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.
Appendix C Vehicle-Miles Traveled: Trends and Implications for the U.S. Interstate Highway System Steven E. Polzin This appendix provides context for the historical trend of vehicle-miles traveled (VMT) and insight into how future conditions might influence travel demand on the U.S. Interstate Highway System. Transportation requires significant investments in infrastructure that often require signifi- cant lead times and produce a host of consequences, from changes in land uses and values to emissions and energy consumption to employment and economic opportunity, all of which are of interest to policy makers and the public. Thus, analysts have long sought to understand future travel demand sufficiently that investment and policy decisions can be implemented to respond to and influence that demand. Although disclaimers are appropriate for any discussion of the future, it should be understood that the current pace of change and uncertainty regarding key factors that influence travel demand is unprecedented in the history of our Interstate System. Demographic and economic conditions continue to fluctuate, fuel price changes affecting travel demand continue to be dynamic, and uncertainties associated with these factors are exacerbated by the transformational changes in technology that are being developed and deployed. These changes and those anticipated will continue to have dramatic impacts on travel. For example, forecasts for self-driving vehicle deployment range from 5 to 50 years in the future and are hypothesized to reduce travel demand by 30 percent or increase it up to 50 percent. The range of consequences associated with evolving technologies is dizzying, with both positive and negative effects on travel demand. Telecommut- ing, e-commerce, online education, and electronic document transfer are examples of technologies reducing the need for travel. Simultaneously, 231
232 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM technologies enable exposure to new travel destinations, provide oppor- tunities for same-day deliveries, and create the prospect for lower-cost travel through vehicle sharing and reductions in the onerousness of travel by virtue of relieving the driver of the need to drive the vehicleâfactors that can induce significant additional travel and affect safety. Visions of empty vehicles shuttling between assignments and low-cost travel in shared, electric, self-driving vehicles bolster scenarios of rapid increases in travel demand, with the fundamental economics of lower cost inducing additional consumption of travel. This appendix first characterizes historic trends in VMT growth with specific attention to the Interstate System, and then addresses our under- standing of what factors influence travel demand and discusses the demand growth and implications. Before venturing into discussions of the future demand for travel, it is helpful to reflect that despite massive intellectual and monetary investments, the record for forecasting election results, stock market prices, and consumer preferences for various products and media is less than stellar. Forecasting human behavior and all the underlying factors that influence it remains difficult. So, too, forecasting future travel demand is highly uncertain and challenging. TRAVEL DEMAND TRENDS Figure C-1 shows the long-term trend in VMT and national population since 1900. World War II and the two energy crises in the 1970s are the only noticeable fluctuations on the trend line until approximately 2005, and the subsequent dip and recent recovery have resulted in an unprecedented approximate 8-year pause in the upward trend in VMT. From 1945 through 2005, VMT increased nearly 12-fold, 4.23 percent per year on average, while population more than doubled, increasing 1.23 percent per year. Dur- ing that same period, the annualized rate of gross domestic product (GDP) growth was 3.15 percent. Seemingly reasonable extrapolations of these VMT growth trends re- sulted in numerous long-range plans produced in the 1990s and early 2000s that predicted gridlock levels of congestion and correspondingly large trans- portation infrastructure needs. The softening of VMT growth, empirically evident in aggregate VMT data in the early 2000s and hinted at by other data sources such as travel surveys even earlier, proved the undoing of the âsky is fallingâ forecasts and gave pause to the use of long-term VMT trends as the basis for future VMT forecasts. Simultaneously, reflections on our theory of travel behavior have pro- vided a logical basis for considering that fundamental trends in VMT were beginning and continue to change. The trend of women joining the formal workforce is substantially complete, an aging population influenced by the
APPENDIX C 233 large baby boomer cohort moving past their peak travel years has lessened their travel demand, auto availability levels have neared saturation, and the flight of urban residents to suburban areas may be playing itself out. Shifts from shared-ride travel, mass transit, and walk modesâsome of the sources for increased VMTâhave less room to drop (Polzin 2006). Substitution of communication for transportation (e.g., e-commerce, distance learning, social media for in-person communications, telecommuting, electronic dis- semination of documents, and voice and video media) and globalization of manufacturingâin effect, exporting the VMT associated with transporting inputs to productionâinfluence travel demand. Between 1945 and 2005, VMT increased more than 45 billion miles per year, and in the next decade, VMT increased, on average, about 10 billion miles per year despite a meaningfully higher population. The bounce-back indicated by preliminary 2016 data shows an increase well over 100 bil- lion miles and is the largest year-over-year increase ever posted. Although a recovering economy and lower fuel prices are often noted as the reasons for this, the changes in VMT since the late 1990s leave analysts with a great deal of uncertainty regarding future forecasts. As noted, this uncertainty is exacerbated by emerging trends related to the prospect of vehicle-sharing and self-driving vehicles. FIGURE C-1 National annual VMT trends and population trends, 1900â2016. SOURCES: FHWA 2017; U.S. Census Bureau 2016.
234 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Figure C-2 shows the trend in VMT and VMT per capita since January 1992. As is apparent in the graph, VMT per capita remains slightly below its peak level in the 2003â2004 time frame, but total VMT has rebounded since the recession and is at all-time highs. Figure C-3 reports monthly VMT and VMT per capita trends in terms of percentage change since January 1992. Geographic Distribution of Travel Demand The magnitude of the implications for the U.S. Interstate System of the new demand for transportation capacity is dependent on the geography distribution of increased travel demand. At the national level, it is common to talk about low single-digit percentage changes in demands or VMT. If we presume that they are uniform across the system, accommodating the increase in travel demand associated with the forecasted approximately 0.8 percent per year growth in population over the next several decades, it does not sound overwhelming. However, growth in demand is not geographically uniform, and hence, the pressure for additional capacity will not be uniform. This results in FIGURE C-2 National VMT and VMT per capita trend, moving 12-month total, 1990â2016. SOURCES: FHWA 2017; U.S. Census Bureau 2016.
APPENDIX C 235 the consequences of demand growth being concentrated and more likely to require more significant capacity increases than can be tolerated or accommodated by incremental operations improvements of existing infra- structure. As an analogy, a 3-inch snow might be passable, but that same snow blown into 3-foot drifts requires plowing. The smaller the share of the system over which new capacity demands occur, the more likely it is to require substantive capacity increases to ensure adequate performance. Other sections of this appendix speak to the various factors that affect demand on the Interstate System, but at the simplest level the geographic distribution of new population creates significant variation in new demands on the Interstate System. Additionally, the disparity of new demands is ex- acerbated in cases in which there is a geographic redistribution of existing population (some areas showing declining population), further influencing the magnitude of new geographic demand. This differential growth was outlined in Appendix E, and its implications in terms of infrastructure needs are enumerated below. Overall growth forecasts are noted in Box C-1. Whereas demands on the Interstate System are affected by many more considerations than just adjacent population, travel demand remains highly correlated with population. At the highest level of geographyâ Census regionsâthe population growth in 2000â2016 has been markedly different across geography. As shown in Table C-1, the historic trend of strong growth in the West and South has continued this century, with these regions growing approximately four times as fast as the Northeast and FIGURE C-3 National VMT and VMT per capita, percentage change from 1992. SOURCES: FHWA 2017; U.S. Census Bureau 2016.
236 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Midwest. When looking at the same issues at the state level, one stateâ Michiganâhas shown a population decline since the beginning of the century. Six statesâ Arizona, California, Florida, Georgia, North Carolina, and Texasâcollectively accumulated more than 50 percent of the national population growth. Perhaps more relevant to the scale of transportation infrastructure needs is looking at growth trends at the county level, which provides additional insight regarding the distribution of population growth over geography. At that scale of geography, the disparity of growth rates and magnitudes is even more pronounced. In total, 12 percent of the 377 coun- ties are responsible for more than 91 percent of the national growth in population since 2000. These counties each had growth in excess of 25,000 persons, suggesting pressure on the roadway and Interstate System in those counties and, most probably, the interconnection of those counties with adjacent geographic areas. Table C-2 enumerates the growth trends since 2000 for the counties and District of Columbia. BOX C-1 U.S. Population Projections The Census Bureau develops long-range population projections, the most recent of which was in 2014 and goes through 2060. The rate of population growth de- clines over time in absolute annual increases from approximately 2.6 million per year currently to an estimated approximately 1.9 million per year in 2060. The increases range from approximately 0.8 percent per year currently to less than 0.5 percent per year in 2060. For context, there are currently approximately 688 lane-miles of Interstate for each million residents. Approximately 41 percent of population growth in the United States since 2010 is attributed to net international in-migration. Going forward, future growth is projected to be even more dependent on net immigration; hence, it will be influenced by policy and economic conditions. SOURCE: https://www.census.gov/data/datasets/2014/demo/popproj/2014-popproj.html. TABLE C-1 U.S. Population Change by Census Region Census Region Net Change 2000â2016 (%) Northeast 4.9 Midwest 5.5 South 22.0 West 21.3 SOURCE: U.S. Census Bureau 2016, Table B01003.
APPENDIX C 237 Notice that 1,295 countiesâ41 percent of the totalâhad declining population since the turn of the century, and 15 had declines of more than 25,000. These population declines would certainly suggest a lessening of pressure for transportation capacity expansion and, perhaps more signifi- cantly, undermine the economic base over which transportation infrastruc- ture investments can be supported. Although the nonuniformity of growth exacerbates the criticality of a substantial capacity expansion, this need is further heightened in situations in which there is redistribution of the exist- ing population (and corresponding travel demand). In the extreme, it cre- ates the possibility of existing capacity in growth areas being overwhelmed while existing capacity in declining areas may be underutilized. Since 2000, national population has increased by 42 millionâapproximately 15 per- cent. In addition, net population relocation of approximately 8 percent of the population from declining counties to growing counties resulted in a total of a 23 percent increase in population occurring in 60 percent of U.S. counties, with the vast majority of that growth occurring in just 12 percent of the counties. Figure C-4 is a visual representation of the geographic disparity of population growth across the United States. TABLE C-2 U.S. County Population Growth Trends, 2000â2016 Growth Category Number of Counties Sum of Change Percentage of Counties Percentage of Growth Counties that grew more than 25,000 377 37,783,846 12.00 91.31 Counties that grew more than 5,000 to 24,999 439 5,160,638 13.97 12.47 Counties that grew more than 1,000 to 4,999 574 1,443,389 18.27 3.49 Counties that grew less than 1,000 457 189,584 14.54 0.46 Counties that shrank from 1 to 999 749 â325,475 23.84 â0.79 Counties that shrank from 1,000 to 4,999 472 â97,254 15.02 -2.41 Counties that shrank more than 5,000 to 24,999 59 â577,965 1.88 â1.40 Counties that shrank more than 25,000 15 â1,299,029 0.48 â3.14 Total 3,142 41,377,734 SOURCE: U.S. Census Bureau 2016, Table B01003.
238 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Urban Versus Rural Interstate Demand Insight into the nature of Interstate demand growth also can be gleaned from looking at the trends with respect to Interstate System availability and utilization in urban and rural areas. As can be seen from Table C-3, system extent and growth are greater in urban areas, and urban area Interstate System volume growth is substantially greater than for rural areas. Urban Interstates comprised 0.5 percent of centerline miles of roadways and 1.2 percent of lane-miles of the roadway system and carried 17.5 percent of the roadway system volume in 2015. Those shares all increased since prior periods. Figure C-5 portrays the relative role of urban and rural Interstates in accommodating VMT. Part of the irregularity of the trends is due to the postâdecennial Census recategorization of urbanized areas (more geography around growing metro areas is classified as urban). Nevertheless, the trends are relatively clear, with urban Interstates playing an increasingly important role in accommodating VMT while the role of rural Interstates is dimin- ished in terms of share of volume. Figure C-6 shows the trend in terms of VMT carried by the components of the roadway system. The role of the Interstate System has continued to increase over time. Interestingly, demand softened in approximately 2002â 2007, whereas volume softened on the non-Interstate highways. As Figure C-7 portrays, volume per lane-mile on the urban Interstate System indicates that it is approximately three times that on the rural FIGURE C-4 Population growth variation across U.S. counties. SOURCE: U.S. Census Bureau 2016, Table B01003.
APPENDIX C 239 TABLE C-3 Changes in U.S. Interstate Extent and Use, 1980â2015 Centerline Milesa Lane-Milesb VMT (millions)c Urban Interstate Change 9,848.1 56,279.5 379,944.3 Percent 106.9 116.1 235.6 Rural Interstated Change â2,914.9 â12,728.3 100,681.6 Percent â9.14 â9.7 74.53 aFHWA 2016b, Table HM-220. Includes 50 states and District of Columbia. bFHWA 2016b, Table HM-260. Data are based on state highway agency estimates reported for various functional systems and include 50 states and District of Columbia. For 1980â1992, the Interstate system is based on 100 percent inventory; non-Interstate arterial and collector functional systems are estimated from sample data; urban and rural local functional systems are estimated assuming two through lanes. For 1993â1995, the Interstate system, other freeways and expressways, and other principal arterial functional systems are based on 100 percent inventory; minor arterial, urban collector, and rural major collector functional systems are estimated from sample data; rural and urban local and rural minor collector functional systems are estimated assuming two through lanes. cFHWA 2016b, Table HM-202. Data based on state highway agency estimates reported for various functional systems; includes 50 states and District of Columbia. dThe decline in centerline miles and lane-miles for rural Interstate attributable to reclas- sification of roadway segments to urban. As urban areas expand, more geography is classified as urban. FIGURE C-5 Role of urban and rural Interstate highways in accommodating VMT. NOTE: Data are based on state highway agency estimates reported for various functional systems and include 50 states and District of Columbia. SOURCE: FHWA 2016b, Table VM-202.
240 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM FIGURE C-6 Role of urban and rural Interstate in accommodating VMT. NOTE: Data are based on state highway agency estimates reported for various functional systems and include 50 states and District of Columbia. SOURCE: FHWA 2016b, Table VM-202. FIGURE C-7 Average daily VMT per lane-mile, urban and rural. NOTES: Data based on state highway agency estimates reported for various func- tional systems; includes 50 states and District of Columbia. For 1980â1992, the Interstate system is based on 100 percent inventory; non-Interstate arterial and collector functional systems are estimated from sample data; urban and rural lo- cal functional systems are estimated assuming two through lanes. For 1993â1995, the Interstate system, other freeways and expressways, and other principal arterial functional systems are based on 100 percent inventory; minor arterial, urban col- lector, and rural major collector functional systems are estimated from sample data; rural and urban local and rural minor collector functional systems are estimated assuming two through lanes. SOURCE: FHWA 2016b, Tables VM-202 and HM260.
APPENDIX C 241 system. In 1980â2015, the daily volume on each lane-mile of urban Inter- state increased from 9,116 to 14,156, or 5,040 per lane-mile, a 55 percent increase. The rural Interstate volume increased from 2,826 to 5,462, about half as muchâ2,636 per lane-mileâbut an increase of 93 percent. As of 2015, daily volumes remained approximately 5 percent below peak levels from the prior decade. Implications of Nonuniform Distribution of New Demands for Capacity Although this review provides insight into the potential geographic distribu- tion of growth and transportation demand at the county level and between urban and rural areas, the fundamental issue of the geographic correspon- dence of transportation capacity relative to demand is relevant at a smaller geographic scale, as revealed by facility-specific demand forecasts. Nonetheless, understanding the recent and forecast population distri- bution and subsequent transportation network impacts requires two key points. First, the distribution of demand across geography affects the na- ture and magnitude of capacity needs. Second, declines in demand in some geographic areas can exacerbate the challenges of accommodating growth, because some existing system links may have declining utilization whereas other areas might see dramatic increases in demand, accommodating both disproportionate shares of growth and redistribution of population. The fact that demand growth has not been uniform across the In- terstate System, nor is it likely to be in the future, has implications for future infrastructure needs. Need for new capacity is most likely to occur in urban high-growth areasâareas in which congestion, higher costs, and challenging right-of-way availability affect the ability to respond to grow- ing demands. Although this discussion is in the context of differential population growth, differential incidence of other conditions known to influence travel demand also may occur. For example, travel demand has long been cor- related with economic health, and economic health may not be uniformly distributed by geography or population. Areas that are growing relatively wealthier might experience more rapid growth in travel demand. Differ- ential trends in various sociodemographic characteristics related to travel demand, technology deployment, and other factors also might contribute to differential demand growth across the system. Factors Influencing Travel Demand In exploring future demand, it is important to touch on our understand- ing of factors that influence travel levels. The criticality of travel for social
242 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM and economic interaction is well established in history and evidenced by the significant role that transportation plays in the economy and the fabric of society. It has been hypothesized that growth in travel demand associ- ated with individuals is attributable to growth in income and growth in knowledge. As characterized in Figure C-8, growth in income and knowl- edge lead to specialization in consumption and activity or use of time, which leads to an increase in the demand for travel and communication. In contemporary terms, a homemaker receives training in a specialized skill area and joins the workforce. Her children are dropped off at daycare and laundry is dropped at the cleaners; prepared meals are purchased on the way home from work. Generic white bread at the local grocery is replaced with whole-grain bread from a more distant natural foods store. The circle of social relationships expands from adjacent neighbors to work colleagues and parents met through childrenâs organized activities. The household has more highly specialized professional, consumer, and social activities, more income, and more travel. This fundamental phenomenon has contributed to significant growth in per capita travel over the past decades and remains central to our under- standing of growth in travel demand. However, beyond these fundamental drivers of demand, the supply and performance of travel options, and, more recently, the ability to substitute communication for travel factor into our considerations of future travel demand as, of course, does basic growth in population. FIGURE C-8 Characterization of drivers of travel demand.
APPENDIX C 243 This section explores factors that researchers have identified as influ- encing the demand for travel. The objectives are to 1. Identify factors that are believed to underlie travel demand going forward, 2. Assess the degree of understanding of their influence on travel de- mand, and 3. Explore the state of knowledge regarding those conditions going forward such that we can speculate on the magnitude of their im- pact on travel. As shown in Figures C-1 and C-2, U.S. historic trends in total roadway VMT and the trend in total VMT per capita have changed markedly since earlier in this century. A number of multidecade trends have played out or at least moderated significantly. This includes growth in real income, growth in auto ownership, suburbanization of population and employ- ment, shifts from alternative modes to single-occupant auto, and growth in trip making accompanying income growth and labor force participation (Polzin 2006). Growth in travel demand decoupled from both population growth (see Figure C-1) and GDP growth (see Figure C-10) as changes in the economy, demographics, and technology altered long-term relation- ships. Trend extrapolation of linear relationships had served well for gaug- ing future VMT but no longer appears appropriate. Perhaps more critical, there is a growing recognition that our understanding of travel demand was perhaps superficial and not up to dealing with emerging trends such as communication substitution for travel, new business models for delivering mobility, new demographic and economic conditions, rapid and significant fuel price changes, and policies and behaviors influenced by environmental and climate change considerations. This pace of change and uncertainty regarding the nature and magnitude of future travel demand heighten the desire to more fully understand future travel. Figure C-9 represents one framework for discussing future travel and is intended to accommodate all types of travel that use roadways, including person and freight travel as well as commercial vehicles, tourists, etc. It is recognized that there are trade-offs between roadway modes and alternative travel modes (rail, air, water) for both passengers and freight movement. In general, demand factors (blue boxes in Figure C-9) change relatively slowly because they include such factors as sociodemographic characteris- tics and land use distribution, all of which, at the national scale, do not change quickly, with perhaps the exception that culture and value fac- tors could shift markedly in response to significant events. The demand factors shown in Figure C-9 are widely acknowledged in the transporta- tion literature, with perhaps the exception of the box embracing business,
244 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM governance, and institutional context. These factors are known to influence travel demand and mode choice, hence providing a basis for forecasting fu- ture demand. Supply factors (shown in the gray box) include characteristics of the transportation system and reflect important factors such as fuel cost, which influences demand. Table C-4 itemizes the estimated breakdown across vehicle types and urban/rural contexts. The best available evidence of the composition of roadway travel demand indicates that approximately 76 percent of national VMT is attributable to household-based person travel, 10 percent to heavy freight travel, 2 percent to public vehicles (e.g., school buses, police cars, emergency response vehicles), and 12 percent to commercial vehicles, which include local freight distribution vehicles, utility vehicles, and vehicles pro- viding services from package delivery to lawn services. Relationship of Supply and Demand It is not uncommon for policy makers to seek insight regarding future travel demand or VMT while implicitly or explicitly presuming that demand can be determined independent of transportation supply. In reality, a host of characteristics, such as time and money cost of travel and available mode FIGURE C-9 Framework for exploring factors influencing travel demand. Land Use Context ï· Regional/national distribution ï· Density ï· Mix of land uses ï· Urban form ï· Urban/network design ï· Contiguousness of development Travel Demand ï· Local person travel ï· Tourism/long trips ï· Freight ï· Commercial travel Socio-Demographic Conditions Household/Person Characteristics ï· Income/wealth levels and distribution ï· Age/activity level ï· Culture/values ï· Racial/ethnic composition ï· Immigration status/tenure ï· Gender ï· Family/household composition Transportation Supply/Performance ï· Modal availability and non-travel options to carry out activity ï· Modal performance o Cost o Speed/congestion o Safety, security o Reliability o Convenience o Image, etc. o Flexibility o Environmental impact o Multi-tasking opportunity Legal/Political Climate Culture Technology Security Economy Business, Governance, Institutional Context ï· Scale of activity concentration ï· Economic structure of service delivery (healthcare, education, government services, etc). Travel Impacts: 1. Change Trip Frequency 2. Change Destination 3. Change Mode 4. Change Path
APPENDIX C 245 choices, are elements of supply that will influence travelersâ actual extent and means of travel. As with consumption of any product or service, there is interplay between supply and demand. Most obviously, cost to travel in both time and money influences the demand for travel. This fundamental metric remains extremely relevant in an era in which energy prices are fluctuating dramatically, technology may increase auto costs significantly, or alternatively, shared use may reduce auto costs, as might reductions in insurance, medical, or property damage costs attributable to technology improvements enhancing safety. Additionally, the prospect that automa- tion will enable travelers to use in-vehicle travel time for purposes other than driving could influence perceived travel costs significantly. Meanwhile underinvestment in infrastructure and transportation services could result in roadway conditions and congestion that could increase travel time and vehicle operating costs. Other aspects of supply also influence demand. For example, the in- troduction of or improvements in transit can increase travel by individuals who do not have private-vehicle mobility options. The prospect of self- driving vehicles is hypothesized to increase travel for those who might be precluded from driving due to mental, physical, or legal/age constraints on driving. Critical aspects of the boxes in Figure C-9 will be discussed in sections below focusing on those factors believed to be most significant influences on future VMT. Sociodemographic and Economic Conditions The traditional focus of VMT forecasting has centered on sociodemo- graphic and economic conditions and travel behavior. This area remains relevant to issues such as an aging population, urbanization, differential millennial behaviors, income distribution, increased diversity, and other fac- tors influencing travel behaviors and travel demand levels. The significance of the geographic distribution of growth was discussed previously. TABLE C-4 Shares of Vehicle-Miles Traveled by Market Segment Light Vehicles (%) Household Based Public Vehicle, Utility, Service Based All Light Vehicles Heavy Vehicles (%) Total (%) Urban 55.52 9.04 64.56 5.45 70.0 Rural 21.90 3.56 25.46 4.54 30.0 Total 77.42 12.60 90.02 9.98 100.0 SOURCES: AASHTO 2013, Table 2-1; FHWA 2015, Tables VM2 and VM4.
246 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM VMT and Economic Activity Not surprisingly, VMT is highly correlated with economic activity. As shown in Figure C-10, the long-term trend in GDP and VMT are highly correlated. Understanding the importance of this relationship is helpful in forecasting VMT; however, the difficulty in forecasting GDP is widely recognized, and hence, understanding the strong relationship is not neces- sarily helpful, absent a way to forecast GDP with confidence. The challenge is compounded by evidence that the GDPâVMT relationship has started to change over the years, with VMT growth not as strongly correlated with GDP. The growth in information and service industries, which are less transportation-intensive than are other industries, is likely contribu- tory to this weakening relationship. Table C-5 itemizes the importance of transportation to various industry categories. The economic sectors that have been and are predicted to continue growing strongly tend to be less transportation-intensiveâfactors that contribute to the lessening signifi- cance of GDP to VMT levels. As noted in Table C-4, the vast majority of VMT comprises household- based travel as individuals carrying out social and economic activities. FIGURE C-10 National VMT and GDP trends. SOURCES: BEA 2017 (current-dollar and ârealâ GDP table); FHWA 2016b, Table VM-202.
APPENDIX C 247 Household income is highly correlated with travel levels, as income relates to workforce participation (commuting trips) and work-derived income enables the social and retail activities that drive household travel demand. Household survey data indicate that the moderation in household travel demand was associated with those households in which resource con- straints appear to be more critical to constraining travel demand. Low- and moderate-income householdsâhouseholds that have not seen real income growthâhave had their travel demand influenced by economic conditions. Figure C-11 shows the relationship between household income for the bottom 80 percent of households and national total VMT. As is apparent, there is a strong correlation between household income in this 80 percent of households and total national VMT. These households may have latent demand for additional travel should more household income become avail- able. Travel-generating activities such as vacations, shopping, recreation, and eating out are highly related to discretionary income and, should in- come growth ramp up for this large share of the population, VMT would expand more rapidly. Collectively, the implications of Figures C-10 and C-11 is that future VMT will be influenced by the nature of business growth and by the dis- tribution of income across the population. Should various activities such as increased manufacturing activities, growth in infrastructure investment, higher minimum wages, or economic growth that results in inflation-beat- ing income growth for low- and moderate-income households occur, the VMT impact may be more pronounced than if current trends continue. TABLE C-5 Transportation Intensiveness of Economic Sectors Sector Amount of Transportation Required to Produce $1 of Output (2014) Contribution to GDP (2015, billions) Natural resources and mining 4.2Â¢ $500.9 Utilities sector 4.6Â¢ $288.3 Construction 3.8Â¢ $716.9 Manufacturing 3.7Â¢ $2,167.8 Wholesale and retail trade 9.9Â¢ $2,130.1 Service Information Financial Professional/business Education and health Leisure and hospitality Other 1.5Â¢ 0.8Â¢ 2.8Â¢ 1.6Â¢ 3.2Â¢ 2.9Â¢ $9,291.7 Government 4.7Â¢ $2,323.6 SOURCE: BTS 2015.
248 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Population Characteristics and Travel Behavior Changing population characteristics, specifically the movement of the large baby boom cohort toward retirement and the entrance of the even larger millennial generation into young adulthood, have been the basis for signifi- cant speculation regarding travel demand implications. Aging baby boom- ers are passing their peak travel years (ages 35â55) and entering the point in their life cycle when travel declines. Declining baby boomer travel demands may be associated with less travel in support of dependent household mem- bers (more likely to be empty-nesters), the prospect of no longer commuting in retirement, and the prospect of declines in physical stamina, discretionary income in retirement, and diminished motivation to accumulate material items. This generation of older adults has higher licensure levels and greater travel levels than prior older adult generations, but diminished travel from earlier in their life. Toward the other end of the age spectrum, millennials have evidenced notably different travel levels than prior generations of young adults, af- fected by significantly higher levels of pursuit of advanced education and delays in marriage, household formation, starting families, and beginning careers. This age cohort is substantially more urban than prior generations and has markedly higher levels of continuing to live in a parentâs household. They use social media to satisfy some of their social interaction needs and FIGURE C-11 National VMT and household income of bottom 80 percent of U.S. households. SOURCE: FHWA 2016b, Table VM-202.
APPENDIX C 249 substitute communications for travel via online shopping, downloading music and videos, and interactive video gaming via the Web. This genera- tion reached adulthood during a recession and, in many cases, accumulated substantial education debt, which slowed their pace of attaining vehicles and independent living. More recent data as the millennials age and the economy improves suggest that this generation is evolving toward more typical travel characteristics as they age and as economic constraints to travel and more travel-intensive lifestyles are diminished. The extent to which their behaviors will continue to vary from historic norms remains to be seen, as does the extent to which subsequent generations will have different travel behaviors. As always, caution should be taken in general- izing behavioral differences across contexts because there is a great deal of diversity within the millennial generation in education levels, residential locations, and travel behaviors. Historically observed race and ethnicity differences in travel might lead to the expectation of moderation in travel demand as the population becomes more diverse. However, many of the travel behavior differences across populations are better explained by location decisions and economic status, because residual differences attributed to race and ethnicity have become more modest over time. Travel behavior predominantly reflects economic conditions and the land use and transportation context. As addi- tional data are assembled and the full impact of the recent economic condi- tions are more comprehensively understood, forecasters will have a stronger basis on which to gauge the impacts of changes in the sociodemographic composition of the population. Land Use Context One of the more powerful influences on travel behavior is the intensiveness of activities or the density and mix of land use. As noted in Figure C-9, a variety of traits associated with land use influence travel. Most obviously, the distance of travel required to access various activities is lessened in environments with more intensive development. Although some of the advantage of proximity is offset by the prospect of additional destination choices spurring more travel, it has long been recognized that routine household-serving activities and their associated travel are minimized in more densely developed urban environments. In addition, these environ- ments make alternative travel means such as biking, walking, and public transit more viable thus further mitigating VMT. Table C-6 summarizes differential per capita VMT for the age 20â39 cohort based on residential location type. These data do not adjust for in- come or other considerations including self-selection but clearly communi- cate the differential VMT per capita as a function of land use intensiveness.
250 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM As alluded to in Figure C-9, a host of characteristics of land use, including mix of activities, nature of the transportation network, physical design of transportation elements, presence of alternative mode options, and other factors contribute to the portfolio of land useârelated characteristics that collectively influence travel. The economic characteristics of business and government also influ- ence travel. The trend toward more-specialized activity types and larger- scale facilities (schools, hospitals, retail centers, etc.) offset some of the proximity advantages of more intense development. Similarly, urban land rent distributions can offset the benefit of intensive development by forcing separation of activities by income levels (low-wage service workers travel long distances for employment because of lack of affordability of nearby residential locations). In the context of long-range VMT forecasts and Interstate highway travel demand, continued urbanization suggests downward pressure on VMT growth. However, the large declines in rural population may have played themselves out, and the full VMT impact of travel patterns within emerging megaregions with continued activity specialization and housing affordability considerations may dampen some of the VMT savings his- torically associated with urban development patterns. As public housing is dispersed and new high-density central city development often is targeted TABLE C-6 Per Capita VMT by Location Type Urban Continuum Daily VMT per Capita (Ages 20â39) Urban 18.0 Second city 23.1 Suburban 27.1 Town and country 32.7 Location in Urbanized Area In an urban area 24.1 In an urban cluster 25.7 In an area surrounded by urban areas 32.9 Not in urban area 35.2 Size of Urbanized Area 1 million + with subway or rail 20.2 1 million + w/o subway or rail 25.8 500,000â999,999 27.9 200,000â499,999 24.7 50,000â199,999 25.9 Not in urbanized area 32.4 Urban/Rural Urban 24.3 Rural 35.2 SOURCE: Polzin et al. 2014.
APPENDIX C 251 to high-income residents, historic empirical relationships between density and travel may be changing. Influence of Transportation Supply and Performance The gray box in Figure C-9 itemizes aspects of transportation choices and characteristics that influence the demand for travel. Figure C-12 character- izes, in greater detail, factors that are frequently recognized as things that influence travel behavior. In reality, the amount of travel both in total and on the Interstate High- way System will be influenced by the characteristics and performance of the system and of the alternative choices available to travelers as noted in Box C-2. Of course, one choice is to travel less. If the cost of transportation is particularly high in either time or money, travelers may choose to forgo trips, make shorter trips to closer destinations, group trips into chains, or travel on different modes or paths. Thus, the demand on the Interstate System will depend on everything from the price of fuel to the relative con- gestion on the Interstate versus parallel arterials to the presence of a transit alternative or the opportunity to substitute communication for travel. Although a multitude of factors make predicting travel demand com- plicated and context specific, there is a generalized understanding of how demand responds to the critical factors of changes in travel in terms of cost of both money and time. These relationships, referred to as the elasticity FIGURE C-12 Characteristics that influence travel decisions.
252 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM of demand as a function of cost and time, compare the percentage change in travel demand in relationship to the percentage change in travel cost or travel time. These relationships give insight into how changes in travel time associated with congestion or circuity (less direct travel path) or changes in travel cost associated with such things as tolls or higher fuel costs might affect demand. They also give insight into the travel consequences of deteriorating performance of the Interstate System should that translate into more congestion (slower travel time equals higher travel cost). In a broader context, cost-competitiveness of Interstate travel will be influenced by other considerations such as the competitiveness of airline, intercity rail, and intercity bus travel options and the comparative performance of rail freight versus truck travel choices. Additionally, the extent to which automated vehicles change the travel costs and perceived onerous of time spent traveling for future urban and intercity travelers will also influence the cost-competitiveness of Interstate travel. Box C-3 highlights the concept of induced travel, explaining the con- cepts and the magnitude of the consequences. Given congestion levels on much of the urban Interstate System, these data suggest that significant additional Interstate VMT would accompany strategic capacity expansions. Time and money cost of travel are recognized as very significant in ex- plaining the mode and amount of travel. Fuel costs are recognized as one of the important cost considerations. Box C-3 reflects research findings on how fuel prices affect VMT. Other vehicle ownership and use costs such as the capital cost of vehicles, financing costs, maintenance costs, and parking and other associated costs also can contribute to decisions on owning and using vehicles. Affordability of insurance in urban environments may be among the factors that affect licensure and driving rates for young people. Similarly, urban environments in which there are parking costs at many destinations, and particularly if there are marginal costs associated with parking a vehicle at oneâs place of residence, can deter vehicle ownership and use. BOX C-2 Generic Travel Choices When Faced with Higher Travel Costs â¢ Forgo travel or substitute communication for travel. â¢ Reduce travel extent by choosing closer destinations or chaining trips together. â¢ Shift travel time (e.g., peak hour to off-peak). â¢ Choose different travel path (e.g., highway or arterial versus Interstate System). â¢ Choose different means of travel (e.g., shift from driving to riding as passenger with others, flying, transit/rail, or bike).
APPENDIX C 253 BOX C-3 Induced Travel Induced travel is the increase in usage of a transportation facility due to a reduc- tion in the cost of travel that results from external changes (e.g., capacity expan- sion to an existing highway). The basic economic theory of supply and demand explains the existence of induced travel: adding capacity decreases travel time and lowers the cost of driving; when the cost of driving goes down, the quantity of driving goes up. Induced travel includes both newly generated travel and travel diverted from other travel paths or destinations. The nature and degree of induced travel may differ in the short run versus the long run. In the short run, generated travel includes longer trips and new trips, and diverted travel includes shifted trips from other routes, times, and modes. In the long run, activity location changes may occur as a result of capacity expansion, and these locational changes can lead to additional travel. The degree of induced travel differs not only between the short versus long runs but also across different scales. If the subject is a single facility, induced travel will appear large in relation to previously existing traffic because induced travel includes both generated travel and travel diverted to the now superior facil- ity. At the regional level, however, induced travel would be smaller as the impact of diverted travel is netted out. At the same time, however, the amount of generated travel due to expanding a single facility is bigger at the regional level because the larger geographic area captures additional generated travel on feeder routes that tie to the expanded facility. The degree of induced travel depends on the level of congestion on the subject facility before expansion. If there is not latent demand, as evidenced by congestion, there will be no induced demand. Induced travel in general is greater from expanding a facility that is more congested. A large of body of empirical evidence confirms the existence and magnitude of induced travel in the case of highway travel at the regional or state levels. A recent review of the evidence suggests a range of 0.3â0.6 for the short-run elas- ticity of VMT with respect to highway lane-miles. This indicates that a 100 percent increase in roadway lane-miles could result in a near-term 30â60 percent increase in VMT. The long-run elasticity is estimated to range from 0.6 to 1.0, indicating that roadway expansion in congested environments might ultimately produce 60â100 percent more VMT as travelers take advantage of the new capacity in the short term and perhaps made residential and travel destination decisions in the long term that further increased their travel. Because of the existence of induced travel, congestion reduction from a given capacity expansion would be lower than otherwise. That is, travelers may still enjoy faster travel but not as fast as they may have expected. At the same time, however, travelers can benefit from traveling at more preferred times and modes, taking more preferred routes, going to more preferred destinations, do- ing more preferred activities, living and working at more preferred locations, etc. Similarly, freight and commercial traffic would benefit from the enhanced mobility as well. SOURCES: Handy and Boarnet 2014; Lee 2002; Noland and Hanson 2013; Pickrell 2001.
254 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Time cost of travel is relevant to future travel demand in two impor- tance ways. First, roadway congestion associated with demand outgrowing capacity can result in additional travel time expenditures by travelers. This additional cost can deter the extent of travel and result in various deci- sions that would dampen VMT. Researchers have estimated the elasticity of travel with respect to changing travel time as being approximately â0.5 (see Box C-4). For example, a 10 percent increase in travel time to work because of growing congestion on the roadway would be expected to reduce VMT by 5 percent. The second major aspect of travel time relevant to future forecasts of demand relates to the prospect that in an era of self-driving vehicles, travelers would consider time spent traveling less onerous if they were able to carry out other activities such as sleeping, reading, or pursuing work or entertainment via a digital device. This has been hypothesized to create sig- nificant additional travel, because travelers are more willing to make travel and location decisions that increase travel if they can spend their travel time in more productive pursuits than driving. This issue is discussed in the sec- tion below on the impact of self-driving vehicles on VMT. As noted in Figure C-10, other transportation system or mode con- siderations also influence the amount and mode of travel. Relative energy efficiency, relative safety, reliability, flexibility, social impacts, and other factors influence travel behavior. Modal safety has generally been improving as has energy efficiency. To the extent that new technologies can continue to improve performance and mitigate the negative consequences of travel (safety, energy use, emissions, noise, etc.), one might expect a positive bias BOX C-4 Travel Time Elasticity of VMT Short-run: â0.38 Long-run: â0.68 SOURCE: Lee and Burris 2005, C-14 Short-run: â0.5 Long-run: â1.0 Urban Rural Short-run â0.27 â0.67 Long-run â0.57 â1.33 SOURCE: Litman 2017, 48.
APPENDIX C 255 toward greater travel. Some of these aspects are further elaborated on in the discussion of self-driving vehicles. Impact of New Technologies and Self-Driving Vehicles on Future Travel Demand This topic area is critical to understanding future VMT. Most unique about self-driving is the fact that it is hypothesized to influence travel demand by virtue of changing both the time and the monetary cost of travel as well as by changing the capacity of the transportation infrastructure system. Thus, the consequences of automation and self-driving are complex and uncertain. In addition, well-credentialed experts have dramatically varying perspectives on the time frame, pace, and consequences of deployment. A host of issues ranging from the pace of technology development and refine- ment to the legal and political context for deployment to market acceptance and economic considerations are likely to influence the magnitude of the impact of technology on transportation. Figure C-13 and Box C-5 highlight some of the anticipated consequences of additional technology deployment in transportation. The complementary and competitive nature of the vari- ous impacts contribute to the complexity and uncertainty of understanding the ultimate impacts. The cost change factors can influence both the total quantity of travel as well as the modal distribution. Mobility services (trans- portation network companies or driverless vehicle services) will be com- peting with public transit, traditional traveler-owned and -driven vehicles and perhaps shared-ownership vehicles for urban trips. There is also likely to be competition for short- to moderate-distance intercity bus, rail, and short-haul air services. Various research initiatives have begun to explore the possible conse- quences of automation on travel demand; Table C-7 enumerates several of these studies. Most often, the studies test scenarios of various conditions and deployment extent to gauge the transportation consequence. As the VMT impact column indicates, results vary widely. In addition, the vast majority of discussions of self-driving vehicles are restricted exclusively to local household-serving person travel. Public and commercial vehicle travel, freight travel, and long-distance travel seldom are incorporated into the respective analyses. The reported results would be relevant only in the context of urban environments and household-based travel. Freight and Commercial Traffic Travel Demand Somewhat distinctive from person travel, freight and commercial travel demand can be affected by different factors. Heavy freight constitutes ap- proximately 10 percent of total VMT, and the remaining commercial and
256 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM public vehicle travel contribute an additional 14 percent. The impact of this VMT is disproportionate to its share of travel by virtue of the fact that these typically larger vehicles, with slower acceleration and deceleration characteristics, have higher levels of energy use and emissions and influ- ence roadway capacity and condition considerably more than do individual personal vehicles. Many of the same considerations that influence person travel will similarly influence freight and commercial travel activity. Demographic distribution, economic conditions, congestion levels, fuel price, intermodal competition, and the prospect of automation are factors that will influence future freight travel demand. Given an advanced economy in which eco- nomic growth is strongest in service and information areas, it is unlikely that growth in freight and commercial travel demand will vary dramati- cally from growth in person travel. In the opinion of this author, the great- est risks for differential growth in roadway freight demand will occur if economic conditions enable broad adoption of very quick deliveries; if the fundamental role of the United States in the global economy with respect to mining, manufacturing, and agricultural production changes significantly; or if technologies incentivise significant mode shiftsâfor example, from rail shipments to truck trains. Regional and facility-specific high-volume longer-distance roadway freight volumes will be influenced substantially by broader logistics and business strategies relating to intermodal and ware- housing facility locations. Thus, geography-specific assessment of freight roadway volumes will be critical in appropriately accounting for freight impacts on Interstate VMT, and, even then, private-sector market decisions may affect line-haul volumes; for example, a business decision of a large international logistics firm in the business of consolidating global freight might significantly alter the volumes emanating from a given port, terminal, or warehousing location. FIGURE C-13 Anticipated consequences of additional technology deployment in transportation. Impacts on Transportation System Demand Impacts on Transportation System Capacity Technology Deployment and Self-Driving Vehicles
APPENDIX C 257 BOX C-5 Technology Impacts on Transportation System Capacity â¢ Greater vehicle occupancyâhence, fewer vehicles requiredâby virtue of technology-enabled ridesharing; â¢ Increased volume by virtue of minimizing incidents and incident delays/ congestion; â¢ Increased volume by virtue of smoothing vehicle flow, which increases throughput; â¢ Increased volume by virtue of closer following distances; â¢ Increased volume through optimized intersection signal systems and lane management; â¢ Increased throughput by virtue of narrower lanes, enabling additional vehicle capacity/lanes on some facilities; â¢ Optimized vehicle logistics/trip circuity via navigation capabilities; â¢ Reduced travel via substitution of communication for travel; â¢ Increased travel via empty shared self-driving vehicles shuttling between vehicle assignments, parking, and service terminal locations; â¢ Increased travel via reduced travel money cost: â reduced insurance costs for safer vehicles, â reduced vehicle capital cost by virtue of sharing capital asset among multiple travelers, â reduced vehicle operating cost by virtue of sharing trips with multiple occupants, â reduced vehicle operating cost by virtue of electrification and logistics optimization anticipated for mobility service providers, â reduced parking/storage costs by virtue of vehicle sharing, â reduced per mile travel roadway infrastructure cost by virtue of greater utilization of facility via self-driving vehicles; â¢ Decreased travel via increased travel money cost: â increased vehicle capital cost due to inclusion of additional technologies and quicker obsolescence, â increased transportation infrastructure cost as infrastructure is modified to accommodate connected and automated vehicles, â increased travel cost if monopoly mobility service providers extract high profits from mobility services, â increased travel costs should reliance on mobility services require con- tingency investments to accommodate special situations such as emer- gency evacuations or operations in inclement weather; â¢ Increased travel via reduced travel time âcostâ: â reduced time cost by virtue of being able to do something of higher value during travel, â reduced time cost by virtue of faster travel speeds on managed or higher-capacity facilities; â¢ Decreased travel via increased travel time âcostâ: â increased travel time cost by virtue of vehicle arrival wait time and trip circuity for shared travel, â increased travel time cost by virtue of slower travel speeds due to in- duced travel from lower travel monetary costs, â increased travel time cost by virtue of slower travel associated with strict speed limit requirements and conservative vehicle interface behaviors.
258 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Forecasts of Future VMT Numerous research efforts have explored modeling VMT trends in an effort to understand future infrastructure needs, energy use, vehicle emissions, traffic accidents, and other considerations impacted by travel demand. Because demand is influenced by a multitude of factors, various strategies and components of demand can be modeled through a number of different means. Despite these numerous strategies, forecasters remain stymied by an inability to predict the fundamental input assumptions for many known explanatory variables such as GDP growth and energy pricing, to say noth- ing of trends in technology evolution and deployment and human travel behavioral changes. Few forecasters predicted the moderation in demand growth that started in the early 2000s, and even fewer predicted the mag- nitude of the bounce-back of the past few years. One of the widely cited uses of VMT forecasts is in identifying potential needs for infrastructure expansion. The Bottom Line Report series from TABLE C-7 Research on VMT Impacts of Automation Reference Primary Variables VMT Impact (%) Auld et al. 2017 100% automated vehicle (AV) penetration, 75% value of time reduction, +77% roadway capacity +28 Childress et al. 2015 65% value of time reduction, â50% parking cost, per mile auto cost of $1.65 â34.5 to +19.6 Correia and van Arem 2016 Free parking, 50% value of time reduction +17 to +49 Corwin et al. 2016 AV adaption and sharing +25 Gucwa 2014 50% value of time reduction +14 Fagnant and Kockelman 2014 Impacts of circuity and deadhead miles only +11 Harper et al. 2016 Induced demand from persons with mobility constraints +14 International Transport Forum 2015 Simulation, mode shifts +6.4 to +90.9 KPMG 2015 Vehicle occupancy scenarios and demographic changes Up to +130 Trommer et al. 2016 Reduced value of travel time, reduced travel time, lower cost of driving to 32% AV penetration +3â9 Wadud et al. 2016 Cost of travel +60
APPENDIX C 259 American Association of State Highway and Transportation Officials ( AASHTO) exemplifies that use. Based on empirical trend analysis and judg- ment, this method accommodates uncertainty by using scenarios to frame future infrastructure needs. That report specifically referenced scenarios with 1 percent per year and 1.4 percent per year rates of national growth in VMT (see Box C-6). Another significant report for Congress, the 2015 Status of the Nationâs Highway Bridges and Transit: Conditions and Perfor- mance (C&P Report), also addresses the issue of future VMT as a basis for understanding future system performance. That report referenced a national level forecast of 1.04 percent per year (FHWA and FTA 2016; see Box C-7). The Federal Highway Administration and Department of Energy de- velop various VMT forecasts for components of the total vehicle demand. Those forecasts disaggregate demand by vehicle class, are sensitive to vari- ous different input variables, and have various update frequencies and fore- cast time frames. A summary of some of these forecasts and their features are contained in Box C-8. BOX C-6 National Growth in VMT An annual investment of $120 billion for highways and bridges between 2015 and 2020 is necessary to improve the condition and performance of the system, given a rate of travel growth of 1.0 percent per year in vehicle miles of travel, which has been AASHTOâs sustainability goal and which represents the likely impacts of both population growth and economic recovery. If travel growth is at 1.4 percent per year, which carries forward the rate employed in the 2009 Bottom Line and is consistent with the long-term trend from 1995 to 2010 and has been indicated in recent months, then needed investment to improve the highway and bridge systems will be $144 billion per year. SOURCE: Pisarski and Reno 2015, 2.
260 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM BOX C-7 Future VMT as a Basis for Understanding Future System Performance Treatment of Traffic Growtha For the Highway Economic Requirements System (HERS) analysis in this report, growth in VMT is based on two primary inputs: HPMS section-level forecasts of future annual average daily traffic that states provide and a national-level forecast developed from a new FHWA model. The national-level forecast serves as a con- trol, which the sum of the forecast section-level changes in VMT must match. To match the national-level control, the section-level forecasts are scaled proportion- ally. For this report, the sum of the section-level forecasts yielded an aggregate average annual VMT growth rate of 1.42 percent that exceeded the national-level forecast of 1.04 percent per year and thus were scaled proportionally downward to match the national-level forecast. The national-level forecast includes separate VMT growth rates for light-duty vehicles, single-unit trucks, and combination trucks; these separate growth rates were applied in the HERS analysis. VMT in light-duty vehicles is forecasted to grow at 0.92 percent per year. VMT for heavy-duty vehicles is forecasted to grow at a rate more than twice that for light-duty vehicles (2.15 percent per year for single-unit trucks and 2.12 percent per year for combination trucks). The higher rate of forecast VMT growth for heavy-duty vehicles reflects a close relation- ship between heavy-vehicle VMT and economic output (GDP or gross domestic product). For the National Bridge Investment Analysis System (NBIAS), these fore- casts build off bridge-level forecasts of future average daily traffic that states provide in the NBI. The sum of the bridge-level forecasts yielded an aggregate growth rate of 1.46 percent per year; growth rates for individual bridges were adjusted downward to match the 1.04 percent control total from the national-level VMT forecast model referenced above. An underlying assumption applied in both HERS and NBIAS is that VMT will grow linearly (so that 1/20th of the additional VMT is added each year), rather than geometrically (i.e., at a constant annual rate). With linear growth, the annual rate of growth gradually declines over the forecast period. New National VMT Forecasting Modelb The Volpe National Transportation Systems Center developed the National Vehicle Miles Traveled Projection for FHWA. The documentation for the model version used for this forecast is posted at http://www.fhwa.dot.gov/policyinformation/ tables/vmt/vmt_model_dev.cfm. The current plan is to release revised national- level forecasts each May; this 2015 C&P Report relies on the 20-year national forecasts for the Baseline Economic Outlook from the May 2015 release. a FHWA and FTA 2016, 7â4. b FHWA and FTA 2016, 9â6.
APPENDIX C 261 BOX C-8 VMT Forecasting Methodologies and Forecasts Volpe Model and FHWA Forecasts Pickrell et al. (2014) developed a set of three models for FHWA to forecast future changes in nationwide VMT for each of three vehicle types: light-duty passenger vehicles, light-duty trucks, and combination trucks. These models were specified on the basis of economic theories and developed with time-series data beginning in 1966. VMT by light-duty vehicles is a function of personal disposable income per capita, fuel cost per mile, and consumer confidence. VMT by light-duty trucks is a function of consumer spending, residential construction activity, and fuel cost per mile. VMT by combination trucks is a function of GDP, fuel cost per mile, and Interstate centerline miles. Since 2014, FHWA has used these models annually to forecast nationwide VMT for a 30-year horizon. Its 2015 version, the forecasts were used as control totals in summing up state-provided VMT forecasts for DOTâs 2015 Conditions and Performance Report (FHWA and FTA 2016). FHWA relies on IHS Inc.âs spring release of the long-term economic outlook for the United States as input data for forecasting VMT. In its 2016 version for the baseline economic growth outlook (medium economic condition), FHWA forecasts an an- nual average growth rate of 0.47 percent for light-duty vehicles, 1.50 percent for single-unit trucks, and 1.87 percent for combination trucks, and 0.61 percent for all vehicles combined during 2014â2044 (FHWA 2016a). DOE Model and Forecasts As part of developing its annual energy outlook, the transportation module of DOEâs National Energy Modeling System also uses a modular system of VMT forecasting (EIA 2016). At the highest level, it consists of the Light-Duty Vehicle Fleet submodule and the Freight Transportation Submodule. â¢ Personal travelâprojects VMT per licensed driver as exponential re- gression function of fuel cost, disposable personal income per capita, employment rate for 16+, light-duty vehicles per licensed driver, and past VMT trends. â¢ Light-duty fleet vehiclesâexpands annual VMT per vehicle by vehicle type (cars or light trucks) and fleet type (private, government, or utility) with vehicle stock by vehicle, fleet, and engine technology fuel type (16 types). â¢ Travel by light commercial trucks (gross vehicle weight = 6,001â10,000 lb)âprojects VMT per truck by growing 1995 base value on relative an- nual growth rates between industry sector output and light commercial truck VMT. â¢ Freight truck VMTâexpands base year VMT per truck by number of trucks with additional adjustment. For its Annual Energy Outlook 2017 (EIA 2017), DOE projects an average annual growth rate of 0.70 percent for personal and light-duty fleet travel, 1.50 percent for light commercial truck travel, and 1.3 percent for freight truck travel during 2015â2050.
262 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Guidance Regarding VMT Growth for Future Interstate Initiative Policy Consideration Based on the body of information presented, a number of observations relevant to policy planning activities for future Interstate Highway System transportation investment can be gleaned: 1. It is clear that the current and anticipated future conditions create a significantly large amount of uncertainty regarding future travel demand and system capacity. Forecasters have struggled to an- ticipate significant changes in factors that influence VMT, such as technology changes, economic conditions, and demographic trends, and the fundamental relationship between some of these conditions and travel levels is not sufficiently well understood to be the basis for highly confident future forecasts. 2. Although demographic shifts, economic cycles, and volatile fuel prices have long influenced VMT trends, the impact of technology changes and their potential to transform the business models for delivering transportation and alter travel needs by virtue of com- munication substituting for travel exacerbate the challenges of anticipating future VMT trends. 3. The presence of high uncertainty favors a strategy of considering various scenarios of future demand as an input for future transpor- tation policy. The use of scenarios allows one to test the robustness of policy and investment initiatives to have intended consequences in light of possible different futures. 4. Existing quantitative models coupled with reviews of empirical trends provide an appropriate basis for testing the sensitivity of differing assumptions about the future. At the aggregate national level, VMT increases over the next two decades can reasonably be expected to range from the rate of growth of population (ap- proximately 0.7 percent per year) to a rate of 2.0 percent per year if modest economic growth is assumed. Sustained rates outside that range would be expected only if there are pronounced changes in the economy. Beyond 20 years, maturation and market penetration of self-driving vehicle technologies might suggest different rates of long-term growth in VMT and different assumptions regarding roadway capacities. 5. Although it might be desirable to have robust long-term forecasts in light of the length of time to plan and implement Interstate System capacity improvements, the practical reality is that the backlog of investment needs and costs and the impact constraints on building capacity for potential long-term needs, coupled with
APPENDIX C 263 uncertainty regarding future lane throughput capabilities, diminish the criticality of long-term needs as a prerequisite for investment decisions. Seldom are we building new capacity on our roadway systems for future demand; most often, we are constrained to build capacity for prior or current levels of demand that are not being adequately accommodated today. 6. VMT scenarios can inform discussion about future Interstate Sys- tem capacity requirements; however, financial and policy decisions, many made locally, will inevitably govern actual capacity expan- sion. The criticality of precision in VMT forecasts is muted by the fact that the Interstate System is part of a broader network of transportation facilities and services in which the consequences of failing to expand capacity are spread over a broader network. 7. Aggregate analyses at the national level are inherently different from those that consider specific facilities. In the case of the latter, more granularity is not only possible but critical. As noted above, the demand for Interstate System capacity is likely to be concen- trated on a relatively small share of the total system, and the total new demands will comprise demand associated with population and travel growth as well as that attributable to population redis- tribution. Thus, the percentage increase in additional capacity to maintain performance will be larger than the average percentage increase in roadway volumes. 8. Current trends suggest that demand growth will be concentrated in growing urban areasâareas in which there is significant latent demand for high-performance roadway travel, and, as such, new capacity will attract and induce significant volumes. Urban Inter- state highways interact and compete with other urban roadway and public transportation services and, as such, the consequence of increased demand will affect other elements of the transportation system and be affected by changes in those systems. Unless there is more active management and pricing of Interstate travel, urban Interstate System performance will be captive to overall corridor performance as travel demand shifts to the highest-performance travel paths. Similarly, urban Interstate performance is somewhat captive to local and state investments in other transportation fa- cilities and services in the respective corridors. Urban Interstates, particularly for large and growing urban areas, capture significant local travel and thus struggle to preserve their intended mission and functional classification. 9. Urban Interstate capacity expansion is the most complex and ex- pensive context for capacity expansion. In urban environments, policy considerations associated with the social, environmental,
264 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM and financial implications of capacity expansion will create huge challenges and place a premium on capacity expansion strategies that can be deployed within existing facility footprints. 10. The collective consequence of the various phenomena noted above create a context in which an annual <1 percent increase in popula- tion and 2 percent increase in GDP might create a 2 percent in- crease in VMT, which might create a need for a 3 percent increase in Interstate Highway System capacity, which might require a 5 or more percent increment in Interstate infrastructure asset value to sustain Interstate System performance. Thus, the aggregate increase in VMT underrepresents the infrastructure investment require- ments for the systemâs performance to be sustained, given the geographic location of demand growth and the financial and policy implications of expanding capacity in areas where new capacity is most critical to sustain the systemâs performance. The investment requirements will be even higher if the cost of consensus results in low-priority investments working their way into the program of improvements. REFERENCES Abbreviations AASHTO American Association of State Highway and Transportation Officials BEA Bureau of Economic Analysis BTS Bureau of Transportation Statistics EIA Energy Information Administration FHWA Federal Highway Administration FTA Federal Transit Administration AASHTO. 2013. Commuting in America 2013. http://traveltrends.transportation.org/ Documents/B2_CIA_Role%20Overall%20Travel_web_2.pdf. Auld, J., D. Karbowski, and V. Sokolov. 2017. Assessing the Regional Energy Impact of Con- nected Vehicle Deployment. Transportation Research Procedia. https://polaris.es.anl.gov/ pdf/WCTR_CAV_paper_v2.pdf. BEA. 2017. Gross Domestic Product. U.S. Department of Commerce, Washington, D.C. https://bea.gov/national/index.htm. BTS. 2015. Industry Snapshots: Uses of Transportation. U.S. Department of Transportation, Washington, D.C. Childress, S., B. Nichols, B. Charlton, and S. Coe. 2015. Using an Activity-Based Model to Explore Possible Impacts of Automated Vehicles. Paper No. 15-5118. Presented at the 94th Transportation Research Board Annual Meeting, Washington, D.C. https://psrc. github.io/attachments/2014/TRB-2015-Automated-Vehicles-Rev2.pdf. Correia, G. H. D. A., and B. van Arem. 2016. Solving the User Optimum Privately Owned Automated Vehicles Assignment Problem (UO-POAVAP): A Model to Explore the Im- pacts of Self-Driving Vehicles on Urban Mobility. Transportation Research Part B, Vol. 87, pp. 64â88.
APPENDIX C 265 Corwin, S., C. Giffi, J. Vitale, and N. Jameson. 2016. Gearing for Change: Preparing for Transformation in the Automotive Ecosystem. Deloitte University Press. https://du- press.deloitte.com/content/dam/dup-us-en/articles/3474_Future-of-mobility-gearing-for- change/DUP_Future-of-mobility-gearing-for-change.pdf. EIA. 2016. Transportation Demand Module of the National Energy Modeling System: Model Documentation. U.S. Department of Energy, Washington, D.C. https://www.eia.gov/ outlooks/aeo/nems/documentation/transportation/pdf/m070(2016).pdf. EIA. 2017. Annual Energy Outlook 2017. Table 7. https://www.eia.gov/outlooks/archive/ aeo17. Fagnant, D. J., and K. M. Kockelman. 2014. The Travel and Environmental Implications of Shared Autonomous Vehicles, Using Agent-Based Model Scenarios. Transportation Research Part C, Vol. 40, pp. 1â13. FHWA. 2015. Highway Statistics Series. U.S. Department of Transportation, Washington, D.C. https://www.fhwa.dot.gov/policyinformation/statistics/2015. FHWA. 2016a. FHWA Forecasts of Vehicle Miles Traveled (VMT): Spring 2016. https://www. fhwa.dot.gov/policyinformation/tables/vmt/may_2016_vmt_forecast_sum.pdf. FHWA. 2016b. Highway Statistic Series. U.S. Department of Transportation, Washington, D.C. https://www.fhwa.dot.gov/policyinformation/statistics.cfm. FHWA. 2017. Traffic Volume Trends. U.S. Department of Transportation, Washington, D.C. https://www.fhwa.dot.gov/policyinformation/travel_monitoring/tvt.cfm. FHWA and FTA. 2016. 2015 Status of the Nationâs Highways, Bridges, and Transit: Condi- tions and Performance. Report to Congress. U.S. Department of Transportation, Wash- ington, D.C. https://www.fhwa.dot.gov/policy/2015cpr/pdfs/2015cpr.pdf. Gucwa, M. 2014. Mobility and Energy Impacts of Automated Cars. Presented at 2014 Auto- mated Vehicle Symposium, July 15â17, San Francisco, Calif. https://higherlogicdownload. s3.amazonaws.com/AUVSI/c2a3ac12-b178-4f9c-a654-78576a33e081/UploadedImages/ documents/pdfs/7-16-14%20AVS%20presentations/Michael%20Gucwa.pdf. Handy, S., and M. G. Boarnet. 2014. Impact of Highway Capacity and Induced Travel on Passenger Vehicle Use and Greenhouse Gas Emissions. Policy Brief for Air Resources Board of California Environmental Protection Agency. https://www.arb.ca.gov/cc/sb375/ policies/hwycapacity/highway_capacity_brief.pdf. Harper, C., S. Mangones, C. T. Hendrickson, and C. Samaras. 2016. Estimating Potential Increases in Travel with Autonomous Vehicles for the Non-Driving, Elderly, and Travel- Restrictive Medical Conditions. Transportation Research Part C, Vol. 72, pp. 1â9. International Transport Forum. 2015. Urban Mobility System Upgrade: How Shared Self- Driving Cars Could Change City Traffic. Organisation for Economic Co-operation and Development. https://www.itf-oecd.org/sites/default/files/docs/15cpb_self-drivingcars.pdf. KPMG. 2015. The Clockspeed Dilemma: What Does It Mean for Automotive Innovation? https://home.kpmg.com/xx/en/home/insights/2015/12/the-clockspeed-dilemma-gary- silberg-head-of-automotive-kpmg-us.html. Lee, D. B. 2002. Induced Demand and Elasticity. Volpe National Transportation Systems Center, U.S. Department of Transportation, Washington, D.C. Lee, D. B., and M. W. Burris. 2005. Appendix C: Demand Elasticities for Highway Travel. In Highway Economic Requirements SystemâState Version, Technical Report, FHWA, Washington, D.C. Litman, T. 2017. Understanding Transport Demands and Elasticities. Victoria Transport Policy Institute, Victoria, British Columbia. Noland, R. B., and C. S. Hanson. 2013. Induced Travel Demand: Research Design, Empirical Evidence, and Normative Policies. Journal of Planning Literature, Vol. 17, No. 1, pp. 3â20.
266 NATIONAL COMMITMENT TO THE INTERSTATE HIGHWAY SYSTEM Pickrell, D. 2001. Induced Demand: Its Definition, Measurement, and Significance. Presented at Eno Transportation Foundation Policy Forum, February 22â23, Washington, D.C. Pickrell, D., D. Pace, and J. Wishart. 2014. FHWA Travel Analysis Framework: Development of VMT Forecasting Models for Use by the Federal Highway Administration. Volpe Na- tional Transportation Systems Center, Cambridge, Mass., and FHWA, Washington, D.C. Pisarski, A. E., and A. T. Reno. 2015. Transportation Bottom Line. AASHTO Bottom Line Report, Executive Version. http://bottomline.transportation.org/Documents/Bottom%20 Line%202015%20Executuve%20Version%20FINAL.pdf. Polzin, S. 2006. The Case for Moderate Growth in Vehicle Miles of Travel: A Critical Juncture in U.S. Travel Behavior Trends. U.S. Department of Transportation, Washington, D.C. Polzin, S. E., X. Chu, and J. Godfrey. 2014. The Impact of Millennialsâ Travel Behavior on Future Personal Vehicle Travel. Energy Strategy Reviews, Vol. 5, pp. 59â65. http://dx.doi. org/10.1016/j.esr.2014.10.003. Trommer, S., V. Kolarova, E. Fraedrich, L. KrÃ¶ger, B. KickhÃ¶fer, T. Kuhnimhof, B. Lenz, and P. Phleps. 2016. Autonomous Driving: The Impact of Vehicle Automation on Mobility Behaviour. Institute for Mobility Research. https://www.ifmo.de/files/ publications_content/2016/ifmo_2016_Autonomous_Driving_2035_en.pdf. U.S. Census Bureau. 2016. American FactFinder. https://factfinder.census.gov/faces/nav/jsf/ pages/searchresults.xhtml?refresh=t. Wadud, Z., D. MacKenzie, and P. Leiby. 2016. Help or Hindrance? The Travel, Energy, and Carbon Impacts of Highly Automated Vehicles. Transportation Research Part A, Vol. 86, pp. 1â18.