Beyond Neighborhood Design: Exploring the Effects of Smart Growth on Older Adults’ Travel Behavior over Time

Abstract

Spanish

Chinese

Previous studies mainly substantiate the influence of neighborhood design on older adults’ travel behavior. This study goes beyond and examines how smart growth affects older adults’ travel behavior over time in the central Puget Sound. Using regression models for the years 1999, 2006, and 2014, we find that smart growth factors have significant but changing effects. The regional growth centers play a growing role in reducing older adults’ travel distance, trip frequency, and promoting non-car commuting modes. This study adds to the knowledge of how older adults’ travel behavior can be affected by the multilevel and long-term urban development strategies.

Keywords

built environment central Puget Sound older adults smart growth travel behavior

Introduction

Demographic aging is a key societal phenomenon worldwide. The share of the U.S. population aged sixty-five and above is expected to rise from 13 percent (40.2 million) in 2010 to 20.2 percent (88.5 million) in 2050 (Vincent and Velkoff 2010). This demographic change has been raising social, economic, and ecological challenges, and its effects on urban travel and transportation have also generated critical concerns. Regarding the planning literature, one of the most heavily examined topics is the neighborhood design characteristics related to older adults’ travel and activity. These studies presuppose that older adults’ daily mobility is largely constrained within a neighborhood, and with most older adults hoping to age in place, the importance of the neighborhood environment in supporting their independent life is further stressed (Kan, Forsyth, and Molinsky 2020).

The studies performed in recent years have added to our knowledge about older adults’ livelihoods and travel patterns, indicating that a higher age does not necessarily mean restricted mobility (Haustein 2012). Cohorts of the new aged are likely to be healthier, wealthier, more educated, and independent. They experience greater mobility through higher trip numbers and longer travel distances, with increased rates of licensing and automobile use than previous generations (Cui, Loo, and Lin 2017; Siren and Haustein 2013). The increased mobility of older adults has also generated broader concerns about environmental impacts and sustainability issues (Banister 2011; Rosenbloom 2001). Older adults’ overall mobility may have been affected by the long-term implementation of urban development strategies, which none of previous studies have discussed. Only investigating the effects of neighborhood designs, the full portrait of older adults’ travel characteristics and their relationship with the larger urban environment may be masked.

To fill this gap, this study examines how smart growth strategies affect older adults’ travel distance, trip frequency, and mode choices over time. We use the central Puget Sound as the case study area. This metropolitan region has implemented smart growth strategies since the 1990s and has a rapidly growing aging population, as well. Based on the Puget Sound Household Travel Survey, we applied linear regression models, Poisson regression models, binary logistic regression models, and the chi-square test of independence for the years 1999, 2006, and 2014. We found that the smart growth factors have significant but changing effects on older adults’ travel behavior over the years. Compact development through the UGBs and higher housing density reduces older adults’ utilitarian travel distances and facilitates their recreational trips. The regional growth centers (RGCs) play a growing role in reducing older adults’ travel distance, trip frequency, and promoting non-car commuting modes.

This paper is organized as follows. The next section reviews the literature on the smart growth, neighborhood design, and older adults’ travel behavior, with research gaps identified and research hypotheses put forward. In the third section, we review the implementation of smart growth strategies and its effects in the central Puget Sound. The fourth section introduces the data, variables, and models used in this study. Model results are presented in the fifth section. In the sixth section, we validate our hypotheses by discussing our model results. Finally, we discuss the associated implications for planning practice and future research directions.

Literature Review

Older Adults’ Travel Behavior: Health and Sustainability

Transport mobility is a key aspect of healthy and active aging that enables participation and autonomy into later life. Ample research affirms that higher mobility in old age is associated with numerous physical health benefits and subjective well-being. In the United States, the automobile provides access to widely distributed services and facilities, and it remains the backbone of an older adult’s active and independent lifestyle (Boschmann and Brady 2013). According to research conducted decades ago, driving one’s own car supports an older adult’s sense of personal control, thus leading to greater well-being (Gonda 1982). By contrast, public transit does not engender a feeling of personal control because older adults often find that public transit has difficulty in accommodating their needs (Carp 1988). Using 2009 Disability and Use of Time (DUST) data, Ravulaparthy, Yoon, and Goulias (2013) find that older adults who walk or ride a bicycle to engage in activities are more likely to be dissatisfied than their counterparts who use an automobile, partly due to a less favorable walking and biking environment and potential physical barriers. Thus, keeping older adults driving as long as safely possible has been claimed to be the optimal mobility option for an aging society (Rosenbloom 2009).

However, the provision and improvements of alternative transportation services, such as public transit, would undoubtedly serve the needs of some older adults, especially those who are not able or not safe to drive. Older drivers have a high rate of fatal car crashes per mile (National Highway Traffic Safety Administration [NHTSA] 2020), which generates growing concern about the risks they cause. Given that public transit has less than a tenth the per-mile traffic casualty rate as automobile travel (American Public Transportation Association [APTA] 2016), transit-supportive policies can provide substantial traffic safety benefits, which include reducing elderly involved car crashes. Besides, active travel, especially walking, has been proved in numerous studies to restrain the risk of physical and cognitive disease for older adults, such as cardiovascular disease (Smith et al. 2007) and dementia (Abbott et al. 2004). Researchers also note that transit use facilitates older adults’ physical activity by encouraging them to walk to access the transit stations from origin and destination (Freeland et al. 2013; Julien et al. 2015).

Trip purpose, which relates to activity participation, is also fundamental for older adults’ mobility-related well-being. Siren, Hjorthol, and Levin (2015) note that previous policies supporting older adults’ daily mobility primarily targeted the utilitarian necessities. They then suggest considering discretionary mobility, such as trips to participate in community activities. Ravulaparthy, Yoon, and Goulias (2013) find that older adults who take more trips for socializing and physical activities (sport, working out, walking, etc.) are more likely to have higher-level life satisfaction.

In addition to health outcomes, a few studies explore older adults’ travel behavior from the perspective of sustainability. Rosenbloom (2001) points out that older drivers may proportionately cause more environmental pollution because of more wasted miles traveled due to wayfinding errors and trip-scouting behavior. Based on a sustainable mobility paradigm, Boschmann and Brady (2013) explore the effects of transit-oriented development (TOD) on older adults’ travel behavior in Denver and find that TOD does lead to more sustainable travel behaviors. This paper examines the effects of smart growth urban development strategies on older adults’ travel behavior from both health and sustainability perspectives, which seek to meet older adults’ travel demand and promote transit use and walking.

Smart Growth and Older Adults’ Travel Behavior—Beyond Neighborhood Design

Smart growth is a comprehensive concept with wide use that can refer to different spatial scales. On the regional or metropolitan scale, it involves the use of planning and zoning tools, such as urban growth boundaries (UGBs) and green belts, to curb low-density sprawl, manage development, and improve the efficiency of public infrastructure investment. On the local or neighborhood scale, it involves planning and designing strategies, such as mixed land use and compact building, to create a livable place that is pedestrian-friendly, has homes near services and amenities, provides various transportation choices, supports different housing types, and encourages social diversity (U.S. Environmental Protection Agency 2011). Inspired by smart growth policies, voluminous research has examined the relationship between the urban built environment and travel behavior since the 1990s.

Previous studies that link urban built environment with older adults’ travel behavior focuses on the effects of neighborhood design on older adults’ travel behavior, especially walking and physical activity. After controlling for sociodemographic factors, the neighborhood environmental attributes that show significant and positive influence typically include walkability (Chudyk et al. 2015), residential density (Nyunt et al. 2015), land use mix (Cerin et al. 2014), street connectivity (Nyunt et al. 2015), green and open spaces (Carrapatoso et al. 2018), aesthetic environment (Zandieh et al. 2016), traffic safety (Carrapatoso et al. 2018), and access to stores and services (Nathan et al. 2012). Kim (2003) investigates the factors influencing older adults’ total travel time and total travel distance. The author concludes that the urban form, in terms of population density and retail employment density, is not significantly associated with older adults’ mobility.

Some studies take self-selection, or attitude factors, into consideration. Residential preference, such as choosing to live in a neighborhood because of its proximity to stores and services, and travel preference, such as a pro-transit attitude, may affect the estimation of environmental influences on travel behavior. Cao, Mokhtarian, and Handy (2010) find that after controlling for attitude factors, the neighborhood environment has limited effects on older adults’ driving and transit use, but access to stores and services is still positively associated with older adults’ walking trip frequency. In another study about the neighborhood environment and older adults’ active travel in China, Cheng et al. (2019) find limited self-selection effects on older adults but significant self-selection effects on younger adults.

After controlling for socioeconomic and attitude factors, previous studies largely substantiate the association between neighborhood design and older adults’ travel behavior. Although overall, today’s older adults still travel shorter distances (Zhai et al. 2019) and may be more influenced by the neighborhood environment compared with their younger counterparts (Yang et al. 2018), they seem to be more mobile and have more extended life-space boundaries than previous generations (Lee and Tan 2019). In addition, labor force participation rates are growing fastest for older adults aged sixty-five and above in the U.S. (Schramm 2018), which indicates increasing commuting trips. Older adults’ overall mobility may have been affected by the long-term implementation of regional and local planning strategies. Because the objectives of these strategies can be realized and improved slowly, observing their effects on older adults’ travel behavior may take a decade or so (Porter, ten Siethoff, and Smith 2005). How the smart growth strategies affect older adults’ travel behavior and how the effects change over time has been rarely explored.

This study performs an empirical analysis with a focus on the central Puget Sound, which is known for its smart growth strategies such as the RGCs, UGBs, mixed land use, TOD, and compact development. We use datasets from 1999, 2006, and 2014 to explore how the effects of smart growth on older adults’ travel distance, trip frequency, and mode choices have changed over the years. Based on the above literature review, we put forward the research hypotheses as follows:

Hypothesis 1 (H1): Smart growth reduces older adults’ travel distance (H1a) and leads to higher trip frequency (H1b).

Hypothesis 2 (H2): Smart growth promotes older adults’ transit use (H2a) and walking (H2b).

Hypothesis 3 (H3): The effects of smart growth on older adults’ travel behavior grow over the years.

Smart Growth and Aging Populations in the Central Puget Sound

The central Puget Sound, which includes King, Pierce, Snohomish, and Kitsap counties (Figure 1), has been identified as an exemplar of a multilevel and comprehensive implementation of smart growth strategies, among other well-known exemplars, such as the Portland, Denver, and San Diego metropolitan areas (Dierwechter 2014; Margerum et al. 2013).

UGBs

The UGBs is one of the most important smart growth strategies. In 1990, the Washington Legislature passed the Growth Management Act (GMA), which mandated that most counties and cities in the state conduct comprehensive land use planning and set UGBs to concentrate urban development inside and conserve rural lands outside. The Puget Sound Regional Council (PSRC), which is the federally designated Metropolitan Planning Organization for the region, has been performing collaborative consistency review of local plans. The PSRC’s planning review process has also been used to monitor the extent to which the local plans are consistent with regional strategies. Early research concludes that the GMA has clearly affected how planning is done in the central Puget Sound (Weitz 1999). King County appears to have taken the strongest approach and has achieved notable results in terms of growth management (Porter, ten Siethoff, and Smith 2005). For Pierce county, Dierwechter and Carlson (2007) find a steady increase in the cluster of residential building permits inside the UGBs and a dramatic decrease of building outside the UGBs from 1991 to 2002. However, by monitoring changes in urban land cover, Hepinstall-Cymerman, Coe, and Hutyra (2013) conclude that the intended effect of UGBs may not have been accomplished in central Puget Sound by 2007.

RGCs

Central Puget Sound is also known for its RGCs, which are strategic places intended to receive a significant proportion of the future population and employment growth compared with the rest of the urban area. The region implemented VISION 2020, a long-range land use and transportation strategy consistent with the GMA objectives in 1990. VISION 2020 initially called for directing growth into twenty-one RGCs. After decades of updates, the region now has twenty-nine RGCs in total. RGC locations are characterized by compact, pedestrian-oriented development, with a mix of different office, commercial, civic, entertainment, and residential uses (Bai et al. 2020). RGCs are also connected by high-capacity transit systems, acting as regional hubs to contain the travel demand within the RGCs and encourage residents outside to use alternative modes when traveling to RGCs. Based on 2010 data, Dalal and Goulias (2014) examine relationships between industry diversity and residents’ travel within RGCs in central Puget Sound. They find that RGCs with diverse and balanced business establishments reduce trip numbers for residents and accommodate more localized travel, although most of the localized trips are still by personal vehicles.

TOD

Central Puget Sound has received acclaim for its TOD strategy, which was developed by a regional coalition called the Growing Transit Communities Partnership. The region has been investing heavily in its high-capacity transit system, greatly expanding the light rail, commuter rail, ferry, streetcar, and bus rapid transit. A majority of the existing and planned transit communities are within or near RGCs, and they serve an important role in accommodating new growth. Others are located outside RGCs or in rural areas and provide important connections to urban areas (PSRC 2020). The Partnership has identified seventy-four transit communities along the region’s long-range light rail corridors as study areas. These transit communities are reported to accommodate 21 percent of the regional total population growth from 2010 to 2016 and 30 percent of the regional total job growth from 2010 to 2015 (PSRC 2013).

Trends in Transit Ridership and Vehicle Miles Traveled (VMT)

The effects of smart growth can be reflected by the trends in transit ridership and VMT in the region. Jun (2008) compares the changes in auto share and transit share from 1990 to 2000 between seventeen selected metropolitan areas with similar populations and employment sizes. The results indicate that the Seattle metropolitan area, which is a main component of the Puget Sound region, has experienced the fastest increase in transit share (a 21.4% increase) and an obvious reduction in auto share (a 0.9% decrease) relative to the seventeen metropolitan areas. From 2005 to 2015, transit boarding numbers in the central Puget Sound grew at 3.6 percent annually. The next fastest growing metropolitan area is Tampa-St. Petersburg, at approximately 3.1 percent. In addition, the transit boarding numbers have increased faster than the population in the region. From 2010 to 2018, population grew 12 percent, while the transit boarding numbers grew 19 percent (PSRC 2013).

As for the VMT, the empirical data show that the regional VMT per capita stopped increasing and instead remained nearly constant over the 1990s (Porter, ten Siethoff, and Smith 2005). From 2010 to 2018, the total daily VMT in the region grew at a much slower rate than the population and employment. While the VMT grew by 6 percent, population grew by 12 percent and employment grew by 22 percent. While the total daily VMT increased, the daily VMT per capita decreased 5 percent from 2010 to 2018 (PSRC 2013).

Aging Populations and Their Travel Patterns

The share of older adults in the central Puget Sound is projected to increase steadily to 18 percent by the year 2030 when the last baby boomers turn sixty-five and is expected to remain at that level through 2050. This is a large increase compared with the past several decades, when the share of older adults was between 9 and 11 percent (PSRC 2018). The PSRC recognizes the rapidly growing aging populations as a great challenge to the regional planning strategy and emphasizes the importance of anticipating and preparing for adequate housing choices, a broad range of transportation options, and access to services and amenities (PSRC 2020).

According to the 2017 Puget Sound Travel Survey, older adults drive their personal vehicles for approximately 49 percent of their trips and take transit for approximately 4 percent of their trips. Although older adults take transit less, they walk more frequently than any other age group. Approximately 30 percent of older adults go for a walk five days or more per week (PSRC 2019). The next section provides an empirical examination of the effects of smart growth on older adults’ travel behavior in the region over the years.

Method

Data

The study sample comprises older adults aged sixty-five or older. We extract data regarding the older adults’ travel behavior, sociodemographic characteristics, and attitudes from the Puget Sound Household Travel Survey collected in 1999, 2006, and 2014. The Puget Sound Household Travel Survey is a regionally representative travel diary survey designed and conducted by the PSRC. The 1999 survey and 2006 survey consist of households keeping track of their travel and activities for a forty-eight-hour weekday period, and the 2014 survey collected travel information for a twenty-four-hour weekday period. The data collected include but are not limited to information about the household (size, income, number of vehicles, residency, etc.), persons in the household (age, gender, employment, education, driver license, workplace, commuting mode, etc.), and travel (travel time, travel distance, trips purpose, mode, etc.).

The land use data regarding smart growth are obtained from different sources. We acquire the geolocations of UGBs and RGCs from the PSRC official website. The land parcel data with property information are obtained from the official websites of the King, Pierce, Snohomish, and Kitsap counties. However, the class codes of land use are not exactly consistent among the different counties. To measure the land use mix at the census tract level, we merge the land parcel data from the four counties and reclassify them into seven uses (residential, industrial, transportation, commercial, public services, recreational, and undeveloped land). The housing unit data are obtained from Census Transportation Planning Products (CTPP), which are special tabulations of American Community Survey (ACS) data. We use the 2000 CTPP data and 2012–2016 CTPP five-year estimates data on housing units to calculate the housing density in 1999 and 2014, respectively. The data of the transit system geographic distribution are acquired from the Washington State Department of Transportation’s official website.

Variables and Measures

Dependent variables

This study investigates older adults’ travel behavior from several aspects, as follows: travel distance, trip frequency, transit use and walking, and commuting mode. For travel distance and trip frequency, we stratify utilitarian and recreational purposes. Researchers have recognized the necessity of distinguishing between travel for different purposes, but their characterizations are not consistent (Kang et al. 2017). In this study, utilitarian trips include trips for work-related activities, shopping and eating, medical appointments, and other personal business. Recreational trips refer to trips for social, physical, and leisure activities (Ravulaparthy, Yoon, and Goulias 2013), such as going to a movie, visiting a friend, participating in community activities, and exercising. For the analysis using the 1999 survey and 2006 survey, the data from the two survey days are pooled (see Kim 2003).

For walking and transit use, we set a dichotomous dependent variable, which is “whether they take transit/walk or not on the survey day.”

Independent Variables

The smart growth variables used in this study cover five domains of land use characteristics, as follows: (1) housing density, (2) transit density, (3) land use mix, (4) accessibility to an RGC, and (5) within or outside the UGBs. Because transit density and land use mix are available only in 2014, we just include the other three smart growth variables for the comparison between three years, which we would elaborate later. The housing density (housing units per square kilometer) and transit density (public transit stations per square kilometer) are calculated at the census tract level. The land use mix is expressed by the entropy index (EI_i) (Cervero and Kockelman 1997), which is calculated as follows: $E I_{i} = - \sum_{k = 1}^{K} p_{k} \times \frac{\ln (p_{k})}{\ln (K)},$

where K is the number of land use categories; p_k is the share of the area of land use category k in tract i relative to the total area of tract i. The EI_i values range from 0 to 1. A higher value represents a higher land use mix, whereas a close-to-0 value indicates a single land use.

The accessibility to an RGC is measured by the distance to the nearest RGC, which is computed from the centroid of the census tract to the centroid of the nearest RGC using ArcGIS. We source the designation information of RGCs from the PSRC official website and identify that there were twenty-one, twenty-five, and twenty-eight RGCs established in the region by the years 1999, 2006, and 2014, respectively. We compute the distances to the nearest RGCs for different years accordingly. “Within or outside the UGBs” for the years 1999 and 2014 is also computed using ArcGIS in the same manner. For the year 2006, we directly obtain the information of “whether a household lives within or outside the UGBs” from the 2006 Puget Sound Household Travel Survey.

Control Variables

Sociodemographic controls

Referring to previous studies, we include eight indicators of personal and household sociodemographic attributes as control variables, as follows: age, gender, education, employment (whether work or not), driver license, household size, household vehicles, and household income. We do not include the race/ethnicity variable because only the 1999 survey contained this information. In addition, no older adults aged above seventy-four were included in the 1999 survey. The education variable is reclassified into three levels, as follows: low (high school or below), medium (bachelor’s degree, associate degree, vocational/technical training, or some college degree), and high (graduate and postgraduate degree).

Self-selection controls

Handy, Cao, and Mokhtarian (2005) point out the necessity of considering self-selection, such as travel attitudes and neighborhood preferences when investigating the relationship between neighborhood characteristics and travel behavior. Because only the 2014 Puget Sound Household Travel Survey contains the self-selection information, we build a separate set of models for the year 2014 to explore the effects of smart growth factors after controlling for self-selection factors. In the 2014 survey, participants were asked to indicate how important some factors are when they chose their current home, on a five-point scale from 1 to 5. We include three factors as self-selection control variables in our analysis, as follows: (1) “being close to public transit”; (2) “being close to the highway”; and (3) “having a walkable neighborhood and being near local activities.”

Analysis and Model

Although we take self-selection into account, showing causality between smart growth and travel behavior is still a challenge. Typically, we first need to demonstrate that smart growth strategies have had impacts on the built environment, of which we tried to give evidence by reviewing the implementation of smart growth strategies and its effects in the region in the previous section. Apart from that, we need to demonstrate that changes to the built environment caused by smart growth have affected travel behavior. However, we do not have the longitudinal data of the same sample who have moved from one place to another (see Cao and Chatman 2016) or who have lived in a place with captured environmental changes. To address this limitation as much as possible, we apply regression models for the years 1999, 2006, and 2014. We try to examine the changes in the effects of smart growth on older adults’ travel behavior over the years.

First, to examine the correlations between smart growth and the daily travel distance of older adults for different purposes (total, utilitarian, and recreational travel), we use the linear regression model. For the dependent variable “travel distance,” a logarithmic transformation (Ln) is used to normalize its distribution to reduce the variance and skewness of the original distribution (Maoh and Tang 2012). Second, we adopt a Poisson regression model to examine the correlations between smart growth and the frequency of older adults’ daily trips for different purposes. Furthermore, we apply a binary logistic model to explore how smart growth affects whether older adults take transit/walk or not. Because two smart growth factors (transit density and land use mix) and the self-selection factors are only available for the year 2014, we separately build models for the year 2014 with these variables included to examine changes in model results compared with models for the year 2014 without these variables included. In all the regression models, the logarithmic transformation (Ln) is applied to the independent variables “accessibility to an RGC” and “housing density” to normalize their distribution. Finally, we examine the relationship between older adults’ workplace/residence and their commuting mode using a chi-square test of independence. We try to explore if working/living within or outside an RGC is associated with older adults’ commuting mode.

Results

Descriptive Summary

Table 1 shows the characteristics of older adults’ daily travel distance and trip frequency for different purposes over the years. From 1999 to 2006, older adults’ average travel distance and trip frequency largely increased, except for the average frequency of recreational trips, which decreased from 1999 to 2006. Compared with 2006, on average, older adults in 2014 have higher number of daily trips and traveled longer distance for recreational activities but shorter distance for utilitarian purpose. In 2014, the average daily travel distance of older adults was 20.71 miles (SD = 31.06; Minimum = 0; Maximum = 334.29), and they made approximately four trips (M = 3.76; SD = 2.93; Minimum = 0; Maximum = 18) per day.

Table 1.

Travel Distance and Trip Frequency of Older Adults in the Sample.

Year	Travel distance (miles)			Trip frequency (number of trips)
Year	Total	Utilitarian	Recreational	Total	Utilitarian	Recreational
1999
M	13.51	9.98	3.53	2.90	1.99	0.91
SD	17.14	13.32	6.19	2.36	1.70	0.95
Min	0	0	0	0	0	0
Max	169.43	161.75	60.53	13	10	7
n	1,797	1,797	1,797	1,797	1,797	1,797
2006
M	19.92	16.19	3.73	3.61	2.81	0.80
SD	21.26	19.13	6.51	2.53	2.34	0.75
Min	0	0	0	0	0	0
Max	154.53	137.19	75.03	17	16	5
n	1,581	1,581	1,581	1,581	1,581	1,581
2014
M	20.71	16.12	4.59	3.76	2.84	0.92
SD	31.06	27.45	10.83	2.93	2.29	1.17
Min	0	0	0	0	0	0
Max	334.29	334.29	120.26	18	13	8
n	1,924	1,924	1,924	1,924	1,924	1,924

Travel Distance and Trip Frequency

Tables 2 and 3 report runs of the linear regression model predicting older adults’ travel distance and the Poisson regression model predicting older adults’ trip frequency, respectively, with a distinction between different purposes. Model results for the three different years are compared. First, after controlling for sociodemographic variables, we find a statistically meaningful relationship between housing density and older adults’ travel distance in 1999 and 2006, but not in 2014. In 1999, we find the coefficient of housing density to be negative and significant at the 10 percent level for total travel distance. For both utilitarian and recreational travel distance, the coefficients of housing density are negative and significant at the 5 percent level. In 2006, higher housing density also correlates with shorter total travel distance (at the 5% level of significance). As for travel for specific purpose, however, housing density only shows a negative and significant correlation (at the 5% level of significance) with utilitarian travel distance but no significant correlation with recreational travel distance.

Table 2.

Results of the Linear Regression Analysis for Travel Distance.

Variables	1999 samples			2006 samples			2014 samples
Variables	T	U	R	T	U	R	T	U	R
Smart growth factors
Ln(housing density)	–0.084*	–0.100**	–0.074**	–0.081**	–0.083**	–0.024	–0.009	–0.022	0.030
Ln(RGC_dist)	0.000	0.006	0.054	0.132***	0.130***	0.031	0.117***	0.104**	0.082**
Within UGBs	–0.313***	–0.247**	–0.067	–0.053	–0.048	0.018	0.061	0.004	0.159*
Sociodemographics
Age
65–74 (Ref.)
75–84	NA	NA	NA	–0.250***	–0.236***	–0.132**	–0.150**	–0.160**	–0.019**
>85	NA	NA	NA	–0.380***	–0.357***	–0.187*	–0.434***	–0.406***	–0.200**
Male	0.036	0.064	–0.011	0.137**	0.152**	0.006	0.104	0.092	0.047
Education
Low	–0.208***	–0.185***	–0.120**	–0.314***	–0.313***	–0.089	–0.263***	–0.203**	–0.254***
Medium (Ref.)
High	0.192**	0.184**	0.058	0.013	–0.018	0.068	0.072	0.053	0.040
Worker	0.345***	0.473***	–0.241***	0.440***	0.513***	–0.244***	0.418***	0.502***	–0.197***
Driver license	0.856***	0.715***	0.525***	0.804***	0.741***	0.277***	0.641***	0.572***	0.250**
Household size	–0.087*	–0.063	–0.150***	–0.053	–0.036	–0.095**	–0.314***	–0.264***	–0.200***
Number of household vehicle	0.006	0.006	0.055*	–0.013	–0.015	0.022	0.149***	0.122***	0.110***
Household income
Level 1	–0.330***	–0.283***	–0.199***	–0.095	–0.074	–0.104	–0.038	–0.010	–0.058
Level 2 (Ref.)
Level 3	–0.154*	–0.170**	–0.090	0.012	–0.030	0.087	–0.087	–0.060	–0.098
Level 4	0.014	–0.001	–0.084	–0.015	–0.021	–0.023	0.098	0.143	–0.111
Sample size	1,797	1,797	1,797	1,581	1,581	1,581	1,924	1,924	1,924

Note: The results of constant were not included for brevity. T = total; U = utilitarian purposes; R = recreational purposes; Ln(RGC_dist) = Ln (distance to the nearest RGC); RGC = regional growth center; UGB = urban growth boundary.

p < .1. **p < .05. ***p < .01.

Table 3.

Results of the Poisson Regression Analysis for Trip Frequency.

Variables	1999 samples			2006 samples			2014 samples
Variables	T	U	R	T	U	R	T	U	R
Smart growth factors
Ln(housing density)	0.018	0.005	0.035	0.038**	0.037**	0.043	0.039***	0.022*	0.095***
Ln(RGC_dist)	0.007	–0.013	0.008	0.027	0.026	0.036	0.028*	0.016	0.069**
Within UGBs	–0.048	–0.017	–0.087	–0.049	–0.058	0.025	0.020	0.007	0.056
Sociodemographics
Age
65–74 (Ref.)
75–84	NA	NA	NA	–0.088***	–0.073***	–0.211***	–0.104***	–0.120***	–0.057
>85	NA	NA	NA	–0.196***	–0.174***	–0.394**	–0.478***	–0.491***	–0.440***
Male	0.046	0.082**	–0.033	0.015**	0.021	–0.033	0.027	0.030	0.021
Education
Low	–0.251***	–0.261***	–0.247***	–0.168***	–0.175***	–0.107	–0.254***	–0.213***	–0.388***
Medium (Ref.)
High	0.106***	0.091**	0.128**	0.061**	0.038	0.248***	0.050*	0.046	0.062
Worker	0.194***	0.344***	–0.186**	0.066*	0.192***	–0.400***	0.110***	0.206***	–0.221***
Driver license	0.774***	0.726***	0.899***	0.333***	0.115***	0.635***	0.422***	0.421***	0.423***
Household size	–0.084***	–0.036	–0.190***	–0.055***	–0.042**	–0.180***	–0.140***	–0.116***	–0.216***
Number of household vehicle	0.041**	0.028	0.070**	0.024	0.022	0.039	–0.003	–0.016	0.039
Household income
Level 1	–0.212***	–0.195***	–0.244***	–0.023	–0.020	–0.050	–0.016	0.012	–0.103
Level 2 (Ref.)
Level 3	–0.008	–0.007	–0.040	–0.039	–0.068**	0.196**	–0.066**	–0.069**	–0.055
Level 4	0.014	0.031	–0.091	0.014	–0.014	0.005	0.071	0.084	0.030
Sample size	1,797	1,797	1,797	1,581	1,581	1,581	1,924	1,924	1,924

p < .1. **p < .05. ***p < .01.

As expected, higher housing density correlates with higher trip frequency for older adults. Contrary to the negative correlation between housing density and travel distance that weakens over time, the positive correlation between housing density and trip frequency grows stronger over the years. Specifically, results for 1999 show no statistically meaningful relationship. In 2006, however, the coefficients of housing density are positive and significant at the 5 percent level for total and utilitarian trip frequency. The year 2014 has a higher coefficient of housing density for total trip frequency (at the 1% level of significance). Notably, in 2014, higher housing density correlates with both higher utilitarian trip frequency (at the 10% level of significance) and higher recreational trip frequency (at the 1% level of significance).

Second, we find positive correlations between the distance to the nearest RGC and older adults’ travel distance as well as trip frequency, meaning that older adults who lived further from the RGC traveled a longer distance and had more daily trips. However, both correlations present changes in the significance and magnitude over the years. In 1999, the distance to the nearest RGC shows no statistically meaningful relationship with older adults’ travel distance. In 2006, the distance to the nearest RGC is positively correlated with older adults’ travel distance, but only for total and utilitarian purposes (at the 1% level of significance). In 2014, although the correlation coefficient for total travel distance gets lower compared with that in 2006, the distance to the nearest RGC becomes positively correlated with both utilitarian and recreational travel distance (at the 5% level of significance). As for the correlation between the distance to the nearest RGC and older adults’ trip frequency, only the year 2014 shows significant results. Specifically, longer distance to the nearest RGC correlated with higher total trip frequency (at the 10% level of significance) and recreational trip frequency (at the 5% level of significance).

Third, we find great changes in the relationship between being within the UGBs and older adults’ travel distance. In 1999, being within the UGBs negatively correlates with older adults’ total travel distance (at the 1% level of significance) and utilitarian travel distance (at the 5% level of significance), showing that older adults who lived within the UGBs traveled a shorter distance, specifically for utilitarian purposes in 1999. In 2006, being within the UGBs presents no statistically significant relationship with older adults’ travel distance. In 2014, we only find the coefficient of being within the UGBs to be positive for the recreational travel distance (at the 10% level of significance), indicating that older adults who lived within the UGBs traveled longer distances for recreational purposes in 2014. However, we find no significant correlation between being within the UGBs and older adults’ trip frequency in all the three years.

As explained in the “Method” section, after comparing model results for the three different years, we further build models for the year 2014 (see Table 4) that include another two smart growth factors (transit density and land use mix) and self-selection variables because these information are only available for 2014. After controlling for sociodemographic attributes and self-selection factors, smart growth factors still show statistically significant relationship with older adults’ travel distance and trip frequency. Compared with previous results, higher housing density no longer correlates with higher utilitarian trip frequency but still correlates with higher total and recreational trip frequency (at the 1% level of significance). Longer distance to the nearest RGC still correlates with longer total and recreational travel distance (at the 10% level of significance) and correlates with higher total and recreational trip frequency (at the 5% level of significance). This indicates that older adults who live further away from the RGC had a longer travel and more trips, specifically for recreational purposes in 2014. The coefficient of being within the UGBs for older adults’ recreational travel distance remains positive and significant at the 10 percent level. Transit density and land use mix, however, show no significant correlation with either travel distance or trip frequency.

Table 4.

Model Results for the Year 2014 After Including Transit Density, Land Use Mix, and Self-selection Variables.

Variables	2014 travel distance			2014 trip frequency			2014 mode
Variables	T	U	R	T	U	R	Transit	Walking
Smart growth factors
Ln(housing density)	0.007	–0.004	0.040	0.034***	0.020	0.079***	0.156	0.091
Transit density	–0.086	–0.097	–0.065	0.004	–0.006	0.030	0.138**	0.387***
Land use mix	–0.003	0.011	0.049	0.040	0.082	–0.083	–0.625	0.606*
Ln(RGC_dist)	0.092*	0.076	0.068*	0.040**	0.026	0.081**	–0.131	0.045
Within UGBs	0.040	–0.016	0.147*	0.010	–0.002	0.047	–0.715	0.077
Sociodemographics
Age
65–74 (Ref.)
75–84	–0.141*	–0.152**	–0.191**	–0.098***	–0.113***	–0.053	–0.320	–0.272*
>85	–0.413***	–0.387***	–0.014	–0.473***	–0.481***	–0.448***	–1.526***	–0.730***
Male	0.099	0.087	0.045	0.034	0.035	0.031	0.202	0.050
Education
Low	–0.267***	–0.209**	–0.262***	–0.254***	–0.214***	–0.383***	0.068	–0.453**
Medium (Ref.)
High	0.087	0.070	0.045	0.036	0.036	0.039	0.163	0.068
Worker	0.421***	0.505***	–0.197***	0.105***	0.202***	–0.229***	0.632***	0.129
Driver license	0.629***	0.559***	0.236**	0.426***	0.428***	0.418***	–0.808***	0.097
Household size	–0.319***	–0.269***	–0.202***	–0.142***	–0.120***	–0.217***	–0.318*	–0.134
Number of household vehicle	0.142***	0.115***	0.105***	0.005	–0.008	0.045	–0.562***	–0.193**
Household income
Level 1	–0.042	–0.017	–0.059	–0.010	0.015	–0.091	0.186	–0.384**
Level 2 (Ref.)
Level 3	–0.090	–0.063	–0.100	–0.066**	–0.068*	–0.057	0.284	0.207
Level 4	0.109	0.153	–0.108	0.074	0.088*	0.029	0.763**	0.835***
Self-selection factors
Res_factor_walk	–0.046*	–0.05*	0.002	0.041***	0.023*	0.101***	–0.059	0.292***
Res_factor_transit	0.017	0.015	–0.002	0.012	0.021*	–0.016	0.558***	0.189***
Res_factor_hwy	0.041	0.035	0.026	0.003	0.001	0.010	–0.258***	–0.143***
Sample size	1,924	1,924	1,924	1,924	1,924	1,924	1,924	1,924

Note: T = total; U = utilitarian purposes; R = recreational purposes; Ln(RGC_dist) = Ln (distance to the nearest RGC); RGC = regional growth center; UGB = urban growth boundary; Res_factor_walk = How important when chose current home: having a walkable neighborhood and being near local activities; Res_factor_transit = How important when chose current home: being close to public transit; Res_factor_hwy = How important when chose current home: being close to the highway. The results of constant were not included for brevity.

p < .1. **p < .05. ***p < .01.

Transit Use and Walking

Table 5 presents the results of the binary logistic model predicting older adults’ transit use and walking. First, as expected, we find housing density positively correlated with older adults’ transit use and walking, although the correlation coefficient changes over the years. Specifically, after controlling for sociodemographic variables, higher housing density correlates with greater odds of transit use in 1999 (at the 5% level of significance) and greater odds of walking in 2006 (at the 1% level of significance). In 2014, higher housing density correlates with greater odds of both transit use (at the 5% level of significance) and walking (at the 1% level of significance). Second, the distance to the nearest RGC has no statistically significant relationship with older adults’ transit use/walking in 1999 and 2006; however, it becomes negatively correlated with older adults’ walking in 2014 (at the 5% level of significance), indicating that older adults who live further from the RGC had smaller odds of walking in 2014. Third, we find no statistically significant correlation between being within the UGBs and older adults transit use/walking in all the three years.

Table 5.

Results of the Binary Logistic Regression Analysis for Transit Use and Walking.

Variables	1999 samples		2006 samples		2014 samples
Variables	Transit use	Walking	Transit use	Walking	Transit use	Walking
Smart growth factors
Ln(housing density)	0.452**	–0.303	0.125	0.336***	0.268**	0.198***
Ln(RGC_dist)	–0.274	0.365	–0.044	–0.002	–0.184	–0.185**
Within UGBs	–0.244	–0.497	–0.131	–0.133	–0.587	0.042
Sociodemographics
Age
65–74 (Ref.)
75–84	NA	NA	–0.212	–0.383**	–0.359	–0.218**
> 85	NA	NA	–0.648**	–0.072	–1.598***	–0.679***
Male	–0.143	–0.894	–0.039	–0.131	0.134	–0.015
Education
Low	–0.582	–0.124	–0.553***	–0.401*	–0.016	–0.481**
Medium (Ref.)
High	0.186	–0.304	0.428***	0.378**	0.237	0.193
Worker	1.704***	–0.231	0.210	–0.145	0.653***	0.157
Driver license	–0.414	0.679	–1.108***	–0.345	–0.913***	–0.037
Household size	0.418*	–0.316	–0.162	–0.193	–0.217	–0.109
Number of household vehicle	–1.602***	–0.195	–0.219**	–0.128	–0.833***	–0.314***
Household income
Level 1	0.647	–0.726	–0.008	0.089	0.237	–0.376**
Level 2 (Ref.)
Level 3	0.904*	–0.599	0.073	–0.231	0.294	0.174
Level 4	0.988	0.970	0.237	–0.067	0.662*	0.806***
Sample size	1,797	1,797	1,581	1,581	1,924	1,924

Note: The results of constant were not included for brevity. Ln(RGC_dist) = Ln (distance to the nearest RGC); RGC = regional growth center; UGB = urban growth boundary.

p < .1. **p < .05. ***p < .01.

Results of a separate model for the year 2014 are also presented in Table 4. After controlling for transit density, land use mix, and self-selection variables, we find housing density and the distance to the nearest RGC losing their significant correlation with older adults’ transit use and walking. As expected, higher transit density correlates with greater odds of older adults’ transit use (at the 5% level of significance) and walking (at the 1% level of significance). Apart from that, we find a positive correlation between land use mix and older adults’ walking (at the 10% level of significance).

Commuting Mode

Table 6 reports runs of a chi-square test of independence to explore if there is an association between older adults’ workplace/residence and their commuting mode over the years. Overall, we find the non-car (transit, walking, and others) mode share among older adults decreases from 1999 to 2006 and increases from 2006 to 2014. Its association with older adults’ workplace and residence in terms of within or outside an RGC are examined separately in Table 6.

Table 6.

Results of the Chi-Square Test of Independence for Workplace/Residence and Commuting Mode.

(a)		Pct. of commuting trips by mode (%)
Workplace	Commuting mode	1999	2006	2014
Within RGC	SOV and HOV	67.9%	66.7%	64.7%
	Transit, walking, and others	32.1%	33.3%	35.3%
Outside RGC	SOV and HOV	86.7%	83.9%	79.2%
	Transit, walking, and others	13.3%	16.1%	20.8%
		Results are significant,	Results are significant,	Results are significant,
		Pearson’s χ² = 8.484,	Pearson’s χ² = 7.313,	Pearson’s χ² = 9.970,
		p < .007	p < .01	p < .003
(b)		Pct. of commuting trips by mode (%)
Residence	Commuting mode	1999	2006	2014
Within RGC	SOV and HOV	73.7%	63.0%	44.1%
	Transit, walking, and others	26.3%	37.0%	55.9%
Outside RGC	SOV and HOV	81.3%	81.9%	82.9%
	Transit, walking, and others	18.7%	18.1%	17.1%
		Results are not significant,	Results are not significant,	Results are significant,
		Pearson’s χ² = 0.628,	Pearson’s χ² = 2.396,	Pearson’s χ² = 139.234,
		p = .537	p = .164	p < .0001

Note: Pearson’s χ² = Pearson’s chi-square; RGC = regional growth center; SOV = single-occupancy vehicle; HOV = high-occupancy vehicle.

As shown in Table 6(a), we find that “whether working within or outside an RGC” is significantly associated with “whether using a non-car commuting mode or not” for older adults in all the three years. Specifically, working within an RGC increases older adults’ non-car commuting mode share by 141.35 percent in 1999, by 106.83 percent in 2006, and by 69.71 percent in 2014. As presented in Table 6(b), we find no significant association between “whether living within or outside an RGC” and “whether using a non-car commuting mode or not” for older adults in 1999 and 2006. However, in 2014, living within an RGC increases older adults’ non-car commuting mode share by 226.90 percent.

Discussion

These results reveal key insights about the effects of smart growth on older adults’ travel behavior. In terms of travel distance and trip frequency, these model results reject our hypotheses that smart growth reduces older adults’ travel distance (H1a) and leads to higher trip frequency (H1b). To be specific, different smart growth factors have different effects. Overall, higher housing density tends to reduce older adults’ travel distance but increases their trip frequency, especially leads to more recreational trips. Previous research suggests that higher housing density may facilitate older adults’ walking via shorter routes to key destinations (Boakye-Dankwa et al. 2019; Chen and Akar 2017), which is consistent with our finding, although we do not focus on older adults’ walking behavior.

As for the distance to the nearest RGC, as expected, older adults who live closer to an RGC appear to have a shorter travel distance. However, we also find that living closer to an RGC lead to fewer recreational trips. While this may be questioned by the previous finding that more recreational trips tend to improve older adults’ life satisfaction (Ravulaparthy, Yoon, and Goulias 2013), it could also indicate that the diversified business and services in the RGCs help meet older adults’ recreational needs by taking fewer trips. Our findings about the RGCs to some extent supports Dalal and Goulias’s (2014) conclusion that business diversification in the RGCs in the central Puget Sound reduces the travel times for residents living within the RGCs.

In addition, we find the effects of living within the UGBs in reducing older adults’ utilitarian travel distance and increasing their recreational travel distance. One would expect a longer utilitarian travel outside the UGBs because suburban housing is often located far from amenities and services (Ettelman et al. 2017). The result of shorter recreational travel outside the UGBs is consistent with the previous research, which indicates that suburban environments often impede older people’s recreational activities and social engagement within their wider community (Zeitler et al. 2012).

Furthermore, land use mix shows no significant effects on older adults’ travel distance and trip frequency. It could be due to the nature of the entropy index, which does not account for detailed information, such as the relative importance of different land uses (Hajna et al. 2014). Transit density also has no significant effects on older adults’ travel distance and trip frequency, as one would expect. Personal vehicles remain the backbone of older adults’ mobility in the United States, which is different from the conditions in some Asian countries (such as China and Singapore), where public transit exerts more influence than automobiles do on older adults’ daily travel (Feng 2017).

The model results support our hypotheses that smart growth promotes older adults’ transit use (H2a) and walking (Hypothesis H2b). Before including transit density, land use mix, and self-selection factors, we find higher housing density promoting older adults’ transit use and walking. We also find that living closer to an RGC facilitates walking for older adults. After controlling for the above-mentioned factors, however, transit density and land use mix replace housing density and the accessibility to an RGC in terms of affecting older adults’ transit use and walking. As expected, higher transit density induces both transit use and walking for older adults, which is consistent with previous studies (Barnes et al. 2016; Cerin et al. 2020). The land use mix modestly facilitates older adults’ walking, which is also supported by previous literature (Nyunt et al. 2015). In addition, our later analysis reveals that working or living within an RGC significantly curbs older adults’ car dependency for commuting.

With respect to our hypothesis that the effects of smart growth on older adults’ travel behavior grow over the years (H3), while we have mixed results for different smart growth factors, some significant growing effects have been confirmed, and the overall changing effects indicate that the impact of smart growth on older adults’ travel behavior can take years. As mentioned before, higher housing density tends to increase older adults’ trip frequency and reduce their travel distance overall. However, the former effect strengthens over the years and the latter effect weakens over time. The role of housing density in affecting older adults’ travel distance was gradually replaced by the RGC. Specifically, living closer to an RGC shows no significant impact in 1999. Over the years, however, it tends to reduce both older adults’ travel distance and trip frequency. Even after controlling for self-selection factors for the year 2014, living closer to an RGC for an older adult still indicates a shorter travel distance and fewer trips for recreational purposes.

As for the association between older adults’ workplace/residence and their commuting mode, working within an RGC always significantly promoted older adults’ non-car commuting mode over the years, whereas it was not until 2014 that living within an RGC showed a promoting effect on older adults’ non-car commuting mode, albeit a more significant one. A regional travel study conducted by PSRC (2015) demonstrates that non-car commuting had increased and commuting by car had decreased since 2006 across the region. Our findings provide empirical evidence that the development of the RGCs create opportunities for older adults to use alternative modes of transportation on their commuting trips.

We also observe great changes in the effect of UGBs on older adults’ travel behavior over the years. Specifically, living within the UGBs showed the effect of reducing older adults’ travel distance for utilitarian purposes in 1999 but not any significant effect in 2006. However, in 2014, living within the UGBs for older adults indicated a longer travel distance for recreational purposes.

Conclusion

Having a built environment that supports independent mobility and provides opportunities for activities for different purposes is regarded as key to healthy aging. A great number of studies have examined the effects of neighborhood environment on older adults’ travel behavior, and much effort has been made to develop walkable, livable, or age-friendly neighborhoods (Nyunt et al. 2015). These neighborhood designing strategies are important local implementations of the broader smart growth urban development concepts. However, how the multilevel and long-term smart growth strategies have an impact on older adults’ travel behavior over time has been rarely explored.

Our empirical evidence in the central Puget Sound reveals the significant influence of the smart growth urban development strategies on older adults’ travel intensity and mode choices. Compact development through the UGBs and higher housing densities reduces older adults’ utilitarian travel distances and facilitates their recreational trips. The RGCs are regional hubs that contain the travel demands of older adults who live inside or nearby. These thriving centers, interlinked with other TOD communities by high-capacity transit systems, attract older adults throughout the region and support alternative modes of transportation and multipurpose trip making (Charles and Symington 2009; PSRC 2010), which is beneficial to both older adults’ health and the environmental sustainability. Moreover, the TOD and transit service improvements can help older adults reduce the risks associated with driving. According to a report conducted by the APTA (2016), the per capita traffic casualty rate of TOD communities is about one-fifth that of car-oriented communities. Seattle, the biggest city both in the central Puget Sound and the Pacific northwest, is one of the representative cities that have relatively low traffic fatality rates because of the pro-transit policies.

In addition, as reflected in our analysis results, the effects of smart growth factors on older adults’ travel behavior have changed over time because the objectives of urban development strategies have been realized gradually. The regional smart growth strategies have been integrated into local regulatory frameworks through an implementation process (Kavage et al. 2005). In our analysis, smart growth variables with statistically significant coefficients imply specific environmental attributes formed by local planning implementation. For example, our models indicate that the RGCs are playing a growing role in containing older adults’ travel demand and decreasing their automobile dependency for commuting trips. In practice, many local jurisdictions in the central Puget Sound have been using the RGC policy as a basis for their comprehensive planning, and they have made substantial progress in planning, designing, and construction in these centers to support the development goals set by the region (Margerum et al. 2013; Porter, ten Siethoff, and Smith 2005). Even in areas with strong implementation, observing changes in the built environment and travel can take a decade or so (Porter, ten Siethoff, and Smith 2005). However, without strategies that unfold with consistency over a long period and that cover a metropolitan region, it is unlikely that these efforts will be successful in changing the course of urban development (Filion and McSpurren 2007) and further exerting beneficial effects on people’s lives and mobility.

As previously mentioned, this study is subject to several limitations. First, we used different samples collected in three different years due to resource constraint. A longitudinal data of the same sample with corresponding information about environmental changes would be helpful to validate the causality. Second, although we built a separate set of models for the year 2014 to control for self-selection factors, we were not able to control for these factors in the comparison between different years due to data limitation. The possible spurious relationship (Handy, Cao, and Mokhtarian 2005) between smart growth and older adults’ travel behavior was not completely avoided to validate the causality. Third, although our findings help reflect changes in the effects of smart growth strategies, we did not probe into how these strategies were unfolded or improved over time at regional and local levels. The whole connection between the smart growth strategies, multilevel implementations, and the consequences in terms of the built environment and human behavior still merits further exploration.

Footnotes

Acknowledgements

We thank Neil Kilgren from the Puget Sound Regional Council for granting us access to the 2006 and 1999 household survey databases. We also thank the three anonymous reviewers for their thorough reviews of the manuscript and constructive suggestions for improvement of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Xueyin Bai

Wei Zhai

Author Biographies

Xueyin Bai is a PhD student and research assistant in the Center for Health and the Built Environment in the School of Landscape Architecture and Planning at the University of Florida. Her research interests include transport geography, built environment and public health, transportation-disadvantage populations, and aging society.

Ruth L. Steiner, PhD, is a professor and director of the Center for Health and the Built Environment in the School of Landscape Architecture and Planning at the University of Florida. Her research interests include the coordination of transportation and land use, school siting and children’s travel, transportation and aging, and the incorporation of risk into long-range transportation planning.

Wei Zhai, PhD, is an assistant professor in the Department of Geography in the Faculty of Social Science at Hong Kong Baptist University. His research interests include urban resilience, spatial data science, and transport geography.

References

Abbott

R. D.

White

L. R.

Ross

G. W.

Masaki

K. H.

Curb

J. D.

Petrovitch

. 2004. “Walking and Dementia in Physically Capable Elderly Men.” Journal of the American Medical Association 292 (12): 1447–53.

American Public Transportation Association. 2016. “The Hidden Traffic Safety Solution: Public Transportation.” Washington, DC: American Public Transportation Association. https://www.apta.com/wp-content/uploads/Resources/resources/reportsandpublications/Documents/APTA-Hidden-Traffic-Safety-Solution-Public-Transportation.pdf.

Bai

Zhai

Steiner

R. L.

. 2020. “Exploring Extreme Commuting and Its Relationship to Land Use and Socioeconomics in the Central Puget Sound.” Transportation Research Part D: Transport and Environment 88:102574.

Banister

2011. “Cities, Mobility and Climate Change.” Journal of Transport Geography 19 (6): 1538–46.

Barnes

Winters

Ste-Marie

McKay

Ashe

M. C.

. 2016. “Age and Retirement Status Differences in Associations between the Built Environment and Active Travel Behaviour.” Journal of Transport & Health 3 (4): 513–22.

Boakye-Dankwa

Barnett

Pachana

N. A.

Turrell

Cerin

. 2019. “Associations between Latent Classes of Perceived Neighborhood Destination Accessibility and Walking Behaviors in Older Adults of a Low-Density and a High-Density City.” Journal of Aging and Physical Activity 27 (4): 553–64.

Boschmann

E. E.

Brady

S. A.

. 2013. “Travel Behaviors, Sustainable Mobility, and Transit-Oriented Developments: A Travel Counts Analysis of Older Adults in the Denver, Colorado Metropolitan Area.” Journal of Transport Geography 33:1–11.

Cao

Chatman

. 2016. “How Will Smart Growth Land-Use Policies Affect Travel? A Theoretical Discussion on the Importance of Residential Sorting.” Environment and Planning B: Planning and Design 43 (1): 58–73.

Cao

Mokhtarian

P. L.

Handy

S. L.

. 2010. “Neighborhood Design and the Accessibility of the Elderly: An Empirical Analysis in Northern California.” International Journal of Sustainable Transportation 4 (6): 347–71.

10.

Carp

F. M.

1988. “Significance of Mobility for the Well-Being of the Elderly.” Transportation in an Aging Society: Improving Mobility and Safety of Older Persons 2:1–20.

11.

Carrapatoso

Silva

Colaço

Carvalho

. 2018. “Perceptions of the Neighborhood Environment Associated with Walking at Recommended Intensity and Volume Levels in Recreational Senior Walkers.” Journal of Housing for the Elderly 32 (1): 26–38.

12.

Cerin

Barnett

Zhang

C. J.

Lai

P. C.

Sit

C. H.

Lee

R. S.

. 2020. “How Urban Densification Shapes Walking Behaviours in Older Community Dwellers: A Cross-sectional Analysis of Potential Pathways Of Influence.” International Journal of Health Geographics 19 (1): 1–18.

13.

Cerin

Sit

C. H.

Barnett

Johnston

J. M.

Cheung

M. C.

Chan

W. M.

. 2014. “Ageing in an Ultra-dense Metropolis: Perceived Neighbourhood Characteristics and Utilitarian Walking in Hong Kong Elders.” Public Health Nutrition 17 (1): 225–32.

14.

Cervero

Kockelman

. 1997. “Travel Demand and the 3Ds: Density, Diversity, and Design.” Transportation Research. Part D, Transport and Environment 2 (3): 199–219.

15.

Charles

Symington

. 2009. “From Barriers to Solutions and Best Practices: Urban Centers and TOD in Washington by the Runstad Center for Real Estate Studies for The Quality Growth Alliance.” Reconnecting America. https://www.crwolfelaw.com/downloads/From%20Barriers%20to%20Solutions.pdf.

16.

Chen

Y. J.

Akar

. 2017. “Using Trip Chaining and Joint Travel as Mediating Variables to Explore the Relationships among Travel Behavior, Socio-demographics, and Urban Form.” Journal of Transport and Land Use 10 (1): 573–88.

17.

Cheng

Chen

Yang

Cao

De Vos

Witlox

. 2019. “Active Travel for Active Ageing in China: The Role of Built Environment.” Journal of Transport Geography 76:142–52.

18.

Chudyk

A. M.

Winters

Moniruzzaman

Ashe

M. C.

Gould

J. S.

McKay

. 2015. “Destinations Matter: The Association between Where Older Adults Live and Their Travel Behavior.” Journal of Transport & Health 2 (1): 50–57.

19.

Cui

Loo

B. P.

Lin

. 2017. “Travel Behaviour and Mobility Needs of Older Adults in an Ageing and Car-Dependent Society.” International Journal of Urban Sciences 21 (2): 109–28.

20.

Dalal

Goulias

K. G.

. 2014. “An Analysis of Jobs, Business Establishments, Housing, and Commuting Travel in the Smart Growth Centers of the Puget Sound Region.” https://www.semanticscholar.org/paper/An-Analysis-of-Jobs-%2C-Business-Establishments-%2C-%2C-Dalal/f55d869483a4d18773638ad163b1392d55305211.

21.

Dierwechter

2014. “The Spaces that Smart Growth Makes: Sustainability, Segregation, and Residential Change across Greater Seattle.” Urban Geography 35 (5): 691–714.

22.

Dierwechter

Carlson

. 2007. “Effects of Urban Growth Boundaries on Residential Development in Pierce County, Washington.” The Professional Geographer 59 (2): 209–20.

23.

Ettelman

Moran

Cardenas

Hudson

J. G.

Minot

Lasley

. 2017. Identifying Transportation Solutions that Promote Healthy Aging for Texas. No. PRC 17-83 F. College Station: Texas A&M Transportation Institute.

24.

Feng

2017. “The Influence of Built Environment on Travel Behavior of the Elderly in Urban China.” Transportation Research Part D: Transport and Environment 52:619–33.

25.

Filion

McSpurren

. 2007. “Smart Growth and Development Reality: The Difficult Co-ordination of Land Use and Transport Objectives.” Urban Studies 44 (3): 501–23.

26.

Freeland

A. L.

Banerjee

S. N.

Dannenberg

A. L.

Wendel

A. M.

. 2013. “Walking Associated with Public Transit: Moving toward Increased Physical Activity in the United States.” American Journal of Public Health 103 (3): 536–42.

27.

Gonda

1982. “Transportation, Perceived Control, and Well-Being in the Elderly.” Specialized Transportation Planning and Practice 1 (1): 61–72.

28.

Hajna

Dasgupta

Joseph

Ross

N. A.

. 2014. “A Call for Caution and Transparency in the Calculation of Land Use Mix: Measurement Bias in the Estimation of Associations between Land Use Mix and Physical Activity.” Health & Place 29:79–83.

29.

Handy

Cao

Mokhtarian

. 2005. “Correlation or Causality between the Built Environment and Travel Behavior? Evidence from Northern California.” Transportation Research Part D: Transport and Environment 10 (6): 427–44.

30.

Haustein

2012. “Mobility Behavior of the Elderly: An Attitude-Based Segmentation Approach for a Heterogeneous Target Group.” Transportation 39 (6): 1079–1103.

31.

Hepinstall-Cymerman

Coe

Hutyra

L. R.

. 2013. “Urban Growth Patterns and Growth Management Boundaries in the Central Puget Sound, Washington, 1986–2007.” Urban Ecosystems 16 (1): 109–29.

32.

Julien

Richard

Gauvin

Fournier

Kestens

Shatenstein

Daniel

Mercille

Payette

. 2015. “Transit Use and Walking as Potential Mediators of the Association between Accessibility to Services and Amenities and Social Participation among Urban-Dwelling Older Adults: Insights from the VoisiNuAge Study.” Journal of Transport & Health 2 (1): 35–43.

33.

Jun

M. J.

2008. “Are Portland’s Smart Growth Policies Related to Reduced Automobile Dependence?” Journal of Planning Education and Research 28 (1): 100–107.

34.

Kan

H. Y.

Forsyth

Molinsky

. 2020. “Measuring the Built Environment for Aging in Place: A Review of Neighborhood Audit Tools.” Journal of Planning Literature 35 (2): 180–94.

35.

Kang

Moudon

A. V.

Hurvitz

P. M.

Saelens

B. E.

. 2017. “Differences in Behavior, Time, Location, and Built Environment between Objectively Measured Utilitarian and Recreational Walking.” Transportation Research Part D: Transport and Environment 57:185–94.

36.

Kavage

S. E.

Moudon

A. V.

Mabry

J. E.

Pergakes

. 2005. “Transportation-Efficient Land Use Regulations and Their Application in the Puget Sound Region, Washington.” Transportation Research Record 1902 (1): 44–54.

37.

Kim

2003. “Analysis of Elderly Mobility by Structural Equation Modeling.” Transportation Research Record 1854 (1): 81–89.

38.

Lee

J. H.

Tan

T. H.

. 2019. “Neighborhood Walkability or Third Places? Determinants of Social Support and Loneliness among Older Adults.” Journal of Planning Education and Research. Published electronically August 14. doi:10.1177/0739456X19870295.

39.

Maoh

Tang

. 2012. “Determinants of Normal and Extreme Commute Distance in a Sprawled Midsize Canadian City: Evidence from Windsor, Canada.” Journal of Transport Geography 25:50–57.

40.

Margerum

R. D.

Brody

Parker

McEwen

. 2013. “Metropolitan Smart-Growth Centers: An Assessment of Incentive Policies in Four Regions.” Journal of Transport and Land Use 6 (2): 21–32.

41.

Nathan

Pereira

Foster

Hooper

Saarloos

Giles-Corti

. 2012. “Access to Commercial Destinations within the Neighbourhood and Walking among Australian Older Adults.” International Journal of Behavioral Nutrition and Physical Activity 9 (1): 133.

42.

National Highway Traffic Safety Administration. 2020. “Traffic Safety Facts Annual Report Tables.” National Highway Traffic Safety Administration, June 30. https://cdan.nhtsa.gov/tsftables/tsfar.htm.

43.

Nyunt

M. S. Z.

Shuvo

F. K.

Eng

J. Y.

Yap

K. B.

Scherer

Hee

L. M.

Chan

S. P.

T. P.

. 2015. “Objective and Subjective Measures of Neighborhood Environment (NE): Relationships with Transportation Physical Activity among Older Persons.” International Journal of Behavioral Nutrition and Physical Activity 12 (1): 108.

44.

Porter

C. D.

ten Siethoff

Smith

. 2005. “Impacts of Comprehensive Planning and Smart Growth Initiatives on Transportation: Case Studies of Five Regions.” Transportation Research Record 1902 (1): 91–98.

45.

Puget Sound Regional Council. 2010. “Vision 2040 Plan.” Seattle, WA: Puget Sound Regional Council.

46.

Puget Sound Regional Council. 2013. “Implementing Equitable Transit Communities: Regional & Local Recommendations for the Central Puget Sound Region.” https://www.psrc.org/sites/default/files/gtc_implementation_report_se011116withblankpage.pdf.

47.

Puget Sound Regional Council. 2015. “PSRC’S 2014 Regional Travel Study.” Puget Sound Regional Council, June. https://www.psrc.org/sites/default/files/regionaltravelsurveyscomparison.pdf.

48.

Puget Sound Regional Council. 2018. “Aging Population, Smaller Households in Region.” Puget Sound Regional Council, April 18. https://www.psrc.org/whats-happening/blog/aging-population-smaller-households-region.

49.

Puget Sound Regional Council. 2019. “How Do Older Adults Get around the Region?” Puget Sound Regional Council, June 11. https://www.psrc.org/whats-happening/blog/how-do-older-adults-get-around-region.

50.

Puget Sound Regional Council. 2020. “Vision 2050.” Puget Sound Regional Council, October 29. https://www.psrc.org/sites/default/files/vision-2050-plan.pdf.

51.

Ravulaparthy

Yoon

S. Y.

Goulias

K. G.

. 2013. “Linking Elderly Transport Mobility and Subjective Well-Being: A Multivariate Latent Modeling Approach.” Transportation Research Record 2382 (1): 28–36.

52.

Rosenbloom

2001. “Sustainability and Automobility among the Elderly: An International Assessment.” Transportation 28 (4): 375–408.

53.

Rosenbloom

2009. “Meeting Transportation Needs in an Aging-Friendly Community.” Generations 33 (2): 33–43.

54.

Schramm

Jennifer

. 2018. “An Aging Labor Force and the Challenges of 65+ Jobseekers.” AARP Public Policy Institute, September 17. https://www.aarp.org/content/dam/aarp/ppi/2018/09/an-aging-labor-force-and-the-challenges-of-sixty-five-plus-jobseekers.pdf.

55.

Siren

Haustein

. 2013. “Baby Boomers’ Mobility Patterns and Preferences: What Are the Implications for Future Transport?” Transport Policy 29:136–44.

56.

Siren

Hjorthol

Levin

. 2015. “Different Types of Out-of-Home Activities and Well-Being amongst Urban Residing Old Persons with Mobility Impediments.” Journal of Transport & Health 2 (1): 14–21.

57.

Smith

T. C.

Wingard

D. L.

Smith

Kritz-Silverstein

Barrett-Connor

. 2007. “Walking Decreased Risk of Cardiovascular Disease Mortality in Older Adults with Diabetes.” Journal of Clinical Epidemiology 60 (3): 309–17.

58.

U.S. Environmental Protection Agency. 2011. “Smart Growth—A Guide to Developing and Implementing Greenhouse Gas Reductions Programs.” U.S. Environmental Protection Agency. https://www.epa.gov/sites/production/files/2017-06/documents/sm_growth_guide.pdf.

59.

Vincent

G. K.

Velkoff

V. A.

. 2010. “The Next Four Decades: The Older Population in the United States: 2010 To 2050.” No. 1138. U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau.

60.

Weitz

1999. Sprawl Busting: State Programs to Guide Growth. Chicago: Planners Press.

61.

Yang

Rodriguez

D. A.

Michael

Zhang

. 2018. “Active Travel, Public Transportation Use, and Daily Transport among Older Adults: The Association of Built Environment.” Journal of Transport & Health 9:288–98.

62.

Zandieh

Martinez

Flacke

Jones

Van Maarseveen

. 2016. “Older Adults’ Outdoor Walking: Inequalities in Neighbourhood Safety, Pedestrian Infrastructure and Aesthetics.” International Journal of Environmental Research and Public Health 13 (12): 1179.

63.

Zeitler

Buys

Aird

Miller

. 2012. “Mobility and Active Ageing in Suburban Environments: Findings from In-Depth Interviews and Person-Based GPS Tracking.” Current Gerontology and Geriatrics Research 2012.

64.

Zhai

Bai

Peng

Z. R.

. 2019. “From Edit Distance to Augmented Space-Time-Weighted Edit Distance: Detecting and Clustering Patterns of Human Activities in Puget Sound Region.” Journal of Transport Geography 78:41–55.