Neither on-the-ground observation nor a basic data-driven analysis of high-transit use neighborhoods conclusively identify the factors driving transit use in Phoenix. Sprawl is so dominant as a residential pattern, however, that planners must focus not only on building transit-friendly neighborhoods but also on providing alternative transportation options in the decidedly transit-“unfriendly” subdivisions that make up Arizona’s largest city.
Transit advocates everywhere like to point to the New Yorks, Bostons, and Chicagos of our country when looking towards the future – if only we could recreate the dense urban fabric of Brooklyn out here in Texas, the thought goes. But unless we tear down nearly everything we’ve built in the last fifty years and start over, we’re going to have to figure out not only how to build transit oriented development but also how to build transit networks to serve the decidedly non-transit-oriented development that, for better or worse, dominates most urban areas outside of the east coast. That’s why I set out to visit Phoenix, Arizona – not to look at what could be, but to get an on the ground look at what’s actually driving transit ridership in one of America’s most car-centric and fastest growing cities. Who are the roughly 72 million1 riders who forego their cars for Valley Metro’s trains and buses each year? What can the current ridership patterns tell us about the secrets to transit success in quasi-suburbia?
Phoenix has been widely praised for its relatively new Valley Metro light rail line, which opened in 2008,2 and with good reason. The system’s sole 20 mile route features modern trains that whisk travelers across the vast expanses of America’s quintessential Sunbelt city at average speeds that, thanks in part to traffic signal priority, dwarf those experienced by riders on many more established light rail systems such as Boston’s Green Line.
There are certainly many valuable lessons to be learned from Phoenix’s light rail successes, but I set out to Arizona not only to investigate the flashy new line but also to take a closer look at what remains the Valley Metro system’s workhouse and backbone – the good old city bus. After all, if we’re going to build transit to effectively serve sprawling cities like Phoenix, where crosstown trips are the norm rather than the exception, we’re going to have to look beyond the classic downtown-focused light rail system. And unless cities such as Phoenix suddenly shift huge sums of money into their transit budgets, the type of crosstown, point-to-point transit network that will serve the spread-out destinations of Sunbelt city residents will likely involve buses or, ideally, bus rapid transit.
In fact, Valley Metro already operates an extensive, grid-like bus system that covers the vast majority of the city and enables direct, neighborhood-to-neighborhood journeys, albeit at lower than optimal frequencies.
Instead of relying simply on census data to analyze transit ridership trends, I picked three census tracts with lower than average proportions of commuters who drive alone to work – in other words, neighborhoods where walking, biking, carpooling, and riding transit to work are especially prevalent – and walked the streets, trying to get a holistic sense of what might set these neighborhoods apart from their more driving-inclined peers.
I traveled to three neighborhoods – census tract 1061 in the north Phoenix neighborhood of Alhambra, where roughly 30% of commuters use alternative modes, Encanto’s census tract 1105.02, where 38% forgo driving alone, and finally tract 1109.02, where 31% do so, compared to a citywide average of 25%.3
My journey began in Alhambra, a ten-minute bus ride north of the light rail line’s northwestern terminus. 19th Avenue, a larger arterial road, bisects the neighborhood, and condo complexes line both sides of 19th. In terms of transit, the neighborhood is served by the 19C bus, which connects to the northern terminus of the light rail line, as well as the 70 and 80 buses, which run east-west along Glendale and Northern Ave, respectively. While the apartment complexes on the eastern side of the street give off a very closed look, turning their backs to the street, those on the western side, interestingly enough, faced the street in a way that looked more urban than not. Venturing eastward and westward along the cross street Myrtle Avenue, I found squat ranch single-family houses, standing one after another on small lots. This type of development, in fact, dominated the largest share of every neighborhood I visited. Strip malls and gas stations alternate with condo complexes along 19th, but you could hardly call the development here mixed-use, and not surprisingly, pedestrians were few and far between.
Hoping to find more pedestrians and less suburbia, I moved on to the close-in neighborhood of Encanto, which is lucky enough to have the light rail line right down Central Ave. Central is a secondary central business district of sorts, with mid-rise office towers and, above all, parking lots lining its sidewalks. A few restaurants dot the scene, but given how deserted the street looked on the Sunday I visited, it’s clear that this is a 9 to 5 kind of place. Farther east, the neighborhood transitions abruptly into single family residential streets that could hardly be distinguished from those further out in Alhambra.
Finally, I rode the 41 bus out to Camelback East, northeast of downtown and directly east of Camelback. This corner of Camelback is served once again by two east-west bus routes, the 41 and 29, as well as the 44, which as its name suggests, runs north-south along 44th St. Here, I once again found the all-to-familiar pattern of strip malls lining arterial roads, masking street after street of single-family ranch houses. The residential parts of Camelback seemed even more suburban and better kept than those I visited earlier, interestingly enough.
Why does Encanto have the highest alternative mode use (38%) of the three? Perhaps the proximity to both light rail and downtown encourage transit commuting; it even seems plausible that residents of the single family homes east and west of Central Ave. might walk to jobs in the nearby office towers. In reality, however, increased public transportation use accounts for almost all of the difference. Explaining why the other two sites I visited have higher-than-average alternative mode use isn’t as clear-cut. They’re both chock full of the “strip malls plus single family ranch”-style development that dominates the city of Phoenix outside of downtown.
To delve further into these questions, I pulled a variety of 2013 American Community Survey data. Starting with the basics, these three neighborhoods – and by neighborhood, I mean the census tracts mentioned above – are in fact much denser than the city as a whole. While Phoenix has a density of 2,852 residents per square mile, Alhambra and Encanto clock in at 5,643 and 4,956, respectively, and Camelback East in fact has close to 10,000 per square mile. Density means that more residents have access to transit, and so these differences could well explain the greater transit use seen in the three areas.
In terms of race, Alhambra is actually less diverse than the city as a whole, as 85% of its population is white, compared to 77% of Phoenix. The share of white Encanto and Camelback residents is substantially lower than the citywide proportion, however. Only about a fifth of Alhambra and Encanto residents are hispanic,4 but 45% of Camelback and a similar share of Phoenix residents report the same. Ostensibly, only Camelback is substantially more diverse than the city as a whole as measured by both race and ethnicity, so it’s not obvious that their racial makeup contributes to their lower car commuting rates.
Income isn’t an obvious driver either – while Alhambra has a median income slightly above the citywide household median of $47,139, Encanto and Camelback report median incomes of approximately $35,000 and $31,000, respectively. Home value and median gross rent mirror these trends
Immigrants are often likelier to use transit or at least avoid driving, but the share of Alhambra and Encanto residents who were born abroad is actually lower than the citywide share of 20%.
Finally, the average commute times in each neighborhood hover quite close to the city average of 24 minutes.
So what gives? The residents of these three neighborhoods tend less towards single-person-car commuting than other Phoenicians, but they’re not consistently more diverse, wealthier, less well off, or more international than their peers in other parts of the city. The higher densities in these neighborhoods likely play a role, but on the ground, the built environment seemed to differ little from the rest of the city. Clearly, these questions require a more thorough, ideally an econometric, analysis, using more variables and sources of data.
At the same time, my on-the-ground observations do suggest that the typical built environment, mixed-use development-related drivers of transit use and pedestrian activity may well play less of a role in sprawling cities such as Phoenix. That’s not at all to say that a more “urban” built environment wouldn’t help encourage commuters to get out of their cars – on the contrary, transit oriented development has potential in any city, sprawling or not. However, the neighborhoods that I visited had little, if any such dense, mixed-use development. To boot, their infrastructure and built environment seemed no more pedestrian-friendly than that of other neighborhoods where more commuters drive alone, so there must be other, less obvious factors at play. Perhaps Alhambra, Encanto, and Camelback East have particularly convenient transit connections to major employment centers, or the condo complexes in these neighborhoods are in fact more compact than those elsewhere, or weekday traffic congestion plagues these areas more so than others.
Whatever these factors are, we’d do well to isolate them – transit oriented development is clearly the “first best” strategy for increasing transit use, but rebuilding substantial parts of cities such as Phoenix and Houston to make them more amenable to transit use is a massive undertaking, one that will require major policy changes as well as sustained private and public investment. In the meantime, commuting patterns in cities such as Phoenix could provide clues as to how we could make small changes, “baby steps,” that would encourage transit, bike, and pedestrian commuting
Table: American Community Survey 2013 5 year estimates, Phoenix vs. the 3 census tracts studied
|City of Phoenix||Alhambra (Census tract 1061)||Encanto (Census tract 1105.02)||Camelback East (Census tract 1109.02)|
|Population and density|
|Population Density (per sq. mile)||2,852.0||5,643.8||4,956.1||9,848.7|
|Black or African American Alone||6.8%||6.1%||10.5%||5.0%|
|American Indian and Alaska Native Alone||1.9%||0.5%||12.5%||3.5%|
|Native Hawaiian and Other Pacific Islander Alone||0.2%||0.0%||0.5%||0.6%|
|Some Other Race Alone||7.8%||1.4%||3.2%||15.3%|
|Two or More races||2.9%||4.6%||9.1%||4.4%|
|Not Hispanic or Latino:||59.7%||77.8%||78.2%||55.0%|
|Hispanic or Latino:||40.3%||22.2%||21.8%||45.0%|
|Median household income (In 2013 Inflation Adjusted Dollars)||$47,139||$53,407||$34,804||$30,839|
|Median house value||$158,000||$282,400||$191,500||$167,000|
|Median gross rent||$870||$952||$863||$818|
|Means of transportation to work|
|Car, truck, or van||87.3%||81.1%||74.6%||80.4%|
|Public transportation (Includes Taxicab)||3.5%||6.1%||11.0%||2.8%|
|Worked at home||5.0%||5.6%||6.2%||3.5%|
|Average Commute to Work (in min)||24||23||20||19|
Thanks to the Tufts Undergraduate Research Fund for providing generous support for this research. All photos by author unless otherwise noted.
My name is Lucas Conwell, and I’m a senior at Tufts University majoring in Quantitative Economics and minoring in Urban Studies. My focus is on transportation economics and planning, and I’ve had the opportunity pursue my passion for sustainable transportation through internships at German Railways (DB), the Massachusetts Smart Growth Alliance, and most recently, the Volpe National Transportation Systems Center in Boston.
3 All census data cited in this post is from the American Community Survey 2013 5-Year Estimates.
4 The Census measures ethnicity – whether a person identifies as hispanic – separately from race.