen, have identified Generation C as the hyper-connected group that is currently 18–34 years of age (Nielsen 2012). Born during or after the introduction of digital technology in the 1990s, this generation has been interacting with technology from an early age. The mobile industry is constantly evolving and growing at an astronomical pace. For example, just over six years ago, Apple sold its first iPhone. In 2012, Apple shipped about 250 million iPhones worldwide (Pelson 2012). The intrinsic value of owning a mobile device has significantly changed in recent years. What was once simply a constant voice connection to anyone with a telephone is now enabling communications in a variety of different ways, such as text messages, e-mails, social media updates, instant messages, blog posts, Web searchers, shopping, and much more.

20 The Effects of Socio-Demographics on Future Travel Demand Trend data indicate that these alternative means of communication have thrived among mobile phone users. A 2013 Pew Research Center survey found that 91 percent of American adults own a cell phone and 56 percent of adults own a smartphone (Pew 2013a) (Figure 3-8). Ownership varies by generation—with 79 percent of Millennials reporting owning a smart- phone, compared with 69 percent of Gen X, 47 percent of Baby Boomers, and 18 percent of Depression era respondents. Urban and suburban ownership were virtually the same, at 59 per- cent while rural ownership was 40 percent (Pew 2013b). It appears that the convenience and cost of mobile devices may have created a shift in the need or desire for a fixed-connection computer in the home. Yet, nearly one third of Americans still lack home broadband access (NTIA 2013). This is where the “digital divide” manifests itself. Poor, rural, and ethnically diverse households have long lagged behind in the rate of home broadband adoption. Hispanics are less likely to have broadband at home than any other demo- graphic group, and they are far less likely to have it than whites and Asian Americans: 65 percent of whites and 69 percent of Asian Americans have broadband at home, compared with only 45 percent of Hispanics, 46 percent of Native Americans, and 52 percent of African Americans (Livingston 2011). Also, 18 percent of African Americans and 16 percent of English-speaking Hispanics are mobile phone-only wireless Internet users, compared with 10 percent of whites. According to another measure, Hispanics and African Americans are more than six times as likely as non-Hispanic whites to use their mobile phones as their sole means to access the Inter- net. When put together with the facts that Hispanics are growing as a percentage of the popula- tion and that Millennials and Gen X’ers have very high smartphone usage, this trend suggests that reliance on fixed broadband connections will decline over time. Research suggests that cohort effects alone will change the travel behavior of the next genera- tion in unique ways from their predecessors (Blumenberg et al. 2012). One of the key features of digital technologies is that they enable people to perform activities remotely, rather than in person. This type of freedom could displace the perceived freedom granted by ownership of a car—which has long been a key selling point for auto manufacturers (Goodwin 2012). Hallett and Stokes (1990) made an interesting prediction: Another possibility is that some new product could hit the market which would make the car redundant in the psychological sense . . . some computing product (probably portable) could maybe be produced which would cater to power, or freedom desires. Source: Pew 2013b. 73% 75% 78% 85% 84% 82% 85% 83% 88% 88% 85% 85% 87% 91% 35% 45% 46% 45% 47% 45% 56% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cell phone Smartphone Figure 3-8. Percentage of American adults by cell phone and smartphone ownership.

Key Trends, Drivers, and Projected Impact on Travel Behavior 21 However, they also go on to say that they do not believe this was at all likely at that moment (in 1990). Even among early adopters, researchers focused on whether this technology would result in substitution for actual trip making. Since then, studies have documented various small effects— slightly less or more trip making, depending on the trip purpose. The big effect many researchers anticipated might have been missing, because adults had already established their typical travel patterns by the time they started interacting with digital technology in the 1990s. Might we find different effects among the digital natives? The recent change in driving patterns among younger people may be related to their mobile device use. The majority of American teens today delay getting a driver’s license. According to the AAA Foundation for Traffic Safety, less than half (44 percent) of teens obtain a driver’s license within 12 months of the minimum age for licensing in their state, and just over half (54 percent) are licensed before their 18th birthday (Tefft et al. 2013). This contrasts with ear- lier generations. In 1990, The FHWA’s 1990 Nationwide Personal Transportation Survey found that 41 percent of 16-year-olds, 70 percent of 17-year-olds, and 77 percent of 18-year-olds were licensed drivers (RTI 1991). For today’s Millennials, influences have included additional requirements and costs associ- ated with getting a driver’s license, greater access to mobile technology, and attitudinal shifts that favor low-carbon transport, especially among urban dwellers. More recent data from FHWA’s 2009 NHTS indicates that the share of automobile miles driven by people age 21–30 in the United States fell to 14 percent in 2009 from 18 percent in 2001 and 21 percent in 1995 (FHWA 2012b). Meanwhile, 2010 Census data show the proportion of people age 21–30 increased from 13 percent to 14 percent (from 2000 to 2010), indicating that this age group went from driving a disproportionate amount of the nation’s highway miles in 1995 to under-indexing for driving in 2009 (Census 2012). While earlier generations embraced a concept of mobility structured around highways and automobiles, digital natives appear to be delaying the acquisition of a driver’s license, driving less, and doing more activities digitally. Researchers have examined the role of social network- ing sites in changing travel behavior (Binsted and Hutchins 2012) and trends regarding young people driving less (Davis et al. 2012). But quantifying in a rigorous way such potentially signifi- cant changes for their impact on future travel behaviors is challenging. Trend 8: The Salience of Environmental Concerns The generational divide over the nation’s energy and environmental priorities is still strong but will decrease over time. • Drivers: Age structure, population aging. • Impact on Travel Demand: Lower car ownership, more transit and nonvehicle travel by younger generations due to elderly population shrinking. According to a 2011 Pew Research Center poll, different generations of Americans have starkly different views on some of the social issues facing the United States today (Pew 2011). While the generational differences in views on energy and environmental priorities are not as pronounced as for some other issues (such as diversity, gay marriage, or civil liberties), genera- tional differences exist. In terms of energy and environmental priorities, 71 percent of Millennials say the United States should focus on developing alternative energy sources, rather than expanding oil, coal, and natural gas exploration. Roughly the same percent (69 percent) of Gen X’ers and 60 percent of Baby Boomers agree. But only 47 percent of Depression era respondents agree. The same

22 The Effects of Socio-Demographics on Future Travel Demand attitudinal distribution is seen on the question of whether stricter environmental laws and regu- lations are (1) worth the cost or (2) cost too many jobs and hurt the economy. For Millennials, the split in agreement with these two statements was 57 percent and 35 percent; for Gen X’ers, 57 percent and 37 percent; for Baby Boomers, 54 percent and 40 percent; and for Depression era respondents, 40 percent and 49 percent (Pew 2011). The same Pew Research Center survey asked about policies to address America’s energy sup- ply. Regarding a policy option to spend more on subway, rail, and bus systems, 61 percent of Millennials favored this solution, compared with 62 percent of Gen X’ers, 62 percent of Baby Boomers, and 47 percent of Depression era respondents. With respect to another policy that addressed tax incentives for buying hybrid/electric vehicles, 69 percent of Millennials favored the policy, compared with 67 percent of Gen X’ers, 56 percent of Baby Boomers, and 38 percent of Depression era respondents (Figure 3-9). Generational differences in attitudes relating to environmentally conscious behaviors are also observed in other research. In the 2013 Urban Land Institute study, convenience to public tran- sit in one’s community was important to 57 percent of Millennials, 45 percent of Gen X’ers, 50 percent of Baby Boomers, and 50 percent of Depression era respondents, while walkability was important to 76 percent of Millennials, 67 percent of Gen X’ers and Baby Boomers, and 69 percent of Depression era respondents (BRS 2013). Millennials are the generation most likely to prefer living in a big city and most eager to live in a place with extensive public transit options. More of them use public transit to commute than any other generation. As the U.S. population ages and older generations pass away, attitudes and behaviors of Millennials will have greater and greater influence on Americans’ environmental footprint. These eight trends highlight the range of uncertainties transportation agencies are facing in their long-range planning. They make clear that the potential impact of these trends on travel demand is conflicting, evolving, and incomplete. This situation illustrates the fact that the future is unpredictable and shaped by many interacting factors. This is where scenario planning comes into play, as is discussed in the next chapter. It is not clear which trends will dominate the others or will change over time or remain moving in the same direction. Source: Pew 2011. 71% 61% 69% 69% 62% 67% 60% 62% 56% 47% 47% 38% 20% 30% 40% 50% 60% 70% 80% 90% 100% Focus on Alternative Energy Sources Spend more on Subway, Rail and Bus Provide Tax Incentives for Hybrid/Electric Vehicles Millennials Gen X Baby Boomers Depression Era Figure 3-9. Generational differences in views on energy and environmental priorities: percentage in agreement.