Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 15, Land Use and Site Design (2003)

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.

15-123 Evaluating Criticism of Smart Growth, available at the Victoria Transport Policy Institute website (Litman, 2003). CASE STUDIES Arlington County, Virginia, Transit Oriented Development Densities Situation. Arlington County is part of the Washington, DC, metropolitan region, situated in Northern Virginia just across the Potomac River from the Nation’s Capital and home to the Pentagon. Prior to construction of the Washington Metrorail system, Arlington’s location made it primarily a close-in bedroom suburb, offering convenient access and affordable housing for Federal government workers and military in downtown Washington or the Pentagon. Conscious planning decisions in anticipation of the construction of Metrorail into Northern Virginia, and predicated on a strong market for office construction, have accounted for significant changes in land use development patterns in Arlington. These changes have greatly shaped the economic and community activity levels of Arlington, and transit ridership levels for trips beginning in or destined to the County. Actions. The Washington Metrorail system began operations in 1976, and its first extension outside the City was to Arlington. The County made a conscious decision that it wanted to encourage growth, and to take maximum advantage of the opportunity presented by Metro. Rather than pushing one Metro alignment north into freeway right-of-way, it decided to bring it in subway through the heart of County areas where commercial development and multi-family housing were already established, but beginning to decline. The expressed intent was to locate the service where higher levels of activity already existed, and new development as well as redevelopment of existing resources was wanted. The County established as its primary development goals in conjunction with this decision: (1) achieving a 50/50 tax base mix of residential and commercial development, (2) preserving existing single family and garden apartment residential areas, (3) encouraging mixed use development, and (4) concentrating development around Metro stations. Sector plans focusing areas within about 1/4 mile of each station were developed and pursued with developers, using special exception site plans as the approval mechanism. Some 5 percent of Arlington was replanned. Analysis. A record of actions taken, the accompanying land use development and population and employment shifts, and aggregate impacts on transit use, is maintained by the Arlington County Planning Director and his staff to support furtherance of the program. Results. Since the 1970s and the coming of Metro, the County has experienced major growth and renewal, partly attributable to the growth of the Washington region in general, partly to the attraction of Arlington as an affordable location close-in to downtown Washington, and partly — it is believed — to aggressive efforts to plan and market TOD. Between 1969 and 2000, office space in Arlington has increased from 4.5 to 18.4 million square feet, and high density residential development has gone from 2,600 units to 14,300 units. Growth activity has occurred mainly in the vicinity of the County’s 11 Metro stations, but with the most spectacular growth in relation to the Rosslyn, Ballston, and Court House stations. In 1980, 51 percent of county jobs were located within walking distance of Metro. This was with only one station not already open, and that at a primarily residential location. By 2000, the fraction reached 67 percent, and it is expected to reach 69 percent by 2020. Transit ridership has

15-124 paralleled the growth in development at the three major stations. Between 1991 and 2000, daily Metrorail passenger entries and exits have grown from 13,600 to 30,100 at Rosslyn; from 5,600 to 14,700 at Court House; and from 9,500 to 21,900 at Ballston. More… Clearly, the extension of Metrorail into Arlington in the late 1970’s and early 1980’s has had a major impact on the physical appearance and economic vitality of the County, particularly in the Rosslyn-Ballston Corridor where Metro service was concentrated. Several factors are credited with the County’s success with TOD. First, they developed a County plan and detailed sector plans to communicate clearly to investors and residents what type of development was planned. This was believed to create a sense of integrity in plans and policies that could be relied upon. Helping this, the government has been fairly stable throughout the growth period, meaning that there have been no political shifts to threaten TOD plans or policies. Second, land adjacent to stations was rezoned to higher density as developers came forth with acceptable plans. Initially, Floor Area Ratios (FARs) of 1.5 were the norm throughout the county, but FARs up to 3.8 have been permitted under the TOD plan. Third, county officials have worked continuously at building community consensus and creating value, pushing for top quality development projects and not just settling for generic office buildings. Fourth, they have attempted to make maximum use of public- private partnerships. While visibly successful, the County is still struggling with several issues, including finding the right balance of parking, achieving desired levels of retail development sufficient to support a 24-hour environment, securing a desired balance of affordable housing, obtaining a more uniformly high quality of urban design, and engineering enough public space or green space into the mix to preserve a community feel. Source: Brosnan, R., “Transit Oriented Development,” The Smart Growth Speaker Series. Oral presentation and visuals (updated 2001). Sponsored by the U.S. EPA, ICMA, the National Building Museum, and the Smart Growth Network, Washington, DC (September 5, 2000). San Francisco East Bay Pedestrian Versus Auto Oriented Neighborhoods Situation. Rockridge and Lafayette, California, are two San Francisco East Bay area suburbs with strong similarities except for their urban form. Both are served and bisected by the SR 24 freeway, and by the BART heavy rail transit line between San Francisco and Concord, which runs in the freeway median with a station in the center of each community. Average incomes in 1990 were only 4 percent apart, the age structure is similar, and racial makeup is predominantly white in both cases, although the proportion of African-Americans in Rockridge is higher (16 percent versus less than 1 percent). Some differences notwithstanding, such as nearness to the University of California and more frequent bus service in Rockridge, and separation by a range of steep hills, the two suburbs provide an unusually good laboratory for paired-community analysis of differences in the built environment. Actions. The Rockridge neighborhood of Oakland was developed at the turn of the last century in the prototypical form of a transit-oriented streetcar suburb. At the core is a retail district with pedestrian-scale storefronts directly along the sidewalk, with parking mostly at the curb or in the back. Few retail area sidewalks are interrupted with curb cuts. The street network is grid-like, albeit with irregularities, and blocks are relatively small. The entire neighborhood is linked with an integrated network of sidewalks and pedestrian paths. The

15-125 planting strip between most sidewalks and streets is occupied by shade trees, and mid-block pedestrian paths enhance transit access in some locations. Lafayette, just two stops away on BART, lies behind the East Bay hills. Lafayette remained largely rural until a 1937 highway tunnel facilitated development. New growth, which in the 1950s began in force, dominates. Retail is mostly in stand-alone buildings, with off-street parking in front, along a 4-lane arterial with median. The land use mix transitions from BART station parking (on one side), to relatively large retail parcels, to offices and multi- family housing, to single family residences. The mix has a coarser grain than in Rockridge. The street system features local, collector, and arterial street hierarchy, with little continuity except along arterials. The retail core has sidewalks, but they are sporadic elsewhere. Pedestrian connections are hampered by elongated blocks and circuitry. Comparative socio- demographic, built-environment, and transit service statistics are given in Table 15-59. Table 15-59 Comparison of Rockridge and Lafayette Neighborhoods Socio-demographic Characteristics Rock- ridge Lafay- ette Neighborhood Characteristics Rock- ridge Lafay- ette Median Household Income $58,770 $61,071 Housing units per sq. mi. 2,194 655 Persons per Household 2.2 2.5 Pct. single-family detached 63.6% 78.4% Percent Single Households 33.7% 18.6% Blocks/sq. mi. near BART 103 47 Median Housing Value $322,595 $392,853 Intersections per sq. mi. 127 64 Median Monthly Rent $682 $843 Cul-de-sacs 5 31 Median Age 37.3 39.8 Number of bus routes 3 3 Percent White 73.8% 88.2% Peak bus headways, min. 2.8 9.7 Pct. Adults College Educated 44.5% 40.7% Peak BART headways, min. 3 3 Source: All tables in this case study from Cervero and Radisch (1995). Analysis. A non-work travel survey and a commute trip travel survey were sent to a sample of all types of households in the two communities in 1994. Each survey involved the mail-out of 4,000 forms. Response rates of 15.5 and 21 percent were achieved for the two surveys, respectively. An additional abbreviated survey was also employed. Although the response rates were low, comparison of household data obtained with 1990 Census results indicated that the returned surveys were fairly representative. In addition to various tabulations and comparisons of the results, binomial logit models were calibrated for home-based non-work and home-based work trip mode shares. The non-work dependent variable was non-auto share, and the work purpose dependent variable was non-SOV share. The models, while simple, were designed to correct for any effects of persons-per-household differentials and the proximity of Rockridge to the University of California, Berkeley. Results. The travel surveys revealed important travel behavior differences between the two communities, as detailed in Table 15-60. The overall picture is one of substantially higher utilization of NMT in Rockridge, and significantly higher auto use in Lafayette, aside from almost identical levels of BART rail transit utilization. The mode share findings are interpreted and qualified in the “Response by Type of Strategy” section of this chapter, under “Site Design” — “Site Design and Travel Behavior.” They suggest that the more important impacts of TND built environments may be on local travel access and non-work travel, with

15-126 residents of Rockridge found three times more likely overall to walk to shopping, recreation, and other local activities than their Lafayette counterparts. In addition to the differentials in primary mode share shown in Table 15-60, there were also large differences in the mode of access to BART. While 31 percent of Rockridge residents reached BART by walking, only 13 percent did so in Lafayette. In both locations, 94 percent of walk access trips were under 1 mile in length. Rockridge also had a bus access to BART share higher by 7 percent, and a 56 percent park-and-ride or kiss-and-ride share, versus 81 percent for Lafayette. Table 15-60 Rockridge Versus Lafayette Comparative Travel Characteristics Commute Travel Non-Work Travel Mode Choice Rockridge Lafayette Rockridge Lafayette Auto Drive Alone Carpool Casual Carpool 63% 51 2 10 79% 69 4 6 85% 96% Transit BART Bus 26% 21 5 20% 20 0 5% 2% Walk 7% 1% 7% Bike 4% 0% 3% Total NMT 2% Average Trip Length 6.8 miles 11.2 miles Included in Table 15-60 are the average trip distances for non-work travel, 6.8 miles for Rockridge residents and 11.2 miles for Lafayette residents. The authors ascribe this trip distance differential largely to Rockridge’s more compact structure. Lafayette trip distance averages are probably also pushed up by the less dense development of its immediate neighbors, the presence of several regional-scale parks and recreation lands in the area, and its somewhat further-out location. Overall, the average daily VMT per resident was 10.8 miles in Rockridge and 19.6 miles in Lafayette. More… The coefficients obtained for the binomial logit mode choice models calibrated are given in Table 15-61. In the work purpose model, the Berkeley destination variable serves to help correct for one possible source of bias, and in the non-work model, the persons per household variable does the same. The large San Francisco destination coefficient in the work purpose model reflects the importance of a variable that acts as a surrogate for congested auto travel, high parking costs, and high quality transit service.