This book is an open exploration in mobility trends with a focus on five flourishing and forward-looking cities in the United States. The choice of these five cities is based on indicators of positive growth and development reflected in a multi-dimensional index for socio-economic and travel related metrics. The aim of the book is rooted in data manipulation. It applies both top-down and bottom-up analytical approaches to test emerging trends in mobility: it looks both at large-scale changes across census intervals as well as minute, hour-by-hour changes observed across volumes of location based data from big data sources. This multi-scalar and data-driven approach aims to offer a broad picture of the current mobility landscape in these five progressive cities while imagining potential direction trajectories for the country’s future. Given the COVID-19 moment in which this book was published, we understand the amplified importance of such interrogations at the brink of a turning point; an upheaval of the status quo. We realize that the long-run effects that this moment of disruption may produce in the American mobility landscape are cloaked in uncertainty. Yet, it is precisely because of this uncertainty that the findings within the pages of this publication are so valuable, because they offer a snapshot of some of the defining directions and travel characteristics preceding the major disruption. Beyond results, this book’s contribution also lies in its methodology and approach following cities with progressive traits as trendsetters for the country’s future. Ultimately, the book offers an evidence-driven narrative for future mobility based on a collection of timely data.
Tracing mobility through a multi-focal data lens
In order to trace mobility trends in the United States, data has been compiled, sorted and employed from a number of robust data sources as outlined on the opposite page. Covering areas ranging from sociodemographic and population statistics to origindestination trip data, the collected data is variously employed throughout the body of this book to test existing arguments and convey new ones based on sound empirical evidence. Chapters 3 and 4, in which the bulk of findings are discussed, follow contrasting data approaches. Chapter 3 starts from a number of predetermined problems and tests them through data, whereas Chapter 4 starts from the data to follow a bottomup approach to identify trends and define problems. The analysis presented herein and in the chapters to follow consists of a multi-scalar approach starting from the national scale, whereby a number of mobility trends are interrogated at the widest lens, and ending with the city scale, whereby a predefined list of selected cities are closely examined to extract mobility information at a higher level of detail. This layered analysis requires a critical coordination of diverse data sets to ensure compatibility and comparability across the spectrum.
Multi-dimensional city growth index
The following scale places cities of the U.S. on a compound growth metric based on their performance in a number of growth areas as defined in the key below. This combined index acts as a starting point for defining a selection of well-performing cities for deeper analysis.
Diverse city profiles on the growth spectrum
The radar charts presented on the opposite page visualize some fast-growing and slow-growing city profiles in the U.S with respect to a number of preset indicators. The five selected fast-growing cities are chosen as the focus cities for detailed analysis throughout this book. The charts demonstrate the diversity of the selected cities with respect to overall growth volumes and variable growth patterns. Growth volumes of each city are visualized as plane surface areas, measured as the share of the maximum plane (100% growth in all directions).
Radar chart index
City growth volumes as percentage of maximum
Radar chart index relative to city growth indicators
Tracking emerging mobility trends –
Are there new mobility directions?
Strength in numbers: where are millennial populations growing?
Data clearly shows population surges in America’s Rocky Mountains and Southwest regions
In this study, the geography of the United States is segmented into 55 megaregions, redefining regions as areas with commuting and economic relations. The megaregional map is based on a superimposition of visual and logarithmic computations of megaregions; ensuring it is grounded in empirical analysis but refined using interpretive cartographic methods. A general reading of the millennial population (defined at the time of the study as those falling between the ages of 25-44) in the United States in absolute numbers based on 2016 census data highlights higher concentrations in the coastal regions. This trend is in keeping with general population figures. According to the United States Census Bureau, about 30 per cent of the population lived in counties directly on shorelines by 2017; a 15 per cent increase since 2000. Popular studies and readings on the habits of millennials tend to suggest a clear demarcation from previous generations, and particularly in the United States. Among other things, it is suggested that they are more transit-oriented, less car-oriented and more open to alternative forms of mobility than their predecessors. Looking at the shares of millennials relative to total populations across the country, we find that this share is generally above 20% in most regions today. Comparing this data with the situation just 4 years prior, we find that millennial shares rose in most regions across the United States, with profound increases particularly in parts of central and western America, where increases are recorded at 10% and above. The highest increases in millennial populations are all concentrated in the state of Texas, where three of the five most attractive cities for millennials are located (as previously shown in Chapter Two). What do these demographic changes mean for American mobility trends? The following section studies a number of mobility trends in cities of diverse demographic and economic patterns as a way of investigating emerging mobility cultures. By interpreting data from the most recent census figures, we explore whether these surges in millennial populations are accompanied by any mobility shifts in America’s urban areas.
Millennials are in fact increasing in central areas of all five cities under study and decreasing from some surrounding districts. Save Austin, the same trend cannot be traced for the Over-65 population, which in the short four-year span has tended to increase in suburban areas. In Savannah, a cluster of increases is recorded in the area just northeast of the downtown area.
Sharing mobility: a differential upsurge
Diverse growth trends as new mobility options set their course
From carsharing to bikesharing and scootersharing, much has progressed in the U.S. mobility landscape in the past decade or so. New shared mobility services have been exponentially growing in the past years, with various trends across the country. As the opposite maps show, bikeshare systems, which have been growing since 2010, today range from less than 1000 bikes in some cities to 10,000 bikes in others. Scootersharing, in retrospect, has only been around since 2017. Yet, a number of cities have rapidly surpassed the 10,000 scooter mark. Likewise, the carsharing market has seen a continuous upward trend in the decade 2006-2016, growing by 15 times the number of registered members across North America in 10 years (as shown below). This diverse uptake of shared mobility services is evident in the cities under study. Taking bikeshare as an example, we find that Denver’s B Cycle system is ahead of the race, dating back to 2010. Los Angeles’ Metro bikeshare system on the other hand has only been around since 2015. Even newer, Seattle’s bikeshare system, which as of 2019 is composed of two private providers (JUMP and Lime) and is only just beginning to find ground after failed pilot attempts in 2015/2016.
The selected cities vary in their offerings of shared mobility services. Bikeshare trends are one example: while Denver’s B Cycle system has been around since 2010, Seattle’s bikeshare program only began in 2016 and companies are still changing year on year.
Micromobility services available in selected cities by type and provider. Data source: NUMO (New Urban Mobility Alliance)
Telecommuting on the rise
Prior to the large-scale adoption of home working across the U.S. during the 2020 pandemic situation, a clear rise in telework popularity was already in motion. Between 2001 and 2016, the share of full-time teleworkers more than doubled in the U.S., suggesting that it was already gaining popularity before there was a direct need for it.
Share of full-time teleworkers in the US, 2001 to 2016. Data source: American Community Survey
The big takeaways
Lessons learnt and the main trends observed
The highest increases in millennial populations across the U.S are concentrated in the State of Texas. Three of the five most attractive cities to millennials are located in Texas, including Austin.
Younger generations are not driving far less than others, but they are more multi-modal, in the sense that they walk, cycle and use public transport at far greater levels than older age groups. Across cities, younger groups are at least twice as likely to walk or cycle than older generations and are progressively less likely to use public transport as they get older.
Los Angeles is an outlier in the sense that people are more likely to drive (and own cars), including the younger generations, across the spectrum. The potential to break the driving cycle is generally more difficult in this large, sprawling metropolis.
A ‘Back to Downtown’ movement is prevalent among millennials in all of the focus cities, while the Baby Boomers are mostly pouring out to the suburbs save Austin, where a notable increase is recorded in central areas near millennial clusters.
There is an overall upward trend in shared mobility development but with large variations between cities, highlighting the newness of these systems, which are still largely in their growth phase, pre-consolidation.
Growing even before the pandemic, work from home trends doubled in scale in the first quarter of the 21st century. Prior to the pandemic, the tendency to work from home was higher for older age groups.
Denver, Colorado comes out strong in its share of remote workers among the group, with the highest rate of increase as well. The State of Colorado has the highest share of full-time telecommuters in the country, which makes sense given its concentration of scientific research and high-technology industries.
Seattle is the only city among the group with a declining car ownership. It also shows a high rate of decrease in the overall share of private transport modes, coupled with high public transport ridership and active travel uptake.
Cross-city analysis: trip data under the microscope
Setting the scene
This chapter takes a closer look at movement patterns in the selected cities to see how they differ from one another with respect to a set of trip characteristics. For a deeper understanding of city patterns, this chapter is oriented towards a fragmental analytical approach: it is structured such that in each city, differences in patterns are detected at the scale of classified zones in order to identify similarities/dissimilarities transversally between cities. For a cross-study benefit the same classified zones were identified for each city. These zones are: downtown, university area, high-income residential and suburban residential zones. In continuity with previous chapters, these zone typologies were chosen to reflect contrasting profiles for socio-economic traits that have been identified as criteria for emerging mobility trends. Such traits include millennial population, millennial population growth, active modal shares, etc.
Our initial assumption that the four zone typologies focused on have different movement patterns was supported by the data. At a glance, distinctions of movement patterns could be broadly drawn between two groups of ‘twin’ zones: the two residential types (high-income residential and residential suburban) as one group and downtown and university zones as the other. To that end, it became clear through the analysis that a zone’s functional nature is a strong factor determining the type of trip characteristics it exhibits. This was the case for each of the studied trip traits. In some instances, there are differentiated trends even between twin zones. Movement density is one such diversified trait: high-income residential areas have the lowest overall movement density of all areas and suburban residential areas have three times that density. A similar distinction can also be made between densities of the twin group on the higher end of the scale: on average, downtown generates 3 times the average movement of university zones. In any case, the performance of downtown areas is more closely linked to the morphological structure of the city, as discussed earlier in the role of mono- and poly-centricity. In other studies, distinction between downtown and university zone performance lends itself more to scheduling factors relating to the main function of each area. For example, internal movement in university zones are amplified more broadly over the morning and afternoon period, whereas in downtown areas the rise is more restricted to the short two/three-hour segment corresponding to office lunch breaks.
Another recurring element emphasized throughout the analysis is distinct lunchtime patterns. The data has shown that lunchtime trips tend to be shorter in distance, though not necessarily in time. Example given, non-home based1 movements are more likely to be internal than any other time of day2.These patterns generally tend to register higher in downtown and university areas, downtown areas in particular. Considering the fact that movement density multiplies by large factors in downtown areas relative to other zones, it becomes clear that a micro-mobility transport offering, which consumes less road space, is more energy-and-cost-efficient and is specifically designed for short, single-user trips, could go a long way towards improving the flow of downtown movement, especially those still heavily reliant on car use. In university zones, this potential capture timeframe is stretched over a longer period of time (from morning to the end of lunchtime).
1On average, more than half of all lunchtime trips and more than two thirds of downtown lunchtime trips are non-home-based. 2 Internal trips have an average share of a quarter of all trips during lunchtime.
On the weekend, movement is generally lower in all zones of any one city, though the amount of reduction is linked to the functional relationships between the different zones in that city. For example, it is no rule that downtown movements show greater reductions than residential areas. The extent to which movement is reduced on the weekend in downtown areas will depend on the functional distribution of the city and the level of functional interdependence of its various zones. The congruence of trip generating areas (residential) and trip attracting areas (commercial, business or schools) plays a role in these variations. In terms of hourly distribution, weekend movements exhibit the reverse trend of weekdays: on the weekend, movements reach their peak in the afternoon as opposed to morning and evening peaks of weekday movements. In addition to this, longer distance movements tend to be higher for residential areas in weekends as opposed to weekdays in all cities.
Population groups and time poverty
The amount of time spent traveling per day per person contributes to their ability to utilize time for other purposes including work, rest and leisure – a measure known as ‘time poverty’. From the data, it was deduced that certain demographic factors have higher correlations with long daily travel times, i.e. a high degree of time poverty. These populations include larger households, lower-income households, and people living in cities with a high share of private travel modes (mainly private cars). Time poverty for these groups is caused by a combination of long-distance travels of suburban communities and traffic congestion. On the other end, younger travelers, and populations living in dense urban areas or in areas with high active travel modes (walking and cycling) are more likely to complete their commutes in shorter times. Naturally, there are exceptions to these rules. The data showed that Los Angeles travelers are more likely to be time-poor across the spectrum. This in part owes to the city’s stratified clustering of jobs and services, as well as its sheer size compared to the other cities under study.
Towards a new mobility landscape
The mobility ecosystem is gradually expanding into a more diversified portfolio: the personal and collective mobility spheres are exploring shared, electric, automated, on-demand, micro- and integrated solutions to fit conflicting needs in a fast urbanizing world. Trends discussed within the book point to the growing adoption of American cities of new alternative mobility options.
The progressive cities framework
Within this study, the progressive (versus non- progressive) cities framework was used as a way to read mobility trends. In progressive cities, which were the focus of this book, a number of co-existing socio-demographic, economic and behavioral conditions were used as parameters feeding into a ‘progressiveness index’. These parameters were designed with the aim to highlight cities showing healthy economic growth trends and a significant representative share of progressive behavioral outliers (millennials, in this case). A select number of cities, which performed well on these scales were tested for a number of progressive mobility trends, as signals of a potential model mobility landscape to replicate across the country.
When things took a sharp turn: a final reflection
Throughout the book, the goal was to use a diversified data approach to observe growing mobility trends, understand where they prevail (topographic tendencies), when they prevail (chronological tendencies) and who their main proponents are (demographic structure). While the trends may be forever altered after the current moment of disruption has passed, the methodology remains intact. It will take continued observation of location-based data over the next several years to patternize the country’s new mobility landscape.